0% found this document useful (0 votes)

401 views288 pages

Masonry Construction & Maintenance

Masonry Constr & Maintenance

Uploaded by

florinvaradi8

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

401 views288 pages

Masonry Construction & Maintenance

Masonry Constr & Maintenance

Uploaded by

florinvaradi8

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008

Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
STP992

Masonry: Materials, Design,

Construction, and Maintenance

Harry A. Harris, editor

ASTM
1916 Race Street
% Philadelphia, PA 19103

Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
Libnuy of Congivss Cataloging-in-Publication Data
Masonry: materials, design, construction, and maintenance/Harry A.
Harris, editor.
(STP; 992)
Papers from a symposium held in New Orleans, LA, 2 Dec. 1986 and
sponsored by ASTM Committees C-7 on Lime, C-t2 on Mortars for Unit
Masonry, and C-15 on Manufactured Masonry Units.
Includes bibliographies and indexes.
"ASTM publication code number (PCN) 04-992000-07."
ISBN 0-8031-1168-1
1. Masonry—Congresses. I. Harris, Harry A. 11. ASTM Committee
C-7 on Lime. III. ASTM Committee C-12 on Mortars for Unit Masonry.
IV. ASTM Committee C-15 on Manufactured Masonry Units. V. Series:
ASTM special technical publication; 992.
TA670.M3781988 88-15446
693.1-dcl9 CIP

Copyright © by AMERICAN SOCIETY FOR TESTING AND MATERIALS 1988

NOTE
The Society is not responsible, as a body,
for the statements and opinions
advanced in this publication.

Peer Review Policy

Each paper published in this volume was evaluated by three peer reviewers. The authors
addressed all of the reviewers' comments to the satisfaction of both the technical editors) and
the ASTM Committee on Publications.
The quality of the papers in this publication reflects not only the obvious efforts of the authors
and the technical editor(s), but also the work of these peer reviewers. The ASTM Committee on
Publications acknowledges with appreciation their dedication and contribution of time and ef-
fort on behalf of ASTM.

Printed in Baltimore. MD
July 1988

This publication, Masonry: Materials, Design, Construction, and Maintenance, contains pa-
pers presented at the symposium of the same name held in New Orleans, LA on 2 Dec. 1986.
The symposium was sponsored by ASTM Committees C-7 on Lime, C-12 on Mortars for Unit
Masonry, and C-15 on Manufactured Masonry Units. Harry A. Harris, Ash Grove Cement Co.,
presided as symposium chairman and was editor of this publication.

MATERIALS

Inadequacy of Property Specifications in ASTM C 270—G. C. ROBINSON

AND R. H. BROWN 7

Shear Strength of Masonry Piers Under Cyclic Loading—L. S. P. JOHAL

AND E. D. ANDERSON 18

Investigation of Masonry Bond and Surface Profile of Brick—7. w. RIBAR

AND V. S. DUBOVOY 33

DESIGN

Preconstruction Brick Veneer Evaluation Testing—D. C. RATHS 41

Problems and Solutions to Masonry Buildings—R. M. GENSERT AND W. C. BRETNALL 57

Discussion 91

Influence of Tie Flexibility, Relative Length, and End-Boundary Condition on Brick

Veneer-Metal Stud Flexural Bond Stress—i. R. CHIN, N. V. KROGSTAD, AND
C. B. MONK, JR. 96

Analysis and Test of a Torsional Sensitive C-Shaped Prefabricated Brick Spandrel

Panel—R. c. ARNOLD, L. J. DONDANVILLE, N. V. KROGSTAD, AND C. B. MONK, JR. 118

Test of Model Masonry Single Pier Under Dynamic Shaking and Qoasistatic
Cyclic Loading—H.-L. CHEN AND S. P. SHAH 145

CONSTRUCTION

Statistical Primer for Brick Masonry—c. T. GRIMM 169

An Investigation of an Extended Plastic Life Mortar—j. H. MATTHYS, i. r. HOUSTON,

AND A. DEHGHANI 193

Tlie Effectiveness of Waterproofing Coatings, Surface Grouting, and Tuckpointing

on a Specific Project—w. B. CONEY AND J. G. STOCKBRIDGE 220

Diagnosis of Terra-Cotta Glaze Spalling—s. E. TBOMASEN AND C. L. SEARLS 227

In-Situ Evaluation of Compressive Stresses—D. MANMOHAN, R. L. SCHWEIN,

AND L. A. WYLLIE, JR. 237

Restoration of the John J. Glessner House—T. SOURLIS 251

Masonry Craclts: A Review of the Literature—c. T. GRIMM 257

Indexes 281

Overview

This book stems from the fifth in a series of symposia on masonry sponsored by ASTM Com-
mittees C-7 on Lime, C-12 on Mortars for Unit Masonry, and C-15 on Manufactured Masonry
Units. Like those that have preceded it, this symposium provided a forum for the dissemination
and exchange of information and experiences related to masonry construction.
Three of the four preceding symposia were also published by ASTM:
STP 589—Masonry: Past and Present, published in August 1975, was the first in this series.
It provided the basis for future symposia by reviewing the specifications and test methods from a
historical perspective. Research and new developments in the field of masonry construction
were also covered.
STP 778—Masonry: Materials, Properties, and Performance, published in September 1982,
covered the third symposium on masonry. This publication presented a forum for research on
masonry units, mortar and grout (including their components), and masonry assemblages.
STP 871—Masonry: Research, Application, and Problems, published in April 1985, covered
the fourth symposium. The objective of this symposium was to cover field applications, end-use
problems, and research.
The second symposium in this series was not published except for several papers appearing in
ASTM's Journal of Testing and Evaluation. The scope of this symposium was similar to that of
the first.
For the current symposium. Masonry: Materials, Design, Construction, and Maintenance,
papers dealing with current technology in each of these four major areas of masonry were re-
quested. The areas were selected to provide general coverage of current developments in the
industry and thus provide an update to previous publications on this subject.

Materiak
The first three papers relate to testing procedures and properties of masonry materials and
assemblages. New test procedures are described and data presented on bond and shear
strengths of masonry assemblages. These papers will provide new guidelines for writing future
specifications and codes for masonry materials.
Robinson and Brown examine the existing C 270 requirements (ASTM Specification for Mor-
tar for Unit Masonry) for mortar and its shortcomings. Test data on the bond strengths of
several C 270 mortars are offered as a basis of writing a new performance specification.
Johal and Anderson have evaluated masonry cement mortars when used in the construction
of shear-wall specimens. Both static and cyclic load tests were conducted in this investigation of
concrete masonry block and clay brick walls.
Ribar and Dubovoy have explored the surface characteristics of brick. A new technique for
measuring surface characteristics provides important new insights into factors controlling bond
and shear strengths. Additional research on surface characteristics may provide a means of
evaluating the performance of each material in a masonry assemblage.

Design
Design of masonry construction is dealt with in the next five papers. Subjects range from the
detailing of tie systems to the testing of assemblages under various types of simulated loading.

1
Copyright by ASTM
Copyright' 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
2 MASONRY

This important work will provide the means for better and safer construction even under the
extreme conditions found during an earthquake.
Raths has presented a preconstruction testing program for the selection of materials used in a
brick-veneer cavity wall. The program uses ASTM standards to evaluate materials and their
compatibility as a means of preventing both construction problems and unsatisfactory perfor-
mance.
Gensert and Bretnall have documented the construction and performance of a masonry struc-
ture by means of photography and computer graphics. Methods of analyzing architectural de-
tails and the interactions between masonry and structural frame works are shown.
Chin et al. have examined the relative stiffness between brick veneer and metal studs and its
effects upon wall design. This paper shows that, by using shorter length metal studs, critical
flexural bond stresses do not exist on tjrpical brick veneer/metal stud walls under designed load-
ing conditions.
Arnold et al. compare two methods of analysis for the design of brick spandrel panels. Tor-
sion stresses during placement were of particular interest in this study.
Chen and Shah have studied methods of improving seismic design of masonry structures by
testing masonry single pier models. The behavior of these piers was studied under dynamic
shaking and slowly applied cyclic loading.

Construction
The section on construction provides a direct link between the researcher and field applica-
tion. Three papers are offered in this category, each dealing with a different subject but all of
value to both the university laboratory as well as the masonry contractor.
Grimm's first paper is a review of methods for sampling and statistical data reduction for
brick masonry. Based on the techniques described, a concept of structural reliability is intro-
duced.
Matthys et al. present data on extended life or ready-mix mortars. The paper describes a
study in optimizing a mix to achieve specific strength and setting characteristics. The optimized
mix is then compared to standard portland cement lime Type N mortar for mechanical proper-
ties.
Coney and Stockbridge address water leakage problems through a study of waterproofing
coatings, surface grouting, and tuckpointing. Field studies were conducted prior to and during
the repairs of a building.

Maintenance
The preservation of our national heritage is gaining greater interest year by year, and the
most outstanding symbols of our past are found in masonry structures. To this end we devoted
the final section of this symposia to maintenance and rehabilitation. The four papers presented
here cover unique reviews of major rehabilitation projects. Details are presented on the materi-
als and methods used, where and how failures in the original construction occurred, and how
failures were corrected. These papers will be of great interest and value to those involved in this
area of masonry work.
Thomasen and Searls describe the deterioration of terra-cotta claddings and their repair.
Special emphasis is given to the cause of glaze spalling as related to repair and prevention
measures.
Manmohan et al. have studied the compressive stresses in terra-cotta cladding due to frame
shorting. Methods of relieving these stresses by cutting the bed and head joints are described.
Sourlis presents a detailed review of the restoration of a 100-year-old historical masonry
structure. All aspects of the project from bidding the job to final cleanup and landscape repairs
are described.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
OVERVIEW 3

Grimm's second paper is a study of masonry cracks and how they affect the performance of a
masonry structure. The various types of cracks are described and their causes and method for
repair are given.
This publication is the result of the combined efforts of many people. I want to thank the
members of my subcommittee and those assisting with the presentation and review of papers. A
special thanks to those on the ASTM staff who helped guide me through the many stages of this
process from conception to final publication.

Harry A. Harris
Ash Grove Cement Co., Kansas City, KS
66103; symposium chairman and editor.

Inadequacy of Property Specifications in

ASTIVI C 270

REFERENCE: Robinson, G. C. and Brown, R. H., "Inadequacy of Property Specifications In

ASTM C 270," Masonry: Materials, Design, Construction, and Maintenance, ASTM STP 992,
H. A. Harris, Ed., American Society for Testing and Materials, Philadelphia, 1988, pp. 7-17.

ABSTRACT: Bond strength and water leakage were determined for Types M, S, and N mortars
joined to six different brick types. The air content, water retention, and flow rate were varied for
each mortar-brick combination. It was found that air content, water retention, and compressive
strength were inadequate indices of bond strength and water leakage. Yet these are the only prop-
erties specified in ASTM Specification for Mortar for Unit Masonry (C 270-84). A plea is made to
add a bond strength requirement to present property requirements. The initial rate of absorption
(IRA) of brick was found to be significant in predicting water leakage. Low IRA brick provided
the highest probability of zero leakage.

KEY WORDS: masonry, mortar, bond strength, water retention, air content, compressive
strength, water leakage

Consumers of mortars have supported the concept of a performance standard for mortar. As
a consequence, ASTM Specification for Mortar for Unit Masonry (C 270-84) includes a prop-
erty specification as well as a proportion specification. Inclusion of the property specification
option is a move in the right direction, but the designated qualities are inadequate for assuring
quality mortar. It is intended to plea for the addition of a bond strength requirement to present
property requirements and possibly the addition of a water leakage specification.
Property requirements should relate to key performance qualities desired in a mortar. The
most rigorous mortar requirements are to provide adequate and uniform bond strength and to
prevent water leakage. All other mortar functions are easy to fulfill. ASTM C 270 fails to ad-
dress the most demanding mortar qualities and instead specifies only three properties: compres-
sive strength, water retention, and air content. It is intended to present test results that will
question the ability of these three properties to assure quality mortar performance.

Compressive Strength
ASTM C 270 prescribes a minimum compressive strength for each type of mortar. Does this
relate to bond strength or water leakage? Figure 1 shows the insignificance of compressive
strength to bond strength. Three different types of mortar and four to six different types of brick
for each mortar type were evaluated for bond strength. A skilled brick mason constructed a
brick prism consisting of four bricks and three mortar joints for each combination for mortar
type and brick type. The mason proportioned the mortars by volume as prescribed by ASTM C
270. He used conventional field procedures for mixing the mortar and adjusted the mixing wa-
ter until he judged the mixture to be of desired consistency. The prisms were constructed and

'Professor of Ceramic Engineering, Clemson University, Clemson, SO 29634-0907.

^Professor and Head of Civil Engineering, Clemson University, Clemson, SC 29634-0911.

7
Copyright by ASTM
Copyright' 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
8 MASONRY

auu Uortar Type

HZ
w iV —• •• o " M
'
150
o
o 0 ° o
o

i 100
o 0 o o

g 50

o
0 1 ~ I 1 1 I 1 . ...J
500 1000 1500 2000 2500 3000 3500 4000
COMPRESSIVE STRENGTH, psi
FIG. 1—The bond strengths of mortars of various compressive strengths.

then cured in a room environment for 28 days under a cover of plastic sheeting. The bond
strength was determined by the bond wrench method [ASTM Standard Method for Measure-
ment of Masonry Flexural Bond Strength (C 1072-86)]. Each point of Fig. 1 represents the
average results from several brick types as shown in Table 1. The compressive strengths of the
mortars were determined on mortar cubes as instructed by ASTM C 270 and ASTM Test
Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-in. or 50-mm Cube
Specimens) (C 109).
All the bond strengths were within plus or minus 10 psi (69 kPa) of 135 psi (931 kPa). At the
same time, the compressive strength of the mortars varied from 800 to 3800 psi (5.5 to 26.2
MPa). This suggests that compressive strength is valueless as an index of bond strength.

Water Retention
ASTM C 270 specifies that mortars of all types shall have a minimum water retention of 75%.
This seems to make sense because a mortar should hold water and remain fluid against the
suction of the masonry unit, but does it work to assure quality mortar? The results of Fig. 2
suggest that this property too provides little assurance of quality mortar. The figure shows the
relation between water retention and bond strength. Each point is an average of several brick

TABLE 1—The number of brick types and joints tested for

each datum point of Figs. 1 and 2.

Number of
Number of Joints Tested
Type Mortar Brick Types Per Datum Point

Type M 18
Types 15
TypeN 12

en
a,
1501.

I G .
-sr

CO
T®
Q 100 JL
50 60 70 80 90
o
WATER RETENTION, %
FIG. 2—The influence of water retention on bond strength.

types as shown in Table 1. The water retention follows the expected pattern of being higher with
Type N and lowest with Type M. The unexpected result was the little change in bond strength
with large changes in water retention. All of the combinations showed good bond strength, and
yet all of the M mortars were below the prescribed water retention, one being only 45%. Two
out of three of the Type S mortars failed to meet the water retention standard while one of the
Type N mortars was below the standard. Nevertheless, they all gave approximately the same
level of bond strength as the mortars that did meet the specified 75% water retention. There is
even a suggestion with the Type M mortars that substandard water retention provides a slight
improvement in bond strength. This supports the work by Hogberg as cited by Goodwin and
West [1]. Hogberg demonstrated that mortars with low water retentivity gave better bond
strength on absorbent brick. Low water retention coupled with extra water in the mortar
quenches the thirst of high IRA brick and improves bonding. This was demonstrated by Robin-
son and Salmond [2]. It is suggested that three mechanisms may explain the observed results on
water retention.
The first is that the mortar should give up easily its water to high IRA brick in order to nullify
the deleterious effect of high suction rate. Low water retention coupled with high water content
mortars provides for satisfaction of the suction demand of the brick while retaining sufficient
fluidity of the mortar for penetration into the surface structure of the brick. In contrast, increas-
ing mortar water retention through additives can retain the water against the suction of the
brick and retard filtration of the mortar paste into the brick surface.
The second consideration is for low IRA brick. Their limited water suction suggests the desir-
ability of a mortar that would stiffen with very little water removal.
The third consideration applies to all units. Dewatering of the mortar at the brick mortar
interface provides a more favorable water-to-cement ratio for strength development.
Once again the inclusion of the property "water retention" bears little relation to mortar
performance and is harmfully restrictive in designing a mortar-brick combination for optimum
performance.

Air Content
The last quality specified is the air content of the mortar. A maximum air content of 12% is
specified for portland cement lime mortars—Types M and S—and 14% for Type N. In con-
trast, masonry cements are limited to 18% maximum air content only if structural reinforce-
ment is incorporated. Otherwise no maximum is listed, although C 270 does state that the ma-
sonry cement should conform to the requirements of ASTM Specification for Masonry Cement
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
10 MASONRY

(C 91). C 91 lists minimum air contents of 12% by volume and maximum contents of 20% for
Types M and S and 22% for Type N.
The influence of air content on bond strength is shown in Fig. 3. The same procedure was
used as described in the section entitled "Compressive Strength." The results of Fig. 1 were
limited to PCL mortars with less than 5.3% air content. These specimens were included in Fig.
3 together with additional specimens with higher air quantities and mortars made with masonry
cements. There is considerable scatter in results, but there appears to be a trend of decreasing
bond strength with increasing air content. In fact, it appears that the air content is more signifi-
cant than the type of mortar. Thus Types N, S, and M almost overlap at 3% air, and Types N
and M and masonry cement overlap at 12% air. The masonry cement seems to form a continua-
tion of the curve to lower levels of bond strength because of the higher permissible levels of air
content. Thus Type N masonry cement with 18% air (near its maximum limit) showed a bond
strength of about 35 psi (241 kPa), whereas Type N-PCL (portland cement-lime) mortar at 10%
air content (near its maximum limit) showed bond strengths of 75 psi (517 kPa). However, the
masonry cement compared to the PCL at the same air content showed about the same bond
strength. It would seem that masonry cements do themselves a disservice by specifying a mini-
mum air content. This is further illustrated by the bond strength designated as M-MC in Fig. 3.
These results are from a different investigator and were prepared with a Type M masonry ce-
ment in which the air content was controlled by mixing procedure. It will be observed that this
produced somewhat higher bond strengths than the Type M-PCL mortar when compared at
equivalent air content. Furthermore, it seems to follow a similar trend of decreasing bond
strength with increasing air content.
These results suggest that air content is a more meaningful quality indicator than the other
property requirements. However, air content alone is not sufficient to assure mortar quality, as
shown in Fig. 4. This shows the results on seven different mortar batches of similar air content.

150

w
^.100

i
CO

§50
O

10 15 20
AIR CONTENT, %

FIG. 3—The influence of mortar air content on bond strength.

CO o M 18 (Air Content)
D. \
150 \
tq \
E-. V M 18.2
y12; \
^ 100 \
S
K
\ ^
M iflV^N^
^^
Cd
50 \Ready Mix 18
^-^^SOAP 18
oq M 19
0 1 1 1 1 1 L I 1

MORTAR TEST
FIG. 4—Variation in bond strength of mortars with similar air content.

However, the air was introduced by use of different air entraining agents or different proce-
dures. This shows a change from 179 to 30 psi (1234 to 207 kPa) bond strengths for the various
trials. All of the compositions were proportioned to be Type M except for the one labeled S and
Ready-Mix. The Type S was a Type S masonry cement and the Ready-Mix was also a Type S
masonry cement. These results point out that although air content may be significant to bond
strength, it is not sufficient to assure bond strength. Instead, the type of entraining agent and
the pore structure will influence bonding. A fine, uniform bubble structure favors strength
while an irregular structure with large bubbles harms strength. Schmidt [3] reports a similar
conclusion of European investigators of bond strength.
Once again the property specification of C 270 is inadequate. Replacement of the air require-
ment with a bond strength requirement would assure adequate performance and permit ma-
sonry cements and PCL mortars to compete without prejudice.

Brick IRA
The properties of masonry units are not part of the scope of C 270. Nevertheless, bond
strength is influenced by the masonry unit, and a bond strength requirement would evaluate the
compatibility of a particular mortar-brick pair. The initial rate of absorption (IRA) has been
suggested as a key quality of brick. Tests were made to evaluate the bond strength of brick of
different IRAs and Figs. 5 and 6 show typical results. It is interesting that the IRA appears to
have little influence on bond strength.
Other investigators have come to a similar conclusion. Gazzola [4] presented a similar result
in his paper at the last masonry symposium presented in Florida in 1983. Additional results
available from the Holmes Laboratory [ 5] in another mortar investigation had a similar finding.
The reason for the contradiction in results between these investigations and earlier ones is being
explored and will be the subject of another report.

Water Leakage
The topic of water leakage has been avoided because of its complexity and the wide scattering
of results. The same mortar-brick combinations previously presented under "Compressive
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
12 MASONRY

-cuw

=ifc
' f ^

w
•^ 150
=115

^ 100

CO •apv / TK \N j ^ - - ^ '•

§ 50 . * I
rT -
O X
oq ^

0 • 1 -1 1 — 1 — 1
10 20 30 40

INITIAL RATE OF ABSORPTION, g/30 in.'/min.

+ LOW FLOW
» MED. FLOW
=ib HIGH FLOW

FIG. S—Tke relationship between bond strength and IRA for a Type N PCL mortar.

200 r-

20 30 40

INITIAL RATE OF ABSORPTION, g/30 in. '/min.

+ LOW FLOW
* UED. FLOW
%, HIGH FLOW

FIG. b—Tke relationship between bond strength and IRA for a Type N masonry cement mortar.

Strength" were evaluated for their tendency towards leakage of water. This was done by pooling
water to a depth of 2 in. on top of horizontal prisms, then measuring the rate in drop of water
level. Each type of brick was tested with Types N, S, and M mortar and each combination was
tested at low, medium, and high air content and low, medium, and high flow rate.
The average water leakage was calculated for each brick type by combining the results for all
mortar types, flow rates, and air contents. The results are shown in Fig. 7 and the number at
each point is the number of prisms tested for each point. The solid curve indicates the average
water leakage. An attempt was made to assign a number to each trial which would indicate the
inches of water drop occurring per hour. There were two problems with this. First, there would
be an initial period where the water would soak into the assembly but not contribute to leakage
through the assembly. The higher the absorption of the brick, the greater the amount of this
initial water level drop. At the other extreme, trials that approached 2 in. of water leakage in the
6-h test period showed a tailing off of the rate of drop when the water level became 0.5 in. or
less. It would appear that the slope of the curve between these extremes better illustrates the
tendency to water leakage. As a consequence, a numerical value expressing water leakage was
obtained by measuring the drop in water level occurring between 1-h elapsed time and 6-h
elapsed time and dividing by five. Furthermore, if the water level dropped to 0.5 in. prior to 6-h
elapsed time, the reading was taken at this point and the time duration adjusted accordingly. In
the case of severe leakage the water would drop 2 in. in less than an hour, and in this event the
slope was evaluated in the period from zero to 1 h.
In addition to a numerical ranking, the specimens were also given a letter grade with A
awarded to those prisms that showed leakages of less than 0.05 in./h, a B for leakages of 0.05 to
0.10 in./h, etc. The broken line curve of Fig. 7 is the inverse of the first and obtained by a
different method. This curve shows the percentage of nonleakers or A grade trials for each type
of brick.

_ 100

5 10 15 20 25 30

BRICK IRA, g/30 in'/min.

FIG. 7- -The relationship between brick IRA and rate of water leakage or percentage ofnonleaking units
(A grade).
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
14 MASONRY

Both methods of expressing results agree that water leakage is lowest with low IRA brick. It
would appear that IRAs of 10 or less are preferable.
The results of the two methods are distinct for higher IRA brick. The water leakage curve
indicates maximum leakage at 23 g IRA and then decreasing leakage as the IRA increases to 38
g. In contrast, the percentage of A prisms decreases to below 12% for all IRAs of 23 and above.
This would suggest that the rate of water leakage may decrease at higher IRA but that the
probability of a joint showing leakage is high for any value of 23 or above. These results were
obtained for brick without prewetting prior to application of mortar.
The scatter in results at higher IRAs precludes definite conclusions. The results do seem to
support the conviction that brick type and IRA have a major influence on water leakage.
Brick continue to imbibe water from mortar beyond the 1-min limit, and it was wondered how
an extended IRA test would rank the brick type. This is shown in Fig. 8, which plots the grams

D &E

0 12 5 10
ELAPSED TIME. Minutes
FIG. 8—The waterpickup versus elapsed time for the types of brick used in bond strength tests.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
ROBINSON AND BROWN ON PROPERTY SPECIFICATIONS 15

-,00

>-
<

K
UJ
If)
I
o
(0 U UJ
I I 2

O O o O HHO»*t K>oe 00

o o o o O O O O
<M 00 1- O <o M OO «1-
to C\i rJ rj

!«<! 'H19N3aiS ONOa

TABLE 2—The influence of air content and mortar flow on resistance to water leakage.
Type N Mortar by Flow Type M Mortar by Flow
Brick Air
Type IRA Content Low Medium High Low Medium High
Low A A F A B A
18 Medium C C F
High J^ A A C A A
Low D F A D D C
27 Medium C F A
High D C D^ F F A_
Low F B A F C C
2S Medium F D C
High F F £ F F f^

NOTE: A = no leakage (less than 0.05 in./h); F = severe leakage (greater than 0.39 in./h).

for water pickup per net 30 in.^ of surface area against minutes of exposure. It will be noticed
that the bad leaking brick (C, D, E, and F) show an increased gap from the other brick at 10
min. Good Brick G and bad Brick D and E gave similar results at 1 and 2 min but a widening
gap with increased time. This persistent suction appears detrimental to leakage resistance.
The results on water leakage cause a suspicion that mortar flow may be significant. Table 2
abstracts leakage rankings (A, B, C, D, or F) from a larger table. The numerical value of leak-
age was obtained by averaging all the values for each brick type. It will be noticed that high
flows gave the best results for problem Brick C and D; however, a reverse result was obtained
with Brick A of lower IRA. Perhaps the flow should be adjusted from low to high as the brick
IRA increases. Also, this table illustrates the ranking of brick in Fig. 7. Note that the percent-
age of A grades decreases from brick Types A to C to D.

Proposed Change to C 270

Test results have been presented that indicate that the present property specification of
ASTM C 270 is inadequate. It does not provide the consumer with assurance of quality mortar.
Such assurance could be provided if the present requirement for water retention, air content,
and compressive strength were replaced with a bond strength test of the intended mortar-brick
combination. Furthermore, it is suggested that such a test could be performed within 24 h of
forming specimens. Figure 9 shows the relation between bond strength and curing time. It will
be noticed that the majority of bond strength develops within 24 h. Bond wrench testing proce-
dures are easy to apply to field testing as well as to the laboratory. The inclusion of a bond test in
C 270 would be a major step forward in quality assurance to the consumer.

Acknowledgments
Grateful appreciation is expressed to the Masonry Research Foundation for its financial sup-
port of this investigation. Appreciation is expressed also to Wilson U. Johnson for his perfor-
mance of most of the test procedures of this project.

References
[/] Goodwin, I. F. and West, H. W. H., "A Review ot Literature on Brick/Mortar Bond," British Ceramic
Research Association, Stoke-on-Trent, ST47LQ, England, December 1980.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
ROBINSON AND BROWN ON PROPERTY SPECIFICATIONS 17

[2] Robinson, G. C. and Salmond, W. H., work in process, Ceramic Engineering Department, Clemson
University, Clemson, SC, 1986.
[3] Schmidt, S. B., president, Addiment, Inc., Doraville, GA, personal communication.
[4] Gazzola, E., Bagnariol, D., Toneff, J., and Drysdale, R. G., "Influence of Mortar Materials on the
Flexural Tensile Bond Strength of Block and Brick Masonry," Masonry: Research, Application and
Problems. ASTM STP 871, American Society for Testing and Materials, Philadelphia, PA, 1985, pp.
15-25.
[5] "Final Test Results Mortar Life Project," H. H. Holmes Testing Laboratories, Inc., Wheeling, IL, June
1983.

Shear Strength of Masonry Piers Under

Cyclic Loading

REFERENCE: lohal, L. S. P. and Anderson, E. D., "Shear Strength of Masonry Piers Under
Cyclic Loading," Masonry: Materials, Design, Construction, and Maintenance, ASTMSTP992,
H. A. Harris, Ed., American Society for Testing and Materials, Pliiladelphia, 1988, pp. 18-32.

ABSTRACT: This paper describes a test program to evaluate shear behavior of masonry piers
constructed with portland cement and masonry cement-based mortars and tested under simu-
lated seismic loading. Both clay brick and concrete masonry block pier specimens were subjected
to in-plane cyclic shear stresses. Several brands of masonry cement and portland cement were used
in fabrication of 32 specimens. The pier portion of specimens was minimally reinforced for flexure
and partially grouted. No shear reinforcement was provided. The specimens were loaded in a
manner to eliminate any axial force in the pier cross section. Results indicate that the shear
strength of masonry piers constructed with masonry cement-based mortars compared favorably
with those constructed with portland cement-based mortars. Values of maximum shear stress
obtained with portland cement and masonry cement-based mortars were close and within the
range of scatter normally expected in experimental data. The shear strength was not influenced
significantly by the use of Types M or S mortars.

KEY WORDS: masonry, masonry piers, shear, shear stress, shear strength, cyclic loading, ma-
sonry cement-based mortar, portland cement-based mortar, Type M mortar. Type S mortar

In masonry construction, shear walls are often used to resist lateral loads through in-plane
stresses. Shear behavior of masonry, therefore, is a significant factor influencing the overall
response of masonry structures. During the last three decades, a limited number of experimen-
tal programs have attempted to identify and investigate parameters affecting the shear strength
of masonry. One of the parameters that has not been adequately investigated is the type of
cement used in the mortar. Both portland cement and masonry cement can be used to achieve a
selected mortar compressive strength. Almost all experimental programs on shear strength,
however, have been based on mortars containing a combination of portland cement and lime. A
shortage of test data on the behavior of masonry walls, using masonry cement-based mortars,
has led to regulatory constraints on the use of masonry cement. The current edition of the Uni-
form Building Code (UBC) does not allow the use of masonry cement in structural frames for
Seismic Zone Nos. 2, 3, and 4.
A test program was undertaken to expand the data base on the behavior of masonry walls.
Both clay brick and concrete masonry block pier specimens were subjected to in-plane cyclic
shear stresses. The overall objective of the program was to evaluate and compare shear behavior
of masonry piers constructed with portland cement and masonry cement-based mortars and
tested under simulated seismic loading. Several brands of masonry cement and portland cement
were used in the fabrication of 32 pier-type masonry shear wall specimens. Sixteen of these
specimens were constructed with hollow clay brick and 16 were constructed with concrete block.

'Senior structural engineer and associate structural engineer, respectively, Structural Experimental Sec-
tion, Construction Technology Laboratories, Skokie, XL.

18
Copyright by ASTM
Copyright' 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
JOHAL AND ANDERSON ON MASONRY PIERS 19

The test region (pier portion) of specimens was minimally reinforced and partially grouted so
that the shear strength of masonry is primarily influenced by the mortar. Only the outer pier
cells were reinforced with a single vertical bar and grouted. In order to determine the shear
strength of masonry only, no horizontal shear reinforcement was provided in the pier portion.
In addition, loads were applied in a manner to eliminate the application of axial load to the pier
portion. This provided lower bound values of shear strength. Shear behavior of the specimens
was then compared to evaluate the performance of different cement t)rpes.

Experimental Program
The experimental program included two categories of mortars. The first category included
mortars obtained from Types M and S masonry cements. The second category had Types M and
S mortars prepared from a combination of portland cement and hydrated lime, with no ma-
sonry cement included. Three representative brands each of Types M and S masonry cement
were used to prepare three mixes each of M and S mortar. One mix each of Types M and S
mortar was prepared by blending three brands of portland cement with lime. Both hollow clay
bricks and hollow concrete blocks were included. Running bond construction pattern was used
throughout this program.

Test Matrix
The test matrix is presented in Table 1. Three representative brands of T3rpe M masonry
cement designated MCI, MC2, and MC3 were used to prepare three mixes of Type M mortar.
Three representative brands of Type S masonry cement designated SCI, SC2, and SC3 were
used to prepare three mixes of Type S mortar. Three representative brands of portland cement
were used to form a blended sample designated PCI. One mix each of Type M and Tjp^ S
mortar was obtained by mixing blended portland cement with hydrated lime.

Specimens
Thirty-two I-shaped pier type specimens were tested. This included 16 specimens constructed
with hollow clay brick and 16 with hollow concrete block. Specimens were cured for at least 28

TABLE \—Test matrix.

Specimen Designation
Mortar Cement
Cementitious Brand Mortar Mix Hollow Clay Hollow Concrete
Materials Designation Type Proportions" Brick Pier Block Pier

Masonry cement. MCI M 0:l:0:2>/2* CMl, CM2 DM1, DM2

Type M MC2 M 0:1:0:21/2* CM3, CM4 DM3, DM4
MC3 M 0:1:0:3 CMS, CM6 DM5, DM6
Masonry cement. SCI S CSl, CS2 DSl, DS2
Types SC2 S 0:1:0:3 CS3, CS4 DS3, DS4
SC3 S CSS, CS6 DSS, DS6
Blended portland PCI M 1:0:1/4:33/4 CPl, CP2 DPI, DP2
cement and lime S 1:0:1/2:41/2 CP3, CP4 DP3, DP4

"Mix proportions are expressed as proportions of portland cement: masonry cement: lime: fine aggregate
(by volume).
'Mix proportions used as specified by the manufacturer.

days in a laboratory maintained at 23° ± 3°C (73° ± 5°F) and 50 ± 5% relative humidity
prior to testing. Specimen designation, type of mortar used, and mix proportions are listed in
Table 1.
Brick Pier Specimens. Sixteen specimens constructed with two-core nominal 200 by 100 by
300-mm (8 by 4 by 12-in.) hollow brick were tested. The specimen included 2.1-m by 0.6-m by
200-mm (7-ft by 2-ft by 8-in.) top and bottom spandrels as shown in Fig. 1. A 0.9-m^ (3-ft^) pier
formed the test portion of the specimen. The spandrels were used to apply horizontal shear
forces that were transferred across the pier portion of the specimens. The pier portion was rein-
forced with one Grade 60 No. 5 bar in each of the two outer cells. These two cells were filled with
Portland cement-based grout. No horizontal shear reinforcement was provided in the pier re-
gion. Thickness of mortar joints was maintained at approximately 13 mm (V2 in.) Full mortar
bedding was used. Mortar joints on both faces were cut flush and then tooled. The spandrels
were overreinforced to force damage into the pier portion. Both spandrels were completely
grouted.
Block Pier Specimens. Sixteen specimens constructed with two-core nominal 200 by 200 by
400-mm (8 by 8 by 16-in.) concrete block were tested. The specimen included 2.0-m by 0.6-m by
200-mm (6-ft 8-in. by 2-ft by 8-in.) top and bottom spandrels as shown in Fig. 2. The test
portion consisted of a 0.6-m^ (2-ft 8-in.^) pier. Thickness of the mortar joint in the pier region
was approximately 10 mm (Vs in.). Mortar was applied only to the face shells. Other details
were similar to those given for brick pier specimens.

7' - 0"

4 - #4 Rebars

Hard Rubber Pad (Typ.)

Load Cell

Steel Strut

Hydraulic Ram

Steel Plate

Metric Equivalents :
1 in. - 2S.4 nun
1 tt . 305 mm

FIG. I—Brick pier specimen.

6' - 8-

2' - 0" Z - 8'

n 2' - 0-
2 - #7 Rebars

6' - 8" 2' - 8"

• - # 5 Rebar&
Hard Rubber Pad

Load Cell

Steel Strut

Hydraulic Ram

*=s
Steel Plate
^ ^

u 2' - 0"

<J7
Metric Equivalents :
1 in. - 25.4 mm
1 ft - 305 mm

FIG. 2—Block pier specimen.

Materials
Masonry Units. Average compressive strengths of hollow clay brick based on gross and net
areas were 22.1 and 32.8 MN/m^ (3210 and 4750 psi), respectively. Average compressive
strengths of concrete masonry block based on gross and net areas were 10.1 and 19.3 MN/m^
(1470 and 28(X) psi), respectively. Net areas of brick and block units were determined in accor-
dance with the ASTM Method of Sampling and Testing Concrete Masonry Units (C 140-75).
Mortar. Types M and S mortar mixes were proportioned in accordance with the ASTM Spec-
ifications for Mortar for Unit Masonry (C 270-84). Materials were proportioned as shown in
Table 1. At least six 50 by 50-mm (2 by 2-in.) mortar cubes were prepared for each mortar mix.
Three of these cubes were cured under the same laboratory conditions as the test specimens and
three were cured in a moist room maintained at 23° ± 2°C (73° ± 3°F) and 100% relative
humidity for comparison. Air content of wet mortar at desired consistency was also determined
for each mortar mix. Compressive strengths and air content levels of mortar for brick and block
pier specimens are listed in Tables 2 and 3, respectively.
Grout. Grout used to fill the outer cells of pier portions of the specimens was prepared in
accordance with the ASTM Specifications for Grout for Masonry (C 476-83). The grout mixture
included by volume one part of Type I/II portland cement, one tenth part of hydrated lime,
three parts of sand, and two parts of pea gravel with a maximum size of 10 mm (Vs in.). In
addition, a grout aid admixture, amounting to approximately 1% of cement by weight, was
added. Water-to-cement ratio was approximately 0.80. Average 28-day grout compressive
strengths measured on 75 by 150-mm (3 by 6-in.) cylinders were 23.0 and 19.3 MN/m^ (3330
and 2800 psi), respectively, for brick and block pier specimens.
Masonry Prisms. At least three brick and three block prisms were constructed and tested for

TABLE 2—Material properties for brick pier specimens.

Mortar Propertiles

Compressive Strength,
kN/m^- Masonry
Mortar Cement Compressive
Mortar Cementitious Brand Cured as Moist Air Content, Strength*,
Type Materials Designation Pier Specimens Cured % ItN/m^

M Masonry cement. MCI 20300 20300 7.3 17500

Type M MC2 14200 17400 12.4 15800
MC3 15700 17300 13.5 11900
M Blended portland
cement and PCI 18300 31700 1.0 18000
lime
S Masonry cement, SCI 1S200 14800 11.4 18300
Type S SC2 15500 15600 11.1 18000
SC3 15000 16300 10.9 17200
S Blended portland
cement and PCI 15400 21400 3.5 14900
lime

"Average mortar compressive strength determined from at least three 51 by 51 mm cubes in accordance
with ASTM C 270-84 (UBC Standard No. 24-20).
'Average masonry compressive strength determined from at least three masonry prisms in accordance
with ASTM E 447-84 (UBC Section 2405 C).

TABLE 3—Material properties for block pier specimens.

Mortar Propertiles

Compressive Strength,
kN/m2" Masonry
Mortar Cement Compressive
Mortar Cementitious Brand Cured as Moist Air Content, Strength*,
Type Materials Designation Pier Specimens Cured % kN/m^

M Masonry cement, MCI 15800 19200 13.4 8610

TypeM MC2 17200 28600 14.4 9230
MC3 19600 22800 18.0 8410
M Blended portland
cement and PCI 22300 38700 1.0 11200
lime
S Masonry cement, SCI 17300 15400 11.2 9230
Types SC2 18100 16500 13.3 11000
SC3 19600 23000 13.6 8410
S Blended portland
cement and PCI 17900 22000 4.0 8340
lime

'Average mortar compressive strength determined from at least three 51 by 51 mm cubes in accordance
with ASTM C 270-84 (UBC Standard No. 24-20).
'Average masonry compressive strength determined from at least three masonry prisms in accordance
with ASTM E 447-84 (UBC Section 2405 C).

each type of mortar mix. Construction and testing of prisms conformed to the requirements of
the ASTM Test for Compressive Strength of Masonry Prisms (E 447-84). The height of the brick
prisms was approximately 400 mm (16 in.) obtained by stacking four units. The height of the
block prisms was approximately 600 mm (24 in.) obtained by stacking three units. Test results
are given in Tables 2 and 3.

Test Equipment
Loads were applied diagonally to the top and bottom spandrel beams as illustrated in Fig. 3.
The loads were applied by 445-kN (100-kip) capacity hydraulic rams using special loading shoes
and transferred through four 16-mm-diameter (Vs-in.) high-strength rods. The horizontal com-
ponent of the applied load represented the shear load transferred across the pier section. Two
struts, one on each side of the pier as shown in Figs. 1 and 2, were provided to reduce rotation of
the spandrel beams. Each strut included a 155-kN (35-kip) capacity hydraulic ram and a 222-
kN (50-kip) capacity load cell. Vertical strut force was applied simultaneously on each side of
the pier. The compression induced by each strut was held to one half of the applied shear force
to prevent application of axial load to the pier section. Thus, the pier portion of the specimen
was only subjected to a combination of shear and flexural loads. A photograph of the represen-
tative test setup is shown in Fig. 4.

Test Procedure
After placing the specimen in the loading rig and attaching all instrumentation, the test was
started by applying load through diagonally opposite corners of the top and bottom spandrel
beams. The load was applied in a series of increments alternating from one direction to the

- 4 - S/8' Oia. Loading Rods

Metric Equivalent:
1 in. . 25.4 mm Hydraulic Ram

Load Cell

FIG. 3—Loading arrangement.

FIG. 4—Representative test setup.

opposite direction. When the specimen exhibited large shear deformations at a specific load
level, the specimen was assumed to be approaching its load capacity. The load was then contin-
uously increased to obtain maximum shear stress. Testing was stopped at a stage when the load
dropped under increasing shear deformation.
Specimens were instrumented to measure applied loads and induced deformations. A layout
of instrumentation is given in Fig. 5. The diagonally applied loads were measured by a pair of
445-kN (100-kip) capacity load cells located behind the loading shoes at the corners of the bot-
tom spandrel. Vertical strut forces on each side of the pier were measured by two 222-kN
(50-kip) capacity load cells. Deformations along the diagonals of the pier were measured by two
linear potentiometers.
For the square pier section, the shear strain, 7, was calculated as follows

(di - dr)d
7 =
lab

where
a.b = height and width of the instrumented region (a = fc in this case),
d — original diagonal length of the instrumented region, and
d\ — di — change in diagonal lengths of the instrumented region at a given load stage.

Metric Equivalent:
1 in. = 25.4 mm

Note :
P j = Diagonally Applied Load
1^ = Vertical Strut Force
° 1 '" 2 = Change in diagonal lengths at a given Load Stage
a, b = Height and Width of Instrumented Region
(38 in. for Bricl< and 34 in. for Block Pier Specimens)
h ^ ,h 2 = Transverse Displacement Measurements

FIG. 5—Layout of instrumentation.

Test Results

Shear Strength
A summary of test results including maximum shear stress, corresponding shear strain, and
shear modulus is listed in Tables 4 and 5 for brick and block pier specimens, respectively. Maxi-
mum shear stress was determined from the measured maximum shear force divided by the net
shear area of the pier section. Shear strain was calculated as described in the previous section.
The values reported corresponded to measured maximum shear forces. As judged from the
slopes of shear stress versus shear strain plots, the shear modulus was determined at shear
stresses of approximately 517 kN/m^ (75 psi) and 345 kN/m^ (50 psi) for brick and block pier
specimens, respectively. Values of average maximum shear stress and coefficient of variation

0^
00
r^^ o TT

00
00

S3E z
o<s
t/i o 5

S OO O 00 00 o.
1^ n<Nf^»/)Q0^O
Q0^fn<>l/)lO
o o o o o o
O N O OOf^
^ < ^ ^*^,
oo oo
>0 ( ^
; »-; O; - ^
\ 0 00
"^ ^
' o c> o o c>

5 « S vO <N l/> O^ O lO O"'<J--^00fS OOO

g 2 ' 00 &> >^ 00
X ez 00 ^ ^ 9^ O ^ o o o o o r - t - t ~ OOO

II
a

i!
i-i
00

11
<
giiiiiggsBssBsgg

111
M
g
8 2
s
8
s8 g
•o .a
lit
ill
•a .S c ~
B —
tt -a
S S 1§
I" s £
I II-

^ fN
;#

O o o
o o o

Z
( J I ^ O^ ^O 0> I
J:-o Z lO t ^ fO ^ O • ^ ^ ^O vO O t ^ ON I
f^ ( S lO ^ fsi •
'"OS t^ m <N fO r s

O C T N O ^ T T I / J O •^00
s
6,

1 ir oooooo oo oooooo

w s
u
2 •5
£ j : S8 E •K CO

1
C/5
£^
1/5 J<
ggSili ^o ^ooooo oo^ ^ ts
2 •o
1 JS u
M +5
1
U
hJ B E-3
n c o
<
11
t«
T-" fS r^ TT I/) \o
«E
SI
OiCLi onv5i/i</}bO(/i auOu
QQQQQQ QQ Q Q Q Q Q Q QQ j :
-t-t

B1
•§ 1 c
n. is .2
13
CO •ti T"
i &
8S
•o J
p
B
11 >
>> o
t4-l

11
T3 . a •4-*

eg. e ~ § O; c—
3j
if I" I" 1 *3

•S 3
H^
es — II
S "§
E>
i 11 o

for all specimens are also listed in Tables 4 and 5. Representative plots of shear stress versus
shear strain are shown in Figs. 6 and 7.

Modes of Behavior
All specimens exhibited a shear mode of failure in the pier portion. Photographs of represen-
tative brick and block specimens after completion of tests are provided in Figs. 8 and 9, respec-
tively. Cracking in most cases initiated at one of the pier corners. These were usually small
cracks. In some cases, cracking appeared along the entire length of a diagonal instantaneously.
In these cases, the mode of behavior was brittle and the failure occurred at loads only 10 to 15%
above the cracking loads. Cracking loads ranged from 40 to 90% of the maximum loads. Early
cracking (below 60% of maximum load) usually resulted in subsequent behavior that was more
ductile. Initiation of cracking was also observed in some specimens in the interior portion of the
pier. As the applied load increased, the diagonal tensile stress reached the tensile capacity of
masonry resulting in large diagonal cracks. Cracks generally extended through horizontal and
vertical mortar joints along the pier diagonals prior to reaching maximum load. At failure
loads, however, the cracks extended through some masonry units also.

Discussion of Test Results

Brick Pier Specimens. As listed in Table 4, average maximum shear stress was 972 kN/m^
(141 psi) for the six brick pier specimens constructed with Type M masonry cement-based mor-
tar. The two corresponding specimens constructed with portland cement and lime reached an
average maximum shear stress of 958 kN/m^ (139 psi). Thus, the maximum shear stress values
in these two series were very close. Also, the average cracking loads and the behavior at maxi-
mum load were similar.

200

-0.01

Metric Equivalents:
I in. = 25.4 mm
I psi =6.89 kPa
- 2 0 0 -J-

FIG. 6—Shear stress versus strain for brick pier specimen CP4.

200 T

-0.01 -0.005 0.005 0.01

Shear Strain, in./in.

Metric Equivalents:
I in. = 25.4 mm
lpsi = 6.89l(Pa

-200 -J-

FIG. 7—Shear stress versus strain for block pier specimen DM3.

For the six brick specimens constructed with Type S masonry cement-based mortar, average
maximum shear stress was 841 kN/m^ (122 psi). The two corresponding specimens constructed
with Portland cement and lime reached an average maximum shear stress of 945 kN/m^ (137
psi). The shear stress values obtained in these two series are within 11% of each other, indicat-
ing similar strengths for masonry cement-based and portland cement-based mortars. The over-
all behavior and cracking pattern of specimens were also similar.
As can be seen in Table 4, the shear strains and shear moduli for all specimens fall within
acceptable range of scatter expected in this type of experimental data. Also, there is no appre-
ciable difference between the results of specimens constructed with Type M and Type S mor-
tars. Within the scope of this test program and within the data scatter normally expected, the
type of cement used in the mortar had no influence on the shear strength of brick pier
specimens.
Block Pier Specimens. As listed in Table 5, the six block specimens constructed with Type M
masonry cement-based mortar reached an average maximum shear stress of 1034 kN/m^
(150 psi). This compares very favorably with the two corresponding specimens constructed with
Portland cement-based Type M mortar. These specimens reached an average maximum shear
stress of 1062 kN/m^ (154 psi). Overall behavior and cracking pattern in the two series were
similar.
For the six block pier specimens constructed with Type S masonry cement-based mortars,
average maximum shear stress was 1062 kN/m^ (154 psi). This is greater than the average maxi-
mum shear stress of 993 kN/m^ (144 psi) obtained in the corresponding portland cement-based
specimens.
There is no appreciable difference between the results of specimens constructed with Type M
and Tjrpe S mortars. Thus, within the scope of the test program and the data scatter normally
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
30 MASONRY

(a) Specimen CMS

(b) Specimen CSS

FIG. S—Representative brick pier specimens after test: (a) Specimen CMS: (b) Specimen CSS.

(a) Specimen DM3

(b) Specimen 0S2

FIG. 9—Representative block pier specimens after test: (a) Specimen DM3; (b) Specimen DS2.

expected, the type of cement used in mortar did not affect the shear strength of block pier
specimens.
Comparison with Uniform Building Code Allowable Shear Stresses. Based on the measured
average compressive strength of masonry used in this program, the UBC allowable shear stress
varies from 276 to 379 kN/m^ (40 to 55 psi). This was determined according to UBC Section
2406(c) for a shear wall with in-plane flexural reinforcement present, and "M/Vd" ratio less
than 1. The maximum measured shear stress ranged from approximately two to four times
these values for individual pier specimens. This is quite significant because the pier portion of
each specimen was only minimally reinforced for flexure, was partially grouted, had no shear
reinforcement, and had no axial force transferred through the pier section. All these factors
combined to provide a lower bound value of the measured shear strength.

Conclusioiu and Recommendations

Comparisons of test results obtained from specimens constructed with portland cement-
based and masonry cement-based mortars indicate that the shear strength of masonry piers
constructed with masonry cement-based mortars compared favorably with those constructed
with Portland cement-based mortars. Values of maximum shear stress obtained with portland
cement and masonry cement-based mortars were close and within the range of scatter normally
expected in experimental data. Overall behavior including cracking loads and damage patterns
at maximum loads were similar.
The air contents for the masonry cement-based mortars, shown in Tables 2 and 3, were less
than or equal to the 18% maximum air content specified in ASTM C 270-82. The air contents
for the portland cement-based mortars, also shown in Tables 2 and 3, were significantly less
than the 12% maximum specified in ASTM C 270-82. In view of the similar performance for
specimens made from masonry cement and portland cement-based mortars, the effect of sub-
stantial variations in the air content was apparently insignificant for these laboratory-prepared
mortars. In addition, the shear strength was not influenced significantly by the use of Type M or
Type S mortar. The measured maximum shear stresses ranged from approximately two to four
times those allowed by the Uniform Building Code.
Based on the results of the test program described in this paper, it is recommended that
restrictions on the use of masonry cement in the structural frames of buildings in Seismic Zone
Nos. 2, 3, and 4 be eliminated from the Uniform Building Code.

Acknowledgments
Work described in this paper was sponsored by the Portland Cement Association under the
direction of George B. Barney, Divisional Director, Engineering Services Division. The investi-
gation was carried out at the Construction Technology Laboratories of the Portland Cement
Association under the direction of Henry G. Russell, Director, Structural Development Depart-
ment. The authors acknowledge review of a report on this work by J. Huisman, R. W. Kriner,
C. W. Miller, W. C. Reed, J. A. Rhineberger, andD. L. Rydquist. Suggestions and comments
during this investigation by members of an unofficial ad hoc committee are greatly appreciated.

Investigation of Masonry Bond and

Surface Profile of Brick

REFERENCE: Ribar, J. W. and Dubovoy, V. S., "Investigatioii of Masoniy Bond and Snrface
Profile of Brick," Afajionry.'Materia/^, Design, Construction, and Maintenance, ASTM STP 992,
H. A. Harris, Ed., American Society for Testing and Materials, Philadelphia, 1988, pp. 33-37.

ABSTRACT; The original objective of this research was to investigate the influence of masonry
cements on flexural bond of mortar to clay brick units and on the water penetration of masonry
assemblies fabricated using these materials. The test variables included 20 masonry cements and
11 types of brick, each representing materials produced in different areas of North America.
Results of flexural bond strength tests on masonry assemblies indicated that factors other than
mortar composition affected bond strength. Accordingly, an investigation was made to evaluate
the effect of surface texture of clay brick units on bond strength. A surface profilometer, from
which results could be quantified, was used for this work. Results suggest that surface texture of
the brick is a primary factor in the development of bond strengths.

KEY WORDS; bond, bond wrench, brick, initial rate of absorption, mortar-brick prisms, brick
surface texture, surface profilometer

The masonry research community has, in recent years, expressed considerable interest and
concern regarding bond between mortar and masonry units. It has enthusiastically accepted the
bond wrench testing technique introduced by Hughes and Zsembery [ /] as a mean of measuring
bond strength because of its versatility and ability to produce data with less effort and cost. This
capability has generated a flood of information that is frequently contradictory and more ques-
tions remain unanswered than are resolved.
Industry research efforts on masonry bond have become restricted due to the multitude of
combinations of materials available and the great variety of individual physical characteristics
of these materials. Research is often limited to mortar types and composition and to not more
than a few brick types and their physical properties. The characteristic of the bricks used in
these studies is almost exclusively the initial rate of absorption. Characteristics of the mortar
include composition, air content, water retention, and compressive strength. When combined
with frequently overlooked variables that exist among aggregates, cementitious materials, and
masonry units, these can produce an infinite number of combinations. Therefore, research con-
fined to a narrow spectrum of materials reflects the performance of the tested materials only.
Masonry bond is an intricate process not completely understood. It is obvious that those param-
eters currently being used to predetermine masonry bond are inadequate.
Construction Technology Laboratories (CTL) has been involved in a research program to
evaluate masonry mortars prepared with 20 masonry cements representing a cross-section of
manufacturers in North America. These cements were combined first with one locally produced
solid brick and later with ten additional bricks representative of brick manufactured in the
eastern two thirds of the United States. This investigation concentrated on testing assemblies

'Principal masonry research engineer and research engineer, respectively. Concrete Materials/Technical
Services Department, Construction Technology Laboratories, Skokie, IL 60077.

33
Copyright by ASTM
Copyright' 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
34 MASONRY

for flexural bond strength with tension normal to the bed joint, and for water penetration. The
primary variables in the bond strength study included masonry cement formulation, brick and
mortar properties, and the sources of brick and cement. A total of 454 clay-masonry prisms
were tested for bond strength, yielding 2844 data points. The water penetration study evaluated
20 test assemblies prepared with each of the 20 masonry cement mortars and the one solid brick
unit.
Results of this research disclosed several definite trends. Characteristics appearing to have
favorable influence on bond with all units tested were higher water content of the mortar and
greater fineness of the cement. A poor correlation was found between bond values and water
penetration and leakage, and between mortar compressive strength and bond strength. Also,
bond tests on seven-unit-high specimens made with Type S mortar that were removed from wall
test assemblies had consistently higher bond values than individually constructed prisms. Half
of the Type N mortars also showed this tendency.
A wide range of bond strengths was measured when each of the 20 cements was coupled with
the ten types of clay brick units. None of the traditional parameters mentioned above was useful
in explaining these variations. The least explainable circumstance occurred when two different
bricks with similar physical characteristics yielded widely contrasting bond data when com-
bined with the same mortar. The physical properties of these two bricks, A and H, are given in
Table 1. Even though the two bricks had identical initial rates of absorption. Brick H produced
significantly higher bond strengths with each of the 20 cements. Attempts to use the traditional
parameters, including those given in Table 1, to explain these differences in bond strength
proved to be inadequate.
The bedding surfaces of Bricks A and H revealed a significant difference in surface texture,
as seen in Figs. 1 and 2. An attempt was then made to quantify this difference by examining the
surface texture using a surface profilometer.
Resulting plots of the surface profiles, with pertinent surface texture parameters for these
bricks, are presented in Figs, lb and 2b. Some of the parameters in Figs, lb and 2b that best
describe and quantify the surface texture are defined as follows:
1. Ra is the most widely used parameter of roughness. It is the arithmetic mean of the depar-
ture of the profile from the mean line.
2. Rt is the maximum peak-to-valley height of the profile in the assessment traverse lengths.
3. Rv'is the maximum depth of the profile below the mean line within the assessment traverse
length.
4. Rp is the maximum height of the profile above the mean line within the assessment tra-
verse length.
5. Pz is the average height difference between the five highest peaks and the five lowest val-
leys within the traverse length.
Comparing these parameters, it can be seen that Brick H has a much rougher surface texture
than Brick A, thus providing for greater absolute surface area contacted by mortar. It is the

TABLE I—Data on brick.

Initial Compressive
Brick Rate of Strength,
Designation Absorption, g Absorption, % psi

A 17 4.0 14,160
H 17 2.2 13,990

Fl - tMlvsis
n - Crtph Hodc Traverse Ltmth Reference I Itnort
F3 - IttiP UHFIITER;I 3.88 in STRAlCHTl »t
F4 - Expand REJJBICK A
FS - Exdudt

I$.3S1 l i n - i

^.t:;^W^-«v-*^^

-21.878 i t n -

?tik To UilUs « 48.421 ain

Fl - Aniluiis
Node Traverse Lenith Reference 1 I more
UHF!l,TERf3 3.88 in STRAtCHT ! 8 «
DEI ittKi A

SLOPE « -.12 Je? Ra • 2 774 l i n

Rfl » 4.8i2 l i n
Lo 2.999 in ftsk < -.7
Rp • 18.333 tin Rkg < 8.S
Rv ' 21.2S1 lin UU ' 14.>( I*f
Rt • 48.184 lin 1.114 • 181.383 l i n
3S.S48 l i n
sS I ' is:.673 l i n
Rz « 21.1«< l i n

FIG. 1—(a) Close-upphotographof bedding surface of Brick A (X10):(h) the surface prof He of Brick A.

Fl - fltuhsis
fZ • Graph .1od« rrjv*rs* Length I Mtrtna i Ijnor*
fj - luii» UHFI'.TE^SII sea in ! jTi;oisnr( a >•
fi - Expand TflH ERICK H
F3 - Exciuiie

25.238 fin-<,

., A^iL Kir
LUh1
-2J.67! i i n -
2.S93 in

?iik To >i\\i') - 43,JS^ l i n

Fl - Hfijl'jsis

T^K ESKX H

310PE .21 !lt4 Rj » 7 Si'* ain

?q = 9.322 i i n
U = 2.599 in Ssk ' .1
Rp • 2S.333 iin Sku » 2.«
Rv =» 23.S71 tin I!il<4 'i 23.S4 3»9
Rt s 't3.939 »iB 1414 » 134.478 i i n
S » 33.211 i j n
SI ' 194.331 i i n
Hz = 41.473 mn

FIG. 2—(a) Close-up photograph of bedding surface of Brick H (XIO); (b) the surface profile of Brick H.

judgment of the investigators that this greater surface area is a primary factor in the greater
bond development measured for Brick H.
It is planned to further evaluate the effect of surface texture of the masonry units on bond
strength in future work at CTL.

ConcluBions
Concepts within the masonry community that masonry performance can be predetermined by
the characteristics of individual components is not justified. The authors agree with Yorkdale
[2] that bond between clay masonry units and mortar is not well understood. Additional re-
search is needed to understand the complex mechanism of masonry bond. This research has
identified a useful tool that may assist research community.

References
[ /] Hughes, D. M. and Zsembery, S., "A Method of Determining the Flexural Bond Strength of Brickwork
at Right Angles to the Bed Joint," Second Canadian Masonry Symposium, Ottawa, June 1980.
[2] Yorkdale, A. H., "Initial Rate of Absorption and Mortar Bond," Masonry: Materials. Properties, and
Performance. ASTM STP 778. J. G. Borcheh, Ed., American Society for Testing and Materials, Phila-
delphia, 1982, pp. 91-98.

Preconstruction Brick Veneer Evaluation

Testing

REFERENCE: Raths, D.C., "PKConstractioii Brick Veneer Evaluation Testinc," Masonry: Ma-
teriah, Design, Construction, and Maintenance, ASTM STP 992, H. A. Harris, Ed., American
Society for Testing and Materials, Philadelphia, 1988, pp. 41-56.

ABSTRACT: This paper discusses and presents a preconstruction testing program for a brick
veneer cavity wall. It presents the testing program and the application of ASTM and industry
standards to the evaluation testing of the brick masonry materials. It provides findings on brick
and mortar characteristics, steps taken to improve the performance of the selected brick wall
materials, and engineering procedures for determining material compatibility and performance.

KEY WORDS: bond, bond wrench test, brick masonry, cavity wall, preconstruction, cope, mor-
tar performance, veneer, water permeability

Contract documents typically present the masonry requirements that must be complied with
during the construction of a project. However, they generally do not address preconstruction
testing that should be conducted to ensure proper selection and compatibility of masonry mate-
rials. This paper explores some of the methods and procedures that may be utilized in making
these often very necessary preconstruction evaluations.
Preconstruction testing and evaluation of the masonry materials that comprise the brick ve-
neer of a typical cavity wall can be an important factor in the prevention of construction prob-
lems and/or unsatisfactory performance of exterior masonry walls. A proper program ensures
the compatibility of materials and establishes construction procedures relating to brick prewet-
ting, tooling, and mortar ingredients. Also, a successful preconstruction testing program estab-
lishes a clear understanding of construction requirements, which should lead to a completed
exterior veneer wall with satisfactory service performance.
To illustrate the benefits of a preconstruction testing program, a recent project is presented
as an example. The project can be described as a complex of several low-rise office structures
with a nominal 100-mm (4-in.) brick veneer and 60-mm (23/8-in.) cavity. The exterior brick
veneer was anchored to the backup system with flexible wire anchors. The typical brick unit was
a standard size and shape with a special feature created by a continuous 16-mm (V8-in.)-radius
quarter circle (cope) removed from the top exterior edge of the brick (Figs. 1 and 2).
Preliminary architectural drawings and masonry specifications were reviewed at the start of
the evaluation testing program to ensure a complete understanding of the project requirements
and typical construction details. Suggested changes were offered to the architect in order to
clarify ASTM testing requirements, material selections, and field workmanship practices dur-
ing construction. Also, two brick types (A and B) were evaluated during the evaluation testing
program.
Based on a review of the project requirements and the general nature of the selected brick
properties, a masonry testing and evaluation program was developed that would include testing
of brick, mortar, and fabricated masonry assemblies.

'Principal, Raths, Raths & Johnson, Inc., Willowbrook, IL 60521.

Copyright by ASTM
Copyright' 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
42 MASONRY

FIG. 1—Brick A with cope feature and 25-mm (1-in.) exposed mortar bed joint.

Ptognun Criteria
Brick A had been initially selected by the architect based on its color, texture, and its aes-
thetic appearance on the building in combination with a colored mortar. The selected brick was
a sand-molded clay product manufactured by the wet-mud process. The project specifications
required the brick to conform to ASTM Specification for Facing Brick (Solid Masonry Units
Made from Clay or Shale) (C 216), Grade SW, Type FBS [7], and mortar to be ASTM Specifi-
cation for Unit Masonry (C 270), Type N {8\. The special feature of a continuous cope along the
top exposed edge of the brick results in a nominal 25-mm (1-in.) mortar bed joint on the ex-
posed masonry wall (Figs. 1 and 2). Concerns were raised by the architect regarding the perfor-
mance and durability of the large mortar bed joint at the exposed face of the brick veneer wall
and in part was the impetus for the evaluation testing program.
The author's firm was requested to develop a masonry testing program to determine compati-
bility of materials and to measure conformance to ASTM requirements. This program consisted
of testing the brick for compliance with ASTM C 216 requirements, compressive strength of
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
RATHS ON BRICK VENEER TESTING 43

16mm (5/8') COPE

{3/8')
25mm MORTAR JOINT

LARGE MORTAR JOINT

LACKS TOOLING
COMPACTION

VOIDS IN MORTAR
WATER PATH ?

APPARENT 5mm (3/16")

ZONE OF FISSURES
OR AQUIFER IN BRICK

FIG. 2—Sketch of possible water paths at bed joint.

mortar per ASTM C 270, water permeance rating of fabricated assemblies in accordance with
field-adapted ASTM Test Method for Water Permeance of Masonry (E 514) [1,2] (Fig. 3), flex-
ural bond strength of wall prisms (cut from test panels) by the bond wrench method, ASTM
Method for Measurement of Masonry Flexural Bond Strength (C 1072) [4] (Fig. 4), and freeze/
thaw testing of a masonry specimen (brick and mortar) for a minimum of 150 cycles in accor-
dance with ASTM Method of Sampling and Testing Brick and Structural Tile (C 67) [6\. Also,
due to the high initial rate of absorption nature of this sand-molded brick, brick wetting proce-
dures were to be evaluated to establish requirements and procedures for field construction.

Phase I—Brick A Testing and ETalnation

This phase of testing resulted in the fabrication of four laboratory masonry test panels. Ap-
proximately 600 bricks (Type A) were received from the brick manufacturer for the testing
work. To ensure meaningful testing data, the actual masonry sand and colored portland cement
to be used in the construction were used in the panel construction.
Due to the high initial rate of absorption (IRA) of the brick, approximately 32 g/min/30 in.^
(Table 1), it was decided to fabricate the four masonry test panels with varying wetting proce-
dures. One panel was fabricated with no prewetting of the brick (as delivered brick), two panels
had water sprayed on the brick to represent a typical field wetting application, and a fourth
panel was constructed with the bricks in a saturated condition (bricks kept moist for several
days and covered).

^All water permeance values reported in this paper are based on the area of ASTM E 514 test chamber.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
44 MASONRY

TO WATBH SUPPLY

BAR CLAMPS
BOTH SIDES

TEST SP£C/M£A/
SPONGE RUBBER
BASKET

FIG. 3—Field-adapted ASTM E 514 water permeability test. Figure takenfrom ASTM E 514 [3].

The four masonry test panels, 137 cm (54 in.) wide by 142 cm (56 in.) high were constructed
outdoors by a journeyman mason employed by the masonry subcontractor for this project (Fig.
5). The panels were constructed of one brick wythe (one vertical section of wall one unit in
thickness) [5] in running bond on a pedestal about 60 cm (24 in.) above grade and were com-
pleted in two days. The mortar ingredients were proportioned by volume in accordance with
ASTM C 270, Type N mortar (1:1:5'/2 to 6 proportions) and mixed for 5 min in a paddle-type
mixer. The head and bed joints were completely filled with mortar as called for in the specifica-
tions. The completed panels were representative of a high level of workmanship (Fig. 6). The
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
RATHS ON BRICK VENEER TESTING 45

FIG. 4a—Brick prisms removed from test wall.

FIG. 4b—Bond wrench test on brick prism.

large 25-mm (1-in.) mortar bed joint was tooled using a section of metal pipe with an approxi-
mate diameter of 100 mm (4 in.) in accordance with the specified architectural requirements.
Brick A test results are presented in Table 1. A sieve analysis of the mortar sand was made
and it was determined that the sand conformed to ASTM Specification for Aggregate for Ma-
sonry Mortar (C 144) requirements. After a period of 28 days, field-adapted ASTM E 514 water
permeance [2] and flexural bond strength tests, using the bond wrench method, ASTM C 1072
[4], were executed on the four masonry test panels. Samples were removed from test panels by a
masonry saw for the bond wrench tests (see Fig. 4). The Phase I test results are presented in
Table 2 and Fig. 7.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
46 MASONRY

TABLE l—ASTM C 67 Tests, Brick A.

Compressive strength Averagefivebricks = 36.0 MPa (5224 psi)

Range 31.8 to 41.1 MPa (4615 to 5958 psi)
Initial rate of absorption Upper surface 24.5 g/min/30 in.^
Lower surface 32.0 g/min/30 in.^
24-h cold soak Average = 6.50%
5-hboil Average = 11.30%
Coefficient of saturation Cold soak/5-h boil = 0.58 ratio
Freeze/thaw test 150 cycles completed on small brick prism specimen (brick and
mortar) with negligible weight loss and no visual deterioration
Comments Fine, powdery residue on upper surfaces of brick may affect bond
strengths
Lower surface exhibited a fissured open surface that yields higher
initial rate of absorption

Additional bond wrench tests were conducted on a series of brick prisms (stack bond) to
make a preliminary evaluation of the significance of the fine powdery residue (common to sand-
molded bricks) which adhered to the upper surfaces of the brick. Four brick specimens were
fabricated by a laboratory technician using a hand-mixed Type N mortar (1:1:5V2 propor-
tions) and bricks uniformly prewetted using a hand sprayer. The specimens were fabricated and
cured in the laboratory at about 75°F (24°C) and then tested at 14 days. The test results are
shown in Table 3.
The Phase I masonry testing results indicated the brick met ASTM C 216 requirements [7],
including the freeze/thaw test on a brick and mortar assemblage for 150 cycles. Because Brick
A exhibited a very high initial rate of absorption [6], they were determined to require prewetting
during construction. The ASTM E 514 water permeance tests indicated water leakage rates well
beyond a 1.89 L/h (V2 gal/h)^ criterion established for a typical good wall [1] on Test Panels 1
and 3. The bond strength values were satisfactory for bricks properly prewetted. The Phase I
testing program results lead to the following observations:
1. Field-adapted ASTM E 514 water leakage rates varied widely according to the degree of
prewetting of the brick.
2. Test Panel No. 3 construction was most representative of field workmanship and its water
leakage rate of 8.50 L/h (2V4 gal/h) was deemed unacceptable.
3. Test Panel No. 4 exhibited the lowest water leakage rate, but was deemed to be impracti-
cal to construct due to the "floating" of the brick and "bleeding" of mortar during fabrication.
This was caused by the saturated condition of the brick at the time of fabrication.
4. Bond strength test values provided an adequate factor of safety compared to recom-
mended practice for brick masonry which allows flexural tension of 193 kPa (28 psi) for Type N
mortar [5].
5. The large volume of mortar at the 25-cm (1-in.) exposed bed joint created by the cope
along the top edge of the brick was difficult to properly compact. Small horizontal voids were
visually detected along the brick surface at the cope area when specimens were broken apart for
examination (Fig. 2).
6. The fine powdery residue on the top surfaces of the brick, resulting from the brick man-
ufacturer's production process, reduced the bond strengths approximately 12 to 15% (see
Table 3).
7. The texture and density of the bricks' bottom surface were suspected to be partly a cause
of high water leakage rates (see Fig. 2).

137cm (54")

CONT. 2" X 4' HORIZ. BRACE

MORTAR CAP
^
# =
^

LEAVE OPENING THIS SIDE i

FOR P P E CLAMP ^

CO
lO

IPARQE THIS END

E
o
CM

LEAVE OPENINQ THIS SOE 4

FOR PIPE CLAMP

FRONT ELEVATION
FIG. 5—Typical brick masonry test panel.

Due to the high water leakage rates, the use of Brick A was not recommended. It was recom-
mended to evaluate an alternate Brick B as well as regular stretcher bricks (solid—no cores)
without the cope feature in the next testing phase. Also, the next test phase needed to more fully
address brick prewetting requirements considering the extreme sensitivity of water leakage to
prewetting moisture contents.

Phase lA—Additional Testing/Evaluation (Bricli A)

Based upon concerns raised regarding the fissured texture and zone of lesser density near the
bottom surface of Brick A, it was decided to examine the surfaces of the brick more closely and

FIG. 6a—Tr^i panel myrkmanihip at head and bed joints, interior surface.

|'1»fi#f|r|4;

;r; - ";-t#'fe' ., "

t?:fe!8'-

FIG. 6b—Close up of head and bed joints, interior surface.

to perform a special water penetration test using a dye to allow identification of actual water
leakage paths through the brick masonry assembly.
A typical brick was sectioned vertically along its long axis to allow a close examination of the
cross section. This examination revealed the presence of a less dense zone of material, about 5
mm (3/i6 in.) thick, at the entire bottom surface of the brick. This led to removal of a large
masonry prism section, approximately 46 cm (18 in.) by 76 cm (30 in.) in size from one of the
completed test walls. It was placed in a horizontal position with the exposed face upward, and
ponded water with a green dye was allowed to penetrate the wall for about 12 h. Upon selective

TABLE 2—Phase I test results of masonry wall panels, Brick A.

Field Adapted Average

E 514 Rate, Flexural
Test Brick L/h Tension,
Panel Prewetttag (gal/h) kPa (psi) Comments
Panel 1 Dry (as delivered) 22.9 (6.0) 365 (S3) Leakage very high, poor bond
Panel 2 Sprayed Tests not run due to cracked bed joint
in test panel
Panel 3 Sprayed 8.5 (2.25) 1531 (222) Typical of field production, good
bond, bricks brushed to remove
powder residue
Panel 4 Saturated
(surface dry) 3.2 (0.85) 1420 (206) Bricks too wet, wall difficult to lay
up—floating

NOTES:
1. Type N Mortar (1:1:6 proportions Panel 1, 1:1:5i/2 proportions Panels 2 through 4).
2. Compressive strength of mortar cubes, ASTM C 270:6123 kPa (888 psi), 6206 kPa (900 psi), and 6295
kPa (913 psi).
3. Flexural tension determined by bond wrench tests, ASTM C1072, on brick prism specimens (removed
from test panel) [4]. Tooled face in tension.

10- • -

9- 34
-- 32
PANEL 1
"""^"^ 30
e - • 28
o
X -• 26
DC
Ul - • 24
(0 -- 22
z
-- 20
s 18
-I
<
o 16
c ftNEL 3
Ul
I- ^-"*-^ 14
< - •

i 12
o
- • 10
•
8
• - 6
^ h ^
_^Pfi NEL 4 4
2
RATHS TY 1CAL WALL
1 1 1 1 1 1 1 1 1 1 L
4 5
TIME - HOURS

FIG. 7—Phase I test results, modified ASTM E 514 water permeance test.

TABLE 3—Bond wrench stack bond tests, brushed versus normal

surfaces. Brick A.

Average
Brick Flexural Tension,
Specimen Surface kPa (psi)

1 Normal 869 (126)

2 Normal 979 (142)
3 Brushed 1083 (157)
4 Brushed 1007 (146)

NOTES:
1. Type N Mortar (1:1:5V2 proportions).
2. Compressive strength of mortar cubes, ASTM C 270: 5116 kPa'(742
psi) and 5461 kPa (792 psi).
3. Flexural tension determined by bond wrench tests, ASTM C 1072, on
brick prism specimens [4], Tooled face in tension.
4. All bricks uniformly wetted by a hand sprayer prior to fabrication of
samples.

demolition of the masonry prism, a detailed examination was made of the mortar/brick inter-
face at the head joints, the bed joints, and in the cope area (top front portion of the brick). This
examination resulted in the conclusion that the water paths were occurring in the bed joint
behind the mortar bond line through the less dense zone at the brick bottom surface, in addition
to the expected water paths at the head joints. Also, it was noted that water appeared to migrate
laterally at the cope area through the veins (voids) in the uncompacted mortar (see Fig. 2).
Additional testing was deemed necessary to further study the effects of prewetting, surface
texture, and tooling of the large exposed bed joint to determine their effects on water permeance
of the brick veneer wall.

Phase n—Brick A and B Testing and Evaluation

The second phase of testing consisted of comparative evaluations of Bricks A and B. Bricks of
both types with the special 16-mm (Vs-in.) cope feature along the top exterior edge. Fig. 1, and
a regular stretcher brick (solid—no cores) were included. This allowed study of the performance
of identical brick materials with and without the cope feature, and comparison of test results
with two different brick bottom surface textures, fissured and smooth.
As a prelude to fabrication of the test panels, a series of brick saturation tests were conducted
to establish the wetting and drying characteristics of Bricks A and B. The results are presented
in Tables 4 and 5.
An additional supply of bricks was received from each Brick A and B manufacturer for con-
structing additional masonry test panels. The same masonry sand obtained from the local job
site source and the actual colored portland cement were used for the mortar mix as in the Phase
I tests.
A series of five masonry test panels, similar in size to the Phase I test panels, were constructed
by the same journeyman mason. The masonry panels were constructed of a single brick wythe in
running bond and were completed in about three days. The mortar ingredients were propor-
tioned in accordance with ASTM C 270, Type N mortar (1:1:5 proportions) with slightly less
sand than Phase I and mixed for 5 min in a paddle-type mixer. All head and bed joints were
filled with mortar as specified. The fifth test panel was constructed utilizing a Type N mortar
with higher lime content than normally allowed (1: 1V2:6 proportions) at the request of the
contractor. The completed test panels were of a high level of workmanship (see Fig. 6).

TABLE 4—Oven dry to saturation data.

Brick A Brick B

Moisture Conditions Weight, g Moisture, % Weight, g Moisture, %

Oven dry weight 1785.0 0 2118.8 0

10-min soak weight 1881.7 83 2220.1 84
4-h soak weight 1890.3 91 2229.0 92
Saturated weight 1901.3 100 2238.8 100

NOTE—Weights based on the average of three bricks submerged in water tank.

TABLE 5—Air drying from saturation data.

Brick A Brick B

Moisture Conditions Weight, g Moisture, % Weight, g Moisture, %

Saturated weight 1917.1 100 2232.7 100

1-h dry weight 1908.7 93 2225.9 94
6-h dry weight 1876.8 67 2195.8 70
24-h dry weight 1851.8 46 2169.8 48
Oven dry weight 1796.5 0 2111.6 0

NOTE: Weights based on the average of four bricks at 58°F (14.44°C) mean temperature.

The tooling of the large mortar bed joint was accomplished in this phase using a metal cylin-
der with an approximate diameter of 60 mm (2'/2 in.), slightly smaller than that used in Phase
I. This smaller diameter tool shape appeared to allow better compaction of the large mortar
volume in the cope area of the brick specimen. A regular mason's jointer was used for tooling
the 10-mm (^/s-in.) joints of the panels fabricated with regular stretcher bricks.
Based on the saturation and drying curves developed for Brick A and B, the moisture levels of
the bricks were monitored by weighing to allow measurement of specific levels of moisture at the
time the masonry test panels were constructed (Tables 5 and 6). Consideration for moisture
levels of brick was based upon water leakage values of Phase I tests, the journeyman mason's
"feel" as to proper timing for laying up and tooling of the mortar joints.
ASTM C 67 tests for Brick B are presented in Table 7. After a period of 28 days, ASTM E 514
water permeance and flexural bond strength tests, ASTM C 1072, were conducted on the ma-
sonry test panels. These results are given by Table 6 and Fig. 8.
Review of the ASTM E 514 water leakage and bond strength test data indicated the Brick A
and B special cope feature did not have any significant influence when compared to stretcher
brick performance. Therefore, it was decided that the coped feature of the brick was accept-
able. The water leakage rates were satisfactory for Brick B as they were within the program
criteria of 1.89 L/h ('/2 gal/h) for a typical good wall [/]. Brick A was very close to the program
criteria with a leakage rate of 2.28 L/h (0.60 gal/h).

Phase m—Brick A Testing and Evaloation, Modified Surface

Based upon our previous testing results and observations, the manufacturer of Brick A resub-
mitted a brick with a modified bottom surface exhibiting a closed, finished surface for addi-
tional testing. The brick production process was altered to allow an enhanced strike-off mecha-

TABLE 6—Phase II test results of masonry wall panels. Bricks A and B.

Modified Average
Brick E 514 Rate, Flexural
Test Brick Moisture L/h Tension,
Panels Prewetting Content, % (gal/h) kPa (psi) Comments

Panel 1 Saturated and 60 2.28 (0.60) 234 (34) Powdery residue

(Brick A— dried both surfaces,
Regular) poor bond
Panel 2 Saturated and 58 2.28 (0.60) 772 (112)
(Brick A— dried
Cope)
Panel 3 Saturated and 85 1.90(0.50) 924(134) Wall slightly
(Brick B— damp wet at lay up
Regular)
Panel 4 Saturated and 72 1.51 (0.40) 876 (127) Low leakage
(Brick B - dried rate
Cope)
Panel 5 Sprayed by 67 5.24(1.40) 469 (68) High leakage
(Brick A— mason rate, poor
Cope) bond

NOTES:
1. Type N mortar: Panels 1 through 4 (1:1:5 proportions). Panel 5 (1: l'/2:6 proportions).
2. Compressive strength of mortar cubes, ASTM C 270: Panels 1 through 4: 6812 kPa (988 psi), 9860
kPa (1430 psi), 10 115 kPa (1467 psi), and 10 329 kPa (1498 psi); Panel 5: 4799 kPa (696 psi).
3. Flexural tension determined by bond wrench tests, ASTM C 1072, on brick prisms (removed from test
panel) \4\. Tooled bed joint in tension.
4. Moisture content based on 24-h cold soak capacity.

TABLE 1—ASTM C67 Tests. Brick B.

Compressive strength Average five bricks = 32.2 MPa (4678 psi)

Range 24.8 to 38.7 MPa (3602 to 5617 psi)
Initial rate of absorption Lower surface 33.3 g/min/30 in.^
24-h cold soak Average = 5.68%
5-h boil Average = 10.37%
Coefficient of saturation Cold soak/5-h boil = 0.55 ratio
Freeze/thaw test To be executed later if brick is selected
Comments Fine, powdery residue on upper surfaces of brick may affect bond
strengths.
Lower surface exhibited a closed, smooth surface.

nism after the clay was compacted in the molds. This represented an improvement in the
finished surface texture and was expected to significantly improve both water leakage rate and
bond strength properties. A further series of small masonry test panels were constructed with
Bricks A and B, including the special cope feature, to allow a further understanding of the
prewetting sensitivity characteristics of the bricks and to evaluate the modified surface of Brick
A. The masonry test panels were constructed 81 cm (32 in.) wide by 46 cm (18 in.) high, one
brick wythe in running bond and Type N mortar (1:1:5 proportions) ingredients identical to
previous tests. The moisture levels of all bricks for the specimens were closely controlled during
fabrication (weighing of oven dry bricks and adding measured amounts of water) to establish
known moisture contents. The ASTM E 514 water permeance values were determined by utiliz-
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
PATHS ON BRICK VENEER TESTING 53

anoH d3d s a a m • aiva M O I J

(o 10 « n CM

anoH U3d SNonivo - aivu Mond

ing a special small scale test chamber 30.0 cm (12 in.) high by 49.5 cm (19V2 in.) wide. The flow
rate and pressure differential were maintained to be equivalent to the standard size ASTM E
514 chamber. The water leakage test rates were then multiplied by the ratio of the standard size
test chamber divided by the small-scale test chamber to allow direct comparison of all ASTM E
514 test values. The results of these tests are shown in Table 8.
Based on the Phase II and III test results for ASTM E 514 water permeance values, accept-
able flexural bond strengths, freeze/thaw durability results, and the designed appearance of the
bricks in the finished wall, the coped Brick A with the modified bottom surface was selected for
construction.

ASTM E 514 Field Testing

The completed in-place brick masonry wall was tested during the construction period to mon-
itor the workmanship of the completed wall and conformance to the established project criteria
of 1.89 L/h (1/2 gal/h). The results from three field adapted ASTM E 514 tests established the
following values:

1. Test Location No. 1—1.89 L/h (1/2 gal/h).

2. Test Location No. 2—1.18 L/h (Vis gal/h).
3. Test Location No. 3—2.36 L/h (Vs gal/h).

Testing Conclusions
The preconstruction masonry testing program established several important factors: compat-
ibility of materials, sensitivity of brick wetting procedures to service performance of the com-
pleted wall, effects of 16-mm (Vs-in.) radius cope along the top exposed face of the brick, and
the influence of surface properties (density and texture of the bottom brick surface) for sand-
molded brick manufactured by the wet mud process.
The program established that the most important parameter, the degree of brick wetting at
the time of construction, related to both water permeability performance and bond strength.
The high initial rate of absorption required controlled wetting procedures in order to ensure
watertight joints between the mortar and the brick units. Technical literature and industry

TABLE 8—Test results field adapted E 514 water permeance.

Equivalent Values,
Moisture Content, % E 514 Leakage Rate, L/h (gal/h)

Panel No. Brick A Brick B Brick A Brick B

1 0 10 6.80(1.80) 5.20(1.43)
2 21 25 4.50(1.12) 2.05 (0.54)
3 44 44 2.30 (0.61) 1.96 (0.52)
4 60 56 2.10 (0.55) N/A
5 70 82 1.63 (0.43) 1.21 (0.32)
NOTES:
1. E 514 water permeance tests run at 14 days with reduced size test chamber,
30.5 cm (12 in.) by 49.5 cm (19'/2 in.). One test executed on each test panel.
2. Tooling of bed joints with 6.35 cm (2i/2 in.) diameter cylinder.
3. Moisture contents based on 24-h cold soak capacity and based on average
weight of five bricks.
4. Reduced size E 514 test chamber 30.0 cm (12 in.) high by 49.5 cm (19'/2 in.)
wide.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
PATHS ON BRICK VENEER TESTING 55

practices may not provide adequate guidance for the designer to specify in detail the brick pre-
wetting requirements for bricks with high initial rates of absorption. Therefore, it becomes criti-
cal to conduct preconstruction testing with ASTM E 514 as well as flexural bond strength tests
to establish the criteria for the specific brick and mortar selected. The preconstruction testing
program described above resulted in the selection of a brick and mortar proportions yielding a
water permeability rating of approximately 1.89 L/h ('/2 gal/h) for brick test panels, thus meet-
ing the program criteria. This was accomplished after refinement of brick wetting procedures
and modification of the texture of the lower surface of Brick A.
The influence of the special cope feature on the exposed top edge of the brick created a large
mortar volume at the exterior bed joint. The results of preconstruction testing for water permea-
bility and bond strength indicated this feature was neither significant nor detrimental to perfor-
mance. However, examination of demolished masonry test panels revealed that proper tooling
was important to obtain full compaction of mortar in the area of the cope (see Fig. 2). The
radius and width of the jointer were varied in order to establish better mortar compaction tech-
niques for the large exposed joint. The tooling technique was unique for this bed joint shape and
resulted in establishment of tooling methods with a specially designed jointer.
The fine powdery residue that adheres to the upper surfaces (constructed position) of the
sand-molded brick was related to the molding process in manufacturing. Testing indicated its
presence reduced bond strengths by up to 15% but did not affect permeance performance.
The preconstruction testing program yielded positive results by quantifying the critical items
of performance and identifying required construction methods and procedures. Also, the de-
signer, contractor, mason subcontractor, and owner jointly participated directly in the brick
masonry test program, which resulted in understanding of the project requirements among all
the parties. Procedures developed during the masonry testing program led to the development
of additional field quality control measures regarding brick prewetting and mortar tooling.
These items critical to the performance of the completed masonry wall were identified well in
advance of construction to allow the contractor and masonry subcontractor to plan and develop
wetting procedures for construction.
The preconstruction and construction testing program identified some important factors
which may apply to other masonry construction:
1. Initiate preconstruction testing program sufficiently in advance of construction to allow
evaluation of brick and mortar and time to effect any material changes without delaying the
project.
2. Importance of a closed finish surface on sand-molded brick as it relates to water per-
meance.
3. Determination of compatible mortar proportions.
4. Degree of brick prewetting and identification of initial rate of absorption level where this
is no longer required.
5. Development of field procedures to achieve the required prewetting.
6. The testing program results can be translated into actual construction methods which
produce in-place walls having the same performance as tested walls.
The benefits of a proper preconstruction testing program should promote understanding of
project requirements, decrease construction problems, increase the owner's assurance of ob-
taining quality constructed masonry, and improve the service performance of the completed
brick masonry wall.

References
[/] Raths, C. H., "Brick Masonry Wall Nonperformance Causes," Masonry: Research, Application, and
Problems. ASTM STP 871. J. C. Grogan and J. T. Conway, Eds., American Society for Testing and
Materials, Philadelphia, 1985, pp. 182-201.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
56 MASONRY

[2] Monk, C. B., Jr., "Adaptations and Additions to ASTM Test Method E 514 (Water Permeance of
Masonry) for Field Conditions," Masonry: Materials, Properties, and Performance. ASTM STP 778, } .
G. Borchelt, Ed., American Society for Testing and Materials, Philadelphia, 1982, pp. 237-244.
[3] ASTM Test Method for Water Permeance of Masonry (E 514-74), American Society for Testing and
Materials, Philadelphia, 1974.
[4] ASTM Method for Measurement of Masonry Flexural Bond Strength (C 1072-86), American Society
for Testing and Materials, Philadelphia, 1986.
[5] "Recommended Practice for Engineered Brick Masonry," Brick Institute of America, McLean VA,
1978.
[6] ASTM Method of Sampling and Testing Brick and Structural Clay Tile (ASTM C 67-83), American
Society for Testing and Materials, Philadelphia, 1983.
[7] ASTM Specification for Facing Brick (Solid Masonry Units Made from Clay or Shale) (ASTM C 216-
84), American Society for Testing and Materials, Philadelphia, 1984.
[8] ASTM Specification for Mortar for Unit Masonry (ASTM C 270-84), American Society for Testing and
Materials, Philadelphia, 1984.

Problems and Solutions to Masonry

Buildings

REFERENCE: Gensert, R. M. and Bretnall, W. C, "Problems and Solutions to Masomy Bnlld-

ings," Masonry: Materials, Design, Construction, and Maintenance, ASTM STP 992, R. A.
Harris, Ed., American Society for Testing and Materials, Philadelphia, 1988, pp. 57-95.

ABSTRACT: The authors have been actively investigating masonry buildings for over 25 years
and have documented problems associated with differential movement within masonry systems
and between systems and their supporting structures. While they have found many buildings and
components which perform well, it has been their experience that masonry buildings with prob-
lems have often been brought to the attention of the public without full and in-depth understand-
ing of the source of the trouble.
One cannot analyze a building's problems by a quick visual inspection and a first reading of the
drawings and specifications. It is necessary to document all signs of distress as an overlay upon the
structural frame. Materials must be sampled and tested for quality and compatibility. Walls must
be probed with fiber optic instruments to determine the quality of assemblages and their attach-
ments to the structure. Architectural details, such as drainage, flashing, movement joints, bond-
ing, etc., are important to the well-being of masonry walls and must be examined. Structural
support of masonry and its interaction with the supporting structural component and with the
structural frame as a whole is also critical in terms of differential movements. The authors have
documented this experience by means of photography computer graphics instrumentation, and
this paper will share with you the knowledge gained by this work effort.

KEY WORDS: masonry, prefabricated panels, rigid connections, deflection, differential move-
ment, interaction, movement joints

Before we look at specific problems, we must first understand some basic principles of mate-
rials and the action of structural frames.
Masonry curtain walls, whether laid in place or prefabricated, must be free to move indepen-
dently of their supporting structures. If a masonry wall is restrained by the supporting frame
from expanding or contracting, excessive stresses will be introduced into the wall. Just as impor-
tantly, a structural frame which is more limber than the stiff masonry will induce excessive
stresses into the masonry by means of deflection. The different properties of masonry and struc-
tural frames can be itemized to draw some comparisons. Table 1 compares the weaker proper-
ties of masonry to the comparable properties of structural steel and reinforced concrete.
When a structural frame deflects laterally from wind loads or seismic loads, the deflection is
often called a drift and the drift comparison between various structural frames is called a drift
ratio. Thus, a braced frame in steel or a shear panel system in concrete or masonry would be
very stiff and would have a drift ratio of 1. By comparison, a steel frame with shear connectors
and no moment-resisting joints would be very flexible with a drift ratio of 10. Table 2 gives
reasonable values for the drift of different structural systems.
The shortening of the height of a structural frame may also induce restraints and resulting
stresses in masonry walls. This shortening of the frame may be due to axial loads in the columns

'Principals, Gensert Bretnall Associates, Cleveland, OH 44114.

57
Copyright by ASTM
Copyright' 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
58 MASONRY

Sr^u:

^ut
qAmi^^
'I tl
8 if
o
-' X
X 1/5

1^ j<
<N '

<s o

I
t;
o

\0 1-H CQ
.J
<

II I I
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GENSERT AND BRETNALL ON MASONRY BUILDINGS 59

TABLE 2—Comparison of typical lateral deflections

of structural frames.

Structural Frame Drift Ratio

Braced frame steel 1
Rigid frame steel 3
Semirigid frame steel 6
Shear connections only steel 10
Rigid frame concrete 2
Shear panel metal 2
Shear panel concrete and masonry 1

and is instantaneous for elastic behavior in steel or concrete or may be time related for concrete
columns, walls, or posttensioned slabs as a result of material creep. For steel, the ratio may be
1, but for concrete it may be 3. Volume changes in masonry can work against deflections of the
structure or individual components, inducing cumulative stresses which exceed the strength of
the masonry.
Figure 1-a illustrates the deflection of an unrestrained steel frame under lateral load without
regard to the exterior curtain wall system. Figure 1-b illustrates the same frame under the same
lateral load but this time with the exterior curtain wall added. The frame deflects without re-
straint by the curtain wall because the movement joints within the curtain wall are effective. The
frame deflects the same amount as it did without regard to the curtain wall. Although J-1 is less
than J-2, as shown in Fig. 1-c, the reduction and extension of joints between panels is within the
workable tolerances of the joint materials. The deflection of the frame applies no stress to the
curtain wall.
Figure 2-a illustrates the same frame, without regard to the exterior curtain wall, deflecting
the same amount under the same lateral loads. The frame is illustrated again in Fig. 2-b but this
time with a curtain wall which has inoperable, insufficient, or even no movement joints (Fig.
2-c). The curtain wall now becomes a significant factor in resisting lateral loads, and the frame
deflects less than it would without the curtain wall. If this wall is masonry, an examination of
Table 1 will clearly show that it does not take much relative force to induce high stresses into the
masonry.
Before we look at a classic case history of this type of problem, we should be aware of the
tensile failure mode of masonry when it is under a compressive stress. Figure 3 shows a brick
pier under compression to the point of failure with vertical cracks that result from lateral ten-
sion stresses resulting from Poisson's ratio. The brick pier splits.
What we learn from this basic background is that if short masonry walls or hollow piers are
supported by a structural frame that deflects, and if horizontal movement joints do not exist,
the masonry is subject to compression stresses and splits vertically at the point of highest stress.
Points of highest stress are typically at corners and returns, as this is where the masonry is
stiffest.

Differential Movement Between Masonry and Stmctural Frames

Case History I
Several case histories illustrate what can happen when no allowance is made for these differ-
ential movements. Figure 4 shows a 23-story steel moment resistant frame building. The exte-
rior curtain wall is partly comprised of piers 1.22 m (4 ft.) deep by 0.61 m (2 ft.) wide, built with
102 mm (4 in.) of brick without any backup. The brick is supported at every other floor, and a 6-
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
60 MASONRY

- -—' V -
\
V-
\
\ \
\
- r 4-
—t
—-

DEFLECTED STEEL FRAME STEEL FRAME. WALL PANELS

WITH EXPANSION JOINTS
a

EXPANSION JOIHT

PANEL INTERACTION

FIG. I—Action of curtain wall with expansion joints.

mm (V4-in.)-thick movement joint was provided under the supporting angle at every fourth
floor. Within eight years of completion, cracks were observed near the comers of these piers
(Figs. 5,6).
It is not hard to understand why these cracks occurred. Movement joints were provided only
at every other supporting shelf angle. Where movement joints were not provided, the normal
and expected volume increase in the brick, due to moisture absorption and thermal expansion,
was restrained by the supporting shelf angle and by the shelf angle above. Where the movement
joint was provided, these volume changes significantly reduced the thickness of the joint; the
thickness had already been reduced during construction by deflection of the shelf angle support

DEFLECTED STEEL FRAME STEEL FRAME. WALL PANELS

WITH NO EXPANSION JOINTS
a

HO EXPANalON

PANEL INTERACTION

FIG. 2—Action of curtain wall without expansion joints.

system. The masonry was restrained and compressed by the jaws of a giant vise, the jaws of the
vise being each succeeding set of shelf angle assemblies. Building drift due to wind caused addi-
tional compressive stresses in the masonry, sufficient to cause the masonry to crack at the stiff-
est place: the comers of the piers (Figs. 7,8).
Figure 9 shows lack of vertical expansion. Figure 10 shows compression stress contours as a
result of a finite-element analysis. Interestingly, ten years after the building was built all of the
cracked masonry was replaced, but the system of movement joints in the reconstruction fol-
lowed the original design. Today, seven years after this masonry was replaced, it is cracking in
the same location and in the same general pattern as the original masonry (Figs. 11-14). If the
problem had been understood and a new movement joint system properly designed, there would
be no cracks in the masonry today.

FIG. 3—Tensile splitting due to compressive load.

FIG. 4—Case History I: Overall view of building.

Case History II
Prefabricated masonry panels can have just as much trouble with overdeflecting frames as
laid-in-place masonry. As another example, Fig. 15 shows a portion of a four-story unbraced
steel frame building with precast concrete floors. For reasons that baffle the authors, the build-
ing has lateral tie beams only at the roof; a futile attempt was made to provide a moment con-
nection between the precast concrete slabs and the steel columns.
Figure 16 shows a cross section through the structure. This 15.24-m (50-ft)-high structure
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GENSERT AND BRETNALL ON MASONRY BUILDINGS 63

FIG. S—Case History I: Compression cracks (original masonry).

FIG. 6—Case History I: Compression cracks (original masonry).

FIG. 7—Case History I: Compression cracks (original masonry).

FIG. 8—Case History I: Spoiling of brick (original masonry).

FIG. 9—Case History I: No vertical expansion.

FIG. 10—Case History I: Compression stress contours.

FIG. 11—Case History I: Compression cracks (new masonry).

FIG. 12—Case History I: Compression cracks (new masonry).

would theoretically deflect 300 mm (11.5 in.), a A/H ratio of 0.019, under a code specified wind
load (Fig. 17). A properly designed steel frame for this building would deflect only 38 mm (1.5
in.), or a A/H ratio of 0.003 (Figs. 18,19).
Any four-story building with this amount of drift would be uninhabitable, yet this particular
building has been in use for over ten years. What keeps the drift within tolerable limits? The
prefabricated 122-mm (4-in.)-thick masonry panel curtain wall, being much stiffer than the
building frame, resists all of the wind loads (Fig. 20). However, they are not designed to do so
and have become distressed as a result. The panels are cracked at the comers of the window and
louver openings, exactly in the pattern one would expect from racking panels. Window frames
were also trying to resist these racking forces, and shear cracks occurred in the panels where
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GENSERT AND BRETNALL ON MASONRY BUILDINGS 67

FIG. 13—Case History I: Compression cracks (new masonry).

FIG. 14—Case History I: Compression cracks (new masonry).

window frame connections to the masonry occurred. All in all, the building was on the verge of
collapse. Figures 21-23 show the distressed condition.
Figures 24 and 25 indicate the, shear displacement between adjacent masonry panels as re-
flected by shear displacements in panel-to-panel caulking.
One other aspect of this building, unrelated to differential movement, needs to be mentioned.
Figure 26 details the belt course of masonry. Original design drawings called for a thicker brick
at the belt course so that the back of the panel would be flush. The panel as-built uses a brick
the same size as other bricks in the panel, effectively destroying the structural integrity of the
panel for wind loads and allowing a great potential for water infiltration.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
68 MASONRY

FIG. 15—(A) Case History II: Overall view of building; (B) Case History II: Typical wall panel.

Case History HI
At the other extreme, consider a concrete building which experiences no significant drift from
lateral loads and which is clad with prefabricated masonry panels with all connections rigidly
welded to the supporting structure. This building, Fig. 27, consists of one-way posttensioned
lightweight concrete slabs supported on lightweight concrete bearing walls. The bricks in the
prefabricated panels were not compatible with the mortar, resulting in very low bond strength,
and rather than change either the brick or the mortar, an attempt was made to provide the
necessary strength by threading reinforcing rods through the brick cores and then attaching
these bars at three points to a structural steel T, the assumption being that the brick and the
steel T would act as a composite section (Fig. 28). However, the principles of composite action
were poorly understood, and horizontal cracks soon appeared in the panels (Fig. 29).

Case History IV
In yet another building with no significant wind drift, the prefabricated brick curtain wall
panels failed partly because prefabrication methods led to weak mortar with poor tensile bond.
Not only was the bond of poor quality, but the process of grouting the joints and cores of the
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GENSERT AND BRETNALL ON MASONRY BUILDINGS 69

BSSt..
jak'innt j . - i . ^ - ; ^ .a,(r;t,V.;;> fTJiirt'-aiffia^

B =f
^Ul,"" ^^SSeehoH T^

*IMJ!UtJ
I'"-

W 10
COUMfN

,/-ro LM. ooHcncn

nr= =
^^assf^
m&MN

«B.na4
i^ =3

JjL^^'l ZI^" ittJM-l

FIG. 16—Case History II: Typical wind frame as designed and built.

brick required the panels to be fabricated face down. As a result, the exterior mortar joints were
not tooled and are not water resistant.
As with the previous example, all connections of these panels to the structural steel frame of
this building were welded and are rigid, and shims between adjacent panels were not removed.
And also as with the previous example, the tension field induced in the panels as they cooled
with respect to the supporting structure caused cracks at the reinforcing bars, where the cross
section of the masonry was reduced. Due to the weakness of the mortar and the relatively close
spacing of the reinforcing bars, the cracks also travelled from bar to bar, causing the panels to
split at midthickness. Figures 34 through 43 show the distressed condition of these panels.
More interestingly, vertical cracks soon developed at the location of the outer reinforcing
rods. During the first winter after construction, the panels cooled while the supporting structure
remained at a relatively uniform elevated temperature. This created a tension field between the
rigid connections, and the panels cracked at the reinforcing rods where the cross section was
reduced (Fig. 30). Several years later, these same panels, which are supported by the postten-
sioned slab, cracked at the center reinforcing rod. This crack, which telegraphed through from
the back of the panel, resulted from high compression forces, induced both by creep shortening
of the posttensioned lightweight concrete slab and thermal and moisture expansion of the brick.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
70 MASONRY

*•*»

., 1.

FIG. 17—Case History II: Deflection of typical wind frame.

These forces, applied to the panel eccentrically through its rigid connectors, caused the panel to
fold like an accordion (Fig. 31). The progression of these cracks—horizontal and vertical—was
documented in a series of three detailed surveys made over a five-year period.
One must always ask the question: Can these stresses be induced into a panel before it is
destroyed at the connections? After all, the stresses must be higher at the connections where the
transfer of force occurs between panel and structural frame within a small area. A computer
analysis was made using the finite-element metjiod, and although the stresses were higher at the
connectors, they were confined to a relatively local area where the panel was reinforced. Figure
32 indicates the tension force contours by analysis and Fig. 33 shows the reinforcing that was
present.

Case History V
Tall parapet panels sometimes require extra supports in the form of steel bracing between the
parapet panel and the roof deck. One building in particular (Fig. 44) had parapets with diago-
nal bracing and short parapets without bracing (Figs. 45, 46). Bracing for the tall parapets
restrained the panels from bowing under a temperature rise on the outside face of the panels
(Fig. 47). The restraint occurred at the roof deck and cracked the panels due to a buildup in
negative bending (Fig. 48). Short parapet panels, however, had two lines of support rather than
three lines and so were not restrained and did not crack (Fig. 49).
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GENSERT AND BRETNALL ON MASONRY BUILDINGS 71

-8&^pn.

S^^

FIG. 18—Case History II: Typical wind frame as should have been designed.

Case History VI
Moisture and thermal expansion properties vary with the brick and with respect to the bricks
principle axis. These factors can present problems when different bricks are used in the same
wall and when they are not laid in the same configuration. One building which presented this
problem had a panelized curtain wall consisting of a bottom soldier course of jumbo brick
topped by standard brick laid in running bond. The bricks in running bond expanded at a rate
4 '/2 times that of the soldier course bricks across their width. This differential in expansion
cracked the mortar joints of the soldier coursing as shown in Fig. 50.

Conclusions
These few case histories, selected from the hundreds of buildings the authors have docu-
mented, show that masonry becomes distressed when it is restrained by its supporting structure
or when the supporting structure imposes loads on the masonry which the masonry is not de-
signed to accommodate. Different properties or materials incorporated into a single system can
also lead to distress of the system. The clear lesson to be learned is that designers must closely
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
72 MASONRY

1 | ^ ^ * —

•tF

"^ir

FIG. 19—Caie History II: Deflection of typical wind frame as should have been designed.

STEEL T I E - B A C K
ANGLES

STEEL QRAVrrY
CONNECTION ANGLES

FIG. 20—Case History II: Single window panel, back view.

FIG. 21—Case History II: Crack at unreinforced opening.

FIG. 22—Case History II: Crack at unreinforced opening.

FIG. 23—Case History II: Cracks at window frame connection.

FIG. 24—Case History II: Shear displacement cracks in caulking.

FIG. 25—Case History II: Shear displacement cracks in caulking.

EXTERIOR EXTERIOR
FACE FACE

POTENTIAL
WATER
INFILTRATION

BRICK CORES BRICK CORES

ALIGN NOT ALIGNED

AS DESIGNED AS BUILT

FIG. 26—Case History II: Belt course sections.

FIG. 27—Case History III: Overall view of building.

FIG. 28—Case History III: Back view of north and south panels.

FORCE

ELASTIC/CREEP SHORTENINO
MOISTURE AND OF CONCRETE WALLS
TEMPERATURE
EXPANSION
OF PANEL

FORCE • w it

SECTION

FIG. 29—Case History III: Opposing volume changes in north and south panels.

< • < •
< • < • -CRACKS

4 •____— --^ ^ >

< • < >
<]> <, •
A • FORCE
T
—MA ..^ - .... L-"T

ELEVATION

FIG. 30—Case History HI: Tension due to temperature changes in north and south panels.

-CRACKS

* * UL'jJ

ELEVATION

CLASTIC SHORTENING OF SLAB

TBMmwnMB EXPANSION^
OF PANEL PLAN

FIG. 31—Case History III: Induced lateral force in north and south panels.

-SYMMETmCAL
ABOUT CEKIERUNE

-UNSTRESSED
SHAPE

-800 pti
leoo p*i
-2400 p«l

PARTIAL ELEVATION

FIG. 32—Case History III: Stress map of north and south panels due to lateral tension forces.

FIG. 33—Case History HI: Vertical reinforcing at connectors.

FIG. 34—Case History IV: Building wall.

FIG. 3S—Case History IV: Rusted X-bracing.

FIG. 36—Case History IV: Cracked and spoiled brick.

FIG. 37—Case History IV: Cracked and spoiled brick.

FIG. 38—Caw History IV: Welded connections.

FIG. 39—Case History IV: Cracked panel.

FIG. 40—Case History IV: Cracked panel.

FIG. Al—Case History IV: Split panel.

FIG. 42—Case History IV: Cracked mortar.

FIG. 43—Case History IV: Eroded mortar.

FIG. 44—Case History V: Overall view of building.

urnoML cewMCTOii

FIG. 45—Case History V.- Typical deep parapet panel.

PWFABHCATED
m C X PANEL

QMVITY CONNECTOm

CONCRETE
SPANDHEL BEAM FLOOR DECKINQ

LATERAL CONNECTOR

-OLASS WINOOW

FIG. 46—Case History V: Typical shallow parapet panel.

FIG. 47—Case History V: Deep parapet panel: restrained bowing.

FIG. 48—Case History V: Deep parapet panel: restrained causes cracking.

WARM

FIG. 49—Cflie History V: Shallow parapet panel: unrestrained bowing.

- ^

__ • Cracking

13^ M 2

M 1 > M2

FIG. 50—Case History VI: Differential moisture expansion.

and carefully examine the interaction of the masonry and its supporting structure, as well as the
interaction of the masonry materials with each other, and to design the overall system to interact
freely and without restraint.

DISCUSSION ON MASONRY BUILDINGS 91

DISCUSSION

J. G. Stockbridge^ (written discussion)—I personally had the opportunity to investigate the

cause of the masonry distress in three of the six buildings discussed in this paper, Case Histories
I, III, and V. In two cases, I worked for the building owner and on the third I worked for the
developer. In all three cases, Gensert and Bretnall worked for the company that provided the
mortar which was used in the buildings.
The masonry distress in all three buildings was caused by corrosion-induced cracking rather
than the causes cited in the Gensert/Bretnall paper. The mortar used was breaking down and
releasing tremendous amounts of chloride. Laboratory tests on samples removed from the
buildings discovered chloride levels 10 to 40 times (not percent) higher than the amount known
to cause corrosion of embedded steel. Galvanized coatings were being completely eaten away,
the base metal of embedded steel elements were severely corroding, and the buildup of the rust
product on the steel was causing cracking and spalling of the masonry.
Case History I: a bank—The paper fails to mention that the reason the horizontal joints were
unable to accommodate volume movements was because they were filled with rust product from
the severely corroding horizontal leg of shelf angles. Also, when inspection openings were cut
into the piers, it could be seen that severe rust buildup on the face of the vertical legs of the
angles were prying the piers apart and also causing cracking. The corrosion buildup on the
leading edges of the horizontal legs of the shelf angles at the alternate levels without movement
joints was shoving whole sections of mortar laterally out of the joints.
Case History HI: a hotel—Vertical cracking developed at the vertical reinforcing bars be-
cause they were severely corroding. When inspection openings were cut into the panels, it could
clearly be seen that the rust buildup on bars was causing the cracking. The cracks were radiat-
ing out from the corroding steel. The cracks were not externally induced.
In addition to the cracking at the vertical bars, corrosion-induced cracking and spalling was
occurring at corroding pencil rods at free edges of panels, vertical cracking was occurring at
corroding lifting lugs, and crescent-shaped spalling was occurring at corroding connections.
Cracks took longer to develop at the center bar than at the outer bars because rust-induced
outward bowing of the panel was restrained at the center bar.
Case History V: a mall—The cracking at the roof connections of the high parapets was not
caused by restraint of panel bowing. It was caused by pressure from rust buildup on the severely
corroding anchors at the roof line.
Corrosion-induced cracking was also occurring at anchors in spandrel panels and at lifting
lugs. There was even cracking occurring at anchors which had never been used and were at-
tached to nothing. The reason the cracking was more severe on the high parapet panels than on
the low parapet panels was because the severely corroding connections in the high parapet pan-
els were closely spaced at 16 in. on center while in the low parapet panels they were spread out at
40 in. on center.
W. G. Hime^ (written discussion)—It is distressing that Messrs. Gensert and Bretnall present
an analysis of curtain walls of six buildings without noting that: the masonry of five of the six
was made with mortar additive of the company that employed them; that severe corrosion oc-
curred to metals in contact with or embedded in the mortar; that engineers, chemists, and
metallurgists have attributed most of the distress experienced by those buildings to corrosion

'Wiss, Janney, Elstner Associates, Inc., Northbrook, IL 60062.

^Erlin, Hime Associates Division, Northbrook, IL 60062.
Copyright by ASTM
Copyright' 1988 byInt'l
AS FM(allInternational
rights reserved); Mon May 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
92 MASONRY

caused by the production of chloride ion through degradation of the saran latex component of
this mortar; and that they (Gensert and Bretnall) were experts employed by this company to
investigate those buildings. (The author of this discussion was employed by plaintiffs in lawsuits
filed against this company in relationship to the distress.)
Gensert and Bretnall's first example is particularly of note because it went to trial, and a jury
held against the company that employed them for over 25 million dollars including punitive
damages. According to public records, this company appealed the verdict and later settled for
about 19 million dollars. The records also indicate that the other buildings the authors describe,
but do not name, and which contained the same mortar, were the subject of suits that were
settled by this company for amounts ranging from a few million dollars to over ten million
dollars.
Considering that five of their examples contained the masonry in question, none of these
authors addressed the significance of such statistics.
C. H. Raths^ (written discussion)—Messrs. Gensert and Bretnall in the verbal presentation of
their paper elected to discuss masonry building problems using a bank and a hotel, both located
in Cleveland, a manor in the Philadelphia area, and an apartment building, also near Philadel-
phia, as examples. The bank project was constructed using in-situ brick masonry and the mor-
tar in question. Prefabricated brick panels were employed on the other three projects in which
both the hotel and the manor had panels fabricated with this mortar. The apartment house used
a proprietary grout in the panel instead of mortar.
The presentation of the paper regarding the three panel buildings indicates that a main cause
of the observed cracking and distress is differential structural responses between the panels and
the building frame. These structural responses, according to Gensert and Bretnall, result from
rigid connections attaching the panels to the structural frame. Because the connections are rigid
and restrain movements, they induce cracking stresses into the panels when the panels are acted
upon by temperature, creep behavior of concrete building frames, moisture expansion of brick,
building lateral shear forces, weak planes within panels, etc. Also, it is offered that panel crack-
ing directly over reinforcement results from a stress riser effect of the rebar within the brick
cores.
No mention is made by the authors of the extremely excessive amounts of the Cl^ within the
mortar in question nor the extensive corrosion suffered by the reinforcement and other embed-
ded metal (with and without protective coatings).
This writer, as well as other qualified engineers, has been involved in investigating, analyz-
ing, and testing these four projects. Relative to the panelized projects, visual evidence of distress
and deformation or failure about connections was not typically present. And, considering the
magnitude of forces which would be attracted to rigid connections, shearing and bearing
stresses should occur which would cause complete failure of the connection itself and the adja-
cent brick. Yet, neither distress nor structural failure of the connections has occurred other
than that resulting from corrosion caused by C\~.
The conclusions about panel cracking and distress generated by rigid connections rest upon
analyses which do not recognize load-deformation characteristics of the connections but instead
consider connections absolutely rigid. Structural testing by the writer has provided data on the
load-deformation spring characteristics of these brick panel connections. Analyses using con-
nection springs, which provide significantly reduced restraint to panel movements, has indi-
cated that the magnitude of restraint forces induced into panels are small and will not produce
cracking distress. The analyses coincide with the observed behavior.
Gensert and Bretnall suggest that the cracking in the panels of the apartment house resulted
from a stress riser effect in the brick caused by vertical panel reinforcement. In-depth petro-
graphic and related testing determined that the in-plane cracking of the panel was a direct

^Raths, Raths & Johnson, Inc., Willowbrook, IL 60521.

consequence of proprietary grout freeze/thaw deteriorations. Structural behavior of the panel

was not a factor in the cracking. Reinforcements within the cracked panels were not corroded,
and no chlorides were present in the grout.
Restrained differential movements between in-situ or panelized brick masonry and the build-
ing frame can and do lead to a certain amount of distress as correctly characterized by Gensert
and Bretnall. But, once cracking occurs about connections, the restraint is basically relieved
and further cracking does not develop. And, for mortars and grouts not containing Cl~, corro-
sion of reinforcements at these relief crack locations does not occur. The cracking described in
the paper cannot happen without some other factor being involved. As discussed by this writer,
CI "-caused reinforcement corrosion and freeze/thaw deterioration of grout are the other fac-
tors leading to the paper's reported cracking and distress. Thus, the conclusions reached and
presented by Gensert and Bretnall relative to problem causes are in error and the result of
incorrect analyses and/or improper site investigations.
G. P. Chacos^ (written discussion)—As the structural-engineer-of-record of the building de-
scribed in Case History III, I am obliged to submit additional information for those interested in
understanding the facts of this case. In addition to providing the structural design of the build-
ing, I reviewed the fabricator's design and shop drawings of the brick panels and have closely
studied the distress in these panels.
I take exception to many of the assumptions and conclusions presented by the authors in their
discussion of Case History III. Their investigation was one of many done on behalf ofthe com-
pany that employed them to prepare for lawsuits centering around claimed corrosive tendencies
of a mortar additive supplied by this company. It is surprising, if an unbiased investigation is
the objective, that their discussions do not include some reference to possible detrimental effects
of chemical interaction between the components of the panels.
There have been no problems with the structural performance of the brick panels. Load tests
of full-sized panels verified the composite design of the stiffened, reinforced panels prior to their
installation. Properly designed soft joints and flexible connections were used, and no instances
of fully compressed horizontal joints or fully compressed vertical joints have been reported. The
forces indicated on Figs. 29 and 30 cannot occur with the real connections.
On Fig. 31 the authors show the vertical cracks which occur at the locations of the three
vertical rebar in each of the most common panels on the north and south faces of the building.
The top floor panels, however, have four vertical rebar and four vertical cracks, totally inconsis-
tent with the "accordian" pattern of failure described. It is my observation that the cracks are
caused by corrosion of the rebar, and I dispute the authors' contention that corrosion of the
rebar came after the cracks occurred.
I agree with the authors that it is necessary to examine all possible sources of distress when
analyzing a problem, but the discussion of Case III does not present all of the facts and forms
conclusions based on an inaccurate model.

Author's Closure
The number of discussion papers received indicates that there is keen interest in the subject of
masonry curtain wall distress and that this topic should be a matter for lively debate. It should
be pointed out that the discussers, except for one, are an organized group of professionals work-
ing collectively for the plaintiffs which they cultivated throughout the country.
In response, additional information is offered:

"Gregory P. Chacos, Inc., Cleveland, OH 44115.

Case History I
Horizontal movement joints were provided approximately every 15 m (50 ft) and were filled
with 6.35-mm ('/4-in.) compressible material. Deflection of the supporting cantilever bracket,
moisture expansion of the masonry, and a temperature rise of 28°C (50°F) will close the joint.
The vertical legs of the shelf angles do not have enough stiffness to pry the piers apart. For the
prying force to be a factor, the piers on the front and side faces would have been displaced
outwards. In fact, piers were cracked on the front face only.
It is interesting to note that cracking occurred at high stress concentrations even though rust-
ing was not always present. Text books on masonry expansion and B.I.A. Tech Notes were
totally disregarded.
The recent history of this building is particularly significant. All masonry near steel framing
was replaced in 1979 using mortar containing no additives. All embedded metal was replaced
with new metal protected with multiple coats of epoxy paint. The reconstruction of the facade
followed the original design. Today, eight years later, the facade has become distressed in the
same locations and in the same pattern as the original distress. The origin of the distress is
clearly due to the structural action between the masonry and the supporting frame. The authors
are not alone in this opinion. Charles Raths, who examined the building at about the same time
as the authors, wrote in his summary report:
The visual examinations made by RRI resulted in identification of certain conditions which were of a
repetitive nature. Among repetitive conditions noted for the . . . office building were the following:
B2. Tapered vertical and horizontal brick cracks in the intermediate piers resulting from structural
frame deformations.

Case History HI
The owners of this building documented, very thoroughly and very carefully, the progression
of the distress in three separate (but not consecutive) years. A detailed study of this documenta-
tion shows that there is a very clear pattern, consistent from panel to panel, to the progression of
the distress. The total length of cracking at embedded metal is substantially less than the total
length of cracking away from embedded metal. If corrosion of embedded metals caused the
cracking, there would be no clear consistent pattern, and certainly no cracking where there was
no metal.
The authors are aware of only one load test of a stiffened panel. This test was performed by
laying a panel flat and having a group of men stand on it. The test was not conducted in a
controlled, scientific manner.
The steel connections between the panels and the structure are rigid and show signs of dis-
tress due to structural interaction between the panel and the structure. There were many docu-
mented instances of fully compressed horizontal joints. These joints closed because of the ex-
pansion of the brick and the creep shortening of the concrete frame.

Case History V
It is significant that the tall parapet panels of this building cracked on only two sides of the
building—those with the greatest thermal exposure. Yet support details and construction are
identical on all four sides. Thermal differentials of as much as 22°C (40°F) were recorded across
the thickness of a cracked panel on a sunny but relatively mild day, with the outside face being
the warmest. Bowing of the panels did occur.
Cracking at the lifting bolts or lugs is due to extremely high prying forces generated during
handling. These lugs were not only used to lift the panels but also to rotate them 90° from their

as-built orientation to their in-place orientation. Entry of water and formation of ice during the
winter propagated initial cracking.

Smnmaiy
The distress seen in all these masonry curtain walls has a common origin: structural interac-
tion between the curtain wall and the supporting frame.
Of conclusive evidence: the patterns of distress in masonry using the mortar additive also
occur in masonry which does not use the additive. Further, nonmasonry curtain wall systems
investigated by the authors exhibited similar distress for the same reasons.
The lesson learned from these case histories is that curtain walls must be designed to allow for
the movements that do and will occur between the curtain wail system and the supporting
frame.
It is most surprising and distressing to the authors that many in the profession investigate
only to the seemingly obvious and compelling conclusion for their clients, yet refrain from fur-
ther research to seek out the underlying and motivating factor of a structural phenomenon, in
this instance that a certain masonry additive causes masonry distress regardless of design appli-
cation or construction technique. Yes! Some chemical additives can accelerate corrosion. But
what is required to initiate corrosion? A basic fact of chemistry is that water and oxygen pro-
mote corrosion. If masonry cracks from structural distress, it is obvious that water enters the
masonry and corrodes the steel. Conversely, if the masonry does not admit water to the steel,
there is no corrosion, as has been documented.
It is our contention, corroborated by several decades of private practice experience and cou-
pled with extensive university teaching and research, that the interaction of structural systems
and the systems they support are so sensitive as to require isolation of the supported system.
Perhaps there is a greater lesson to be learned here. Let us benefit from the documented past.
Let us understand the problems. And in so doing, reduce our forensic efforts and direct our
energies toward creating structures that will stand the test of time.

Influence of Tie Flexibility, Relative

Length, and End-Boundary Condition
on Brick Veneer-Metal Stud Flexural
Bond Stress^

REFERENCE: Chin, I. R., Krogstad, N. V., and Monk, C. B., Jr.,"Iiifliieiice of Tie FlexibUIty,
Relative Lengtii, and End-Bonndar; Condition on Bricli Veneer-Metal Stud Flexural Bond
Stress," Masonry: Materials, Design, Construction, and Maintenance, ASTM STP 992, H. A.
Harris, Ed., American Society for Testing and Materials, Philadelphia, 1988, pp. 96-117.

ABSTRACT: Based on insights from the Clemson test report on the wall system under discussion,
the authors analytically examined the issue of the relative stiffness between the brick veneer and
the metal stud.
Much controversy has surrounded the contention that the inherent stiffness of the brick veneer
will result in critical flexural bond stresses in the veneer when laterally supported on metal studs.
Using ordinary conventional relative stiffness methods of analysis, the brick veneer can attract
85% or more of the wind load, whereas the metal studs attract 15% or less. Even if the studs are
designed to take full wind load at a span-deflection ratio of 360 to 600, ordinary analysis may still
indicate that critical flexure bond stresses will occur in the veneer prior to full participation of the
stud.
This paper shows that if the inherent tie flexibility, normal shorter metal stud length, and usu-
ally free top boundary wall condition are taken into account, critical flexural bond stresses do not
exist in the brick veneer on typical brick veneer/metal stud walls. Recommendations on the utili-
zation of inherent boundary conditions to minimize the development of critical flexural bond
stresses in the veneer are presented.

KEY WORDS; ordinary conventional analysis, relative stiffness, computer analysis, brick veneer,
metal studs, flexural bond stress, span-deflection ratio, metal tie flexibility, spring constant, fix-
ity, soft joint, relative height

Brick veneer with metal studs (BV/MS) construction has been used in the United States since
the mid 1960s. Early field experience was obtained in the walls of the Malaysia Pavilion build-
ings of the New York World's Fair (1963). Thus, there has been over twenty years of commercial
application. Development of the concept was initially m a d e by the metal stud producers
through their trade association, the Metal Lath/'Steel Framing Association ( M L / S F A ) , and was
based upon the successful use of brick veneer/wood stud walls in low-rise residential construc-
tion. By 1979 the Brick Institute of America (BIA) published Technical Note No. 28B, incorpo-
rating the brick industry's recommendation for B V / M S construction. The BIA criteria was at
variance with the earlier practice of M L / S F A in at least two ways:

'Senior consultant, architect II, and senior consultant, respectively, Wiss, Janney, Elstner Associates,
Inc., Northbrook, IL 60062.
^Metric conversion for units in this paper: in. = 2.54 cm; 1 ft = 30.48 cm; 1 psi = 6894.0 N/m^; 1 psf =
47.88 N/m^; 1 lb/in. = 175.1 ti/mK

1. The use of adjustable 9-gauge wire ties instead of 16 gauge, Vs-in.-wide corrugated ties to
tie the brick veneer to the metal stud back-up wall.
2. Limiting span-deflection ratios of the metal studs under full wind load to 600 to 720 with
900 preferred instead of 240 to 360.
By 1980 ML/SFA had had over a decade of apparently successful BV/MS wall installations
on thousands of buildings with their criteria. The BIA recognized that BV/MS walls were in-
creasingly being used on high-rise construction, well over three stories, where high lateral forces
are encountered. Accordingly, their criteria were directed at assuring greater wall stiffness to
minimize the development of critical flexural bond stresses in the brick veneer.
To resolve these conflicting criteria, the two trade associations (BIA and ML/SFA) sponsored
a research program at Clemson University to investigate the BV/MS system. A report was is-
sued 23 April 1982 [/].
During the 1970s designers became aware that while the majority of BV/MS walls performed
satisfactorily, some did not. Five papers plus a workshop were dedicated to this subject at the
recent Third North American Masonry Conference (June 1985). This demonstrated an active
professional interest in the subject. No paper offered evidence of any structural collapses; how-
ever, some papers reported on ordinary structural analysis, which showed that ultimate flexural
bond stresses in the brick veneer on BV/MS walls were sometimes exceeded, which could pre-
sumably lead to excessive water penetration. Following the June conference, the BIA sponsored
a round-table conference in September 1985 to resolve the growing concern over this matter.
The results of this conference have not yet been made public.
The authors have been challenged in reviewing this issue to explain the apparent success of
the majority of BV/MS installations. In their own and in their colleagues' field experiences,
many more successes than failures of BV/MS walls have been observed. This paper shows why
this is the case. A careful review of the Clemson tests provided some helpful insights.
The classical ordinary strength of materials analysis, based on the relative stiffness between
the brick veneer and the steel stud, usually indicates that critical flexural bond stresses exist in
the brick veneer. However, the Clemson tests show that if the difference in length between the
brick veneer and the metal studs, the boundary conditions of the brick veneer and the metal
studs, and the flexibility of the metal ties connecting the brick veneer to the studs are taken into
account, critical stresses may be avoided. This paper examines these conditions.
In examining these conditions, a metal stud span-deflection ratio of 600 under full wind load,
as recommended by the BIA, is principally utilized. A metal stud span-deflection ratio of 360,
as earlier recommended by ML/SFA, is also utilized to explain the apparent success of typical
BV/MS walls designed under this earlier criteria.

Material Properties
At the outset it is well to establish a reasonable range of brick masonry properties based on
the best available test data. The brick industry, through its Structural Clay Products Research
Foundation (SCPRF), published in October 1964 the results of a National Testing Program
based on 31 varieties of regional brick. Much of this work can be found in "Progress Report No.
1 Small Scale Specimen Testing, October 1964" by SCPRF [2]. Figures 1 and 2 show the results
of the test program, which used Type S mortar throughout. From this work the statistics in
Table 1 have been compiled.
The recommended allowable flexure bond stress for masonry walls by the Building Code Re-
quirements for Engineered Brick Masonry [3] is:
1. Type S and M mortar = 36 psi.
2. Type N = 28 psi.

M«0 r 4 t ^ 0 ^ i e \ f ~ f ~ 9 i 9 t o o ^ mt^tf r~ h- ^ to n oi fn oi ^ M ^ O
OO O^ <D OK 0^ ^ ' Q 9 i r - r - - O t « < 0 o^(^ao ( h i * ^ h - h - o t o

® tn (^ o v%mmcD o n roo) 9 s 9 i 9 t m m r ^ m o \ H v o V M cu
IS S S asssssssaissas'-adsa

I
^ ^ ^ ^ o ^^ I J« J so \0 lOtf\\o lO « xo \o \o «
J J «
: :
O C 0 l ' ^ O o « O H 0 - * Oi-» O l ^ o < n m ^ ^ < ^ • » » ^ o - » h-r-^"©
''^sa-4'n uNir\.« r— < 0 ' * «*» «> w - cu m vo so >o

W O -^f-\0 O Ol^O^HtO'^eO'* OICO tr\ tn m rrt Ci r - t * - h - « 0 irs*o O s \ o

>H <n ^ N

s a ^ ^ t?\ ^ O- f j

^ ^ \0 ^ ^ ^ ^ J t •" ! \0 \ 0 lA \ 0 vO VO •O vO \ 0 \D \0
t i
i I

I
O O O O i O V O - ^ ' n i A 00 VO' f^ ^ o -* o ^ >o
'»*^ r^ r^ n-i
8 S 2 S S S^
^ S) f - ^ y ) ON O *A
nd (M O O O O

^n O N N - 4 C O > C n c o t o i o f - CO lA
ro ^ - i
ni r^ -^ aa
H CO (*^ n o
m oi cu
^ r> O O -* >4
*"' i* si s
>^ n tf\ o o o ^ < * ^ o - » o f - w - » i —
jy c^ *o i«- ©» f » Ov 1^ ^
8d
o 0 ' O o o o o c ^ - o o o %o O OO ( h V O O O O 9 O
S
H O v ^ S s S ^ o ^ o8 m
j-« o (n >o
Ml
»rt « m o trs (u NO CT> - * <VI 9 \ so

* 1 1 a s 5-1 s $ a g g I s § I H I 2 1 1 1 1
H N oj i > o j f i ! r

oi o m H <o ffi O O - H H W C O O N O O
2:^J:-

t~ 0\ Jf ^ V\ 3 a <^ - ^ ( ^ ^ 0
« H H -^ ftl
S N O UNO
J CS
ifsK^ ||*se§| S 5 •S 9,-»3

nills. niiiiiiiiiir 3 ..!.«

^
I - -s
•H O NO H <n {?N O ^ O •'^ NO
\' 4
rtS'âissâaaaas'âaaaassa'-''"
< D N O M ^ h-WNO

5sa8"sss3&«asasj»aass8i»sa'^'^s
f^-* V \ > 4 N O O ^ C O ^ ^ N O O i f N h - O N O N e O N O

.1 1
<||R£?|0S|gepS«9S^«gg^egR|^gSg3g i
d5£BS;iSE:jeB!3l6aBB9eQSSBBBe ti • fl§

TABLE l—Statistics based on SCPRF brick

testing program.

FLEXURAL BOND STRENGTH

Average = 123.18 psi
Standard deviation = 31.62 psi
Coefficient of variation = 25.67%
Number of samples = 140

MODULUS OF ELASTICITY, E
Average = 2.245 X 10* psi
Standard deviation = 0.882 X 10* psi
Coefficient of variation = 39.28%
Number of samples = 218

For the purposes of this paper, the authors assume that these "allowable" stresses are appli-
cable to BV/MS walls. These values may be increased 331/3% when wind is acting to 48 and 37
psi, respectively. For Type S mortar, the 36 psi could be increased three-fold to 108 psi without
the brickwork cracking. It is conceivable that some walls in service have experienced 108 psi
without failure. Further, very high wind loadings are events that happen relatively infrequently
in the life of a building. In the work to follow, the authors examine the conditions of BV/MS
walls that results in calculated stresses that fall within the 48 to 108 psi range for Type S or M
mortar (or 37 psi to 84 psi range for Type N mortar). Examination of the modulus of elasticity of
brickwork between the values 1 to 3 million psi, which reflect average values, is incorporated in
the analysis.

Ordinaiy Strength of Materials Analysis

The deflection criteria established by BIA and ML/SFA for BV/MS walls assumed that the
metal studs take the total design wind load. Ordinary relative stiffness analysis indicates that
the brick veneer frequently takes more than half the load, which commonly stresses the masonry
to well over the "allowable" stresses, even in excess of ultimate stresses in some cases. Although
the recommended span-deflection ratio of 600 for the metal studs by the BIA is intended to limit
the deflection of the studs, inherently the veneer is very stiff and not able to tolerate span-
deflection ratios of 600 or less without cracking. To examine these ideas, the following plot was
made.
From ordinary flexural theory it may be shown that: (X/A) = (2AE/Sf) -r (£/d). This can be
plotted on a log-log plot as a series of straight lines as follows:

log (Z,/A) = log (24£:/5/) - log (L/d)

Figure 3 is a plot of this equation. The abscissae are L/d values or wall slendemess ratios; the
ordinates a r e i / A values or span-deflection ratios. Two regions of brickwork modulus of elastic-
ity are shown: E = i 000 000 and 3 000 000 psi. Each region is bounded by two levels of flex-
ural bond stress:/;, = 48 and 108 psi. The first represents the "allowable" for Type S mortar
increased by 33 ' / 3 % for wind or 36 psi X 1.33 = 48 psi. The second is an assumed high stress
level where brickwork cracking is not yet expected to develop, 36 X 3 = 108 psi. Values within
this range exceed the "allowable" stress for masonry walls but are below the ultimate cracking
stress. Also plotted are the ranges of slendemess ratios recommended by BIA and ML/SFA for
the design of the backup system. Because the backup system and the brick veneer are assumed
to deflect together in an ordinary analysis, it is clear that the brick veneer have to take the
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
100 MASONRY

BRICK PRISM COMPRESSIVE STREBGTHS (DCS Laboratory T e s t s )

PRISMS
HOPTAR
:OHPRESSIVE STREBGTH " ' m COMPRESSIVE BRICK
STRENGTH COMPRESSIVF
CODE n
X R 1 ' V
E ».
pst
AIR CURED WITB
STRENGTH
pel
a
pBl PRIS>B»*
pal * a X
psl X 1 0 ' X

LL 3805 661 3>iO 8.9 1.82- U78 U39 6066

LM "•ssa 555 283 6.2 i.n J82 11)39 6306
Ft 2658 126 6i> 2.U 1.33 500 1V73 631)0
KM us JO 811 '•52 Lo.r 1.1*9 352 1773 6868
GM l»8U3 521 269 6.3 2.73 6Uj 1157 7501)
JM lt790 1018 519 10.8 1.89 395 1319 7806
HM l>eU9 522 299 6.2 2.63 5U2 1665 8039
DH U58U 350 19k I..2 3.'.5 753 171)8 8717
FM 6516 i»83 265 U.l 2.02 310 lltliO 9161
AL 14002 1039 572 U.J l.ltl 352 1580 9267
KH 5880 1.96 251 U.J 3-33 566 1773 9708
!M 5891 582 306 5.2 I.7U 295 171)8 9730
CL "•583 510 255 5.6 1.83 399 11)1)0 9810
MH 5851 99 52 0.89 2.70 1.61 1319 10711
CM 1)689 371 197 It.2 3.66 781 1363 11502
AH 6186 298 151. 2.5 2.70 Uj6 1319 12015
GH 61.39 59't 326 5-1 3-57 551* 1157 12138
DH itiH3 31'4 161* 3.7 2.70 612 1665 12261)
LH 6105 99 265 U.U 2.38 390 1166 12336
FH 6087 19't 112 1.8 3-03 U98 1773 12788
JB 6k7k ••59 230 J.6 303 it68 1665 13052
AM 6349 296 I6it 2.6 a.uu. jSit 1580 11)388
BE I1867 629 3it5 Y.i 3.08 UUl 1336 lUc60
BM i»803 7'» ftS 0.89 3-33 693 935 15300
HE '»937 210 112 2.3 2.63 533 880 16093
EB 5595 9li6 ••98 8.9 3.70 661 nw 161)1)0
CB 3^ 6Q08 U95 26l 3.8 li.l7 613 1363 17838

HL igfii* 55 28 l.U 0.67 3'tl 11)91 i«i)89

BR i>29e 585 293 6.8 2.11 U92 1580 6370
BM ISl"* 305 153 11.7 0.69 525 1580 3968
t

SCR 5 6281 296 130 2.1 3.03 1)82 1287 U771

..1
•Nodulua of Elasticity, Initial tangent. ••See Table 3-3

FIG. 2—Results from the National Testing Program published by the Structural Clay Productions Re-
search Foundation.

BRICK PRISM CCmmSSm SlREnCTHS

(Co^relal laboratory Teits)

PRISMS MCBTAR
COSE^SSIVE !ilRENCm " ' • COMPRESSIVE miCK
BRICK D
STREHGIH C0MIW8SIV1
psl STRENGTB
X R • V
m Wf, AIR CURED WIIB
PRISMS** psl
pel psl p«l pal X 1 0 ' ^
f X
u, It 225; Ul»l 5U3 2'>.0 0.792 350 2082 6066
tN 5 36S6 16U 617 17.0 0.952 263 2082 6306
FL 5 2391 889 399 16.7 O.6U5 270 2082 63>i0
KM 5 2437 767 313 12.8 0.833 jl^l 1708 6868
CM 5 2631 1093 '•37 16.6 o.goi 3*2 713 750'»
JM 5 3800 620 316 8.3 2.085 5'»9 2779 7806
Ml 5 3825 995 357 9-3 1.1)90 390 713 8039
DM 5 3807 380 lUS 3-7 3.i)U8 906 897 8717
m 5 1>UU3 hos 212 U.8 1.520 3'»2 ac82 9161
AL 5 3729 II18O 576 15.5 0.952 255 1380 9267
KB 5 3U29 1007 He 12.1 1.350 39't 1708 9708
m 5 sgn *95 251 6.U 1.099 281 SUSi 9730
CL 5 5182 1380 502 9-7 2.280 Ititf) 1588 9810
MB 5 35W 615 294 8.3 1.500 ••23 713 10711
CM 5 '•952 620 29lt 5.9 3.060 618 1588 11502
AH 5 5066 II85 220 i>.l> 1.750 Jhi 1380 12015
ca 5 U719 1098 501 10.6 2.220 kio 713 12138
DB 5 li207 575 226 5.lt 3.57JI 8J*9 897 12264
la 3 5617 »63 261 *.7 i.itao 263 2082 12336
ra 5 5357 1055 507 9.5 1.730 323 2082 12782
JB 5 5358 1520 581 10.9 lilt82 277 2779 13052
AM 5 6018 635 SSI 3.8 2.060 3't2 1380 14388
BB 5 5556 1680 678 12.2 2.'•50. Itlfl 215'^ 14760
RM 5 5650 382 169 3.0 SM9 '•32 2875 15300
BB 5 5722 53'' 198 3.5 2.632 I46O 2875 16093
EB 5 i>li08 1090 ItliO 10.0 2.469 560 2'^23 16440
CH 5 5760 530 225 3-9 2.730 Ii7'» 1588 17838

*Hadulu« of EUstlelty, Initial tangent

**Cuplleatlon* of vmlues In this colunn Indicate tbat one set of 6 air cured
2 In. by 2 In. cubes was iBdr froa the aortar batch but tbat tvo or more seta
of prisma vere aide froa tbat batch.

FIG. 2—Continued.

• X.-, . - , , . , 'r' ' ii''CA_DtSIGN

^
Nd, V 1 • • . 1 1 •

FIG. 3—Parametric study showing height/deflection values as a function of slendemess ratio values for
constant values of wall modulus and wall stress.

majority of the load or crack. Brick masonry walls cannot normally tolerate, without cracking,
span-deflection ratios below the £/A = 900 limit.
To examine this issue with greater precision, graphs were developed which express the flex-
ural bond stress of the brick veneer, (f, psi) based on the following equation

/ = l(qL^)/l7.S2l] H- [1 + (zq V)/{25.40SE)]

where
q = wind load, psf,
L = wall height, ft.

E = brick modulus, psi, and

2 = span-deflection ratio of metal studs.
The equation is derived from ordinary strength of materials analysis, assuming that the total
wind load is shared between the brick veneer and the metal studs according to their relative
stiffnesses. For the purposes of simplifying the equation, the reduction of tensile stress due to
the weight of the masonry has not been included. Figure 4 is a plot of the stresses in the brick
veneer for the case of 2 = 600 assuming the stud is hinged top and bottom. For this case, the
maximum stress in the brick veneer ranges from about 120 psi, when E = I 000 000 psi, to
about 260 psi, when E = 3 000 000 psi, showing that the flexural bond stress is sensitive to
values of E.
The effect of metal stud fixity on stresses in the brick veneer with 2 = 600 was also deter-
mined by the ordinary analysis, as shown in Figs. 5 and 6. When the bottom of the metal studs is
fixed and the top of the studs is hinged, the maximum stress in the brick veneer ranges from
about 60 psi when £ = 1 000 000 psi to about 140 psi when £• = 3 000 000 psi. When the top
and bottom of the studs are fixed, the maximum stress in the veneer ranges from about 40 psi
when E = 1 000 000 psi to about 80 psi when £• = 3 000 000 psi. The stresses in the brick
veneer are therefore also very sensitive to stud fixity. Generally, the stresses in the veneer are
greater than the assumed 48 psi "allowable" limit, but many are less than 108 psi when stud
fixity exists.
In actual construction practice it is probable that neither a hinged-hinged or fixed-fixed stud
end boundary condition exists. The bottom of the metal studs is normally square ended and fits
directly against the web of the runner. The top of the metal studs, on the other hand, is gener-
ally held clear of the web of the top runner to allow for vertical building movement.
It is interesting to note that the span-deflection ratio of metal studs hinged at top and bottom
changes from 600 to about 1445 when the bottom of the studs are fixed. This benefit requires no
increase in the metal stud cross sectional area; however, some design effort is needed to assure
bottom fixity.
While taking advantage of stud fixity can reduce flexural bond stresses, other benefits are
possible by a careful review of the Clemson report as discussed in the next section.

Computer Analysis Including Tie Flexibility

The ordinary analysis assumes that the brick veneer and the backup system move together
and share the same span. In actual walls, however, the backup and veneer have complex bound-
ary conditions which allow the walls to move separately. This section examines the benefits of
the following:
1. The tjrpical soft joint installed at the top of the wall under the shelf angle to accommodate
vertical movements.
2. The difference in height, between the usually taller brick veneer and the shorter metal
stud.
3. The utilization of the inherent spring property of the metal ties used to tie the brick veneer
to the metal studs.
The top soft joint, the taller brick veneer, and the spring action of the metal ties all contribute
to reduce wall stresses as will now be shown.
A computer model was developed using the brick veneer, the metal studs, and the metal ties
as a frame network. Figure 7 is a schematic layout of the computer model prepared for running
on a STRESS program (Version 1 11-81). The ties were modeled as pin-ended springs. The
range of spring constants: 200, 2000, and 20 000 lb/in. is based on metal tie tests contained in
the Clemson report. The modulus of elasticity for the brick veneer was again bounded between

® _,

£§
o

I-'
a

+ ^

I
1
^sd u| peoi pujM

^sd u| pBOi pu|M

I
isd u| peoi pujM
1

« g
•H
s.m
o
1
£ o
o
1
•B o
o
o
1 n
II
!
=: m d
5 E

|6d u| peon PU!M

I
i

5
I
|sd u| peon PUjM
I

S
u. a
o

11
s • O

)8Cl U| PBOI PU!M

Flexible Wall Tie

r ^ — •

CD
T-
Masonry Wall O 0
*
" ^ <D

m
CD

*CD

<i °
<D

r
FIG. 7—Diagram of computer model used.

1 000 000 and 3 000 000 psi. Thus, for each wall height studied there are six combinations of
material properties. Four stud heights were studied: 8 ft 0 in., 9 ft 4 in., 10 ft 8 in., and 12 ft 0
in. These were selected as being the range of most practical interest. Common stud heights in
the majority of BV/MS walls range from 8 to 10 ft. The height of the brick veneer was assumed
to be 16 in. taller than the studs to reflect actual installations, jrielding veneer heights of 9 ft 4
in., 10 ft 8 in., 12 ft 0 in., and 13 ft 4 in. All cases studied assumed the base of the brick veneer
to be hinged; the top was free. For the metal studs, the base was hinged or fixed and the top
hinged. Six levels of wind loads were selected: 15, 20, 25, 30, 35, and 40 psf. In order to com-
pare the results with the previous analysis, the reduction of tensile stress due to the weight of the
wall was ignored.
Several parametric studies were undertaken to study the sensitivity of the flexural bond stress
in the brick veneer. As the program modeled each mortar joint as a separate network joint, it
was possible to scan all the joints and print only the maximum flexural bond stress. Figure 8
shows results with the metal studs designed with a span-deflection of 600 under full load. In
addition to selecting the maximum flexural bond stress, the maximum tie load and the displace-
ment at the top of the brick veneer are shown. Figure 9 shows the influence of fixing the base of
the stud.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
108 MASONRY

MAXIMUM STRESS IN MORTAR

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 200. 1000000. 31.6 38.5 45.3 51.9 58.4 64.8

112.0 200. 3000000. 33.6 41.2 48.5 55.6 62.7 69.6
112.0 2000. 1000000. 46.7 56.3 64.4 71.5 78.0 83.7
112.0 2000. 3000000. 56.5 70.3 82.6 93.6 103.9 113.4
112.0 20000. 1000000. lilt.5 56.3 67.0 76.6 85.3 93.3
112.0 20000. 3000000. 55.9 72.4 88.1 102.8 116.7 129.9

128.0 200. 1000000. 41, 50.4 59, 67.3 75.6 83,

128.0 200. 3000000. 45. 56.0 65. 75.2 84.5 93.
128.0 2000. 1000000. 56. 67.5 76, 84,4 91.2 97,
128.0 2000. 3000000. 72, 89.9 105, 119.0 131.5 143,
128.0 20000. 1000000. 53, 66.8 78, 88.8 98.0 106,
128.0 20000. 3000000. 70, 90.7 109, 127.3 143.8 159,

144, 200. 1000000. 51, 61, 71, 81.3 91.0 100 5

144, 200. 3000000. 59, 72, 84. 95.7 107.3 118 7
144, 2000. 1000000. 65, 77. 86. 93.8 100.2 105 7
144, 2000. 3000000. 89, 109, 127. 143.1 157.3 170 3
144. 20000. 1000000. 61, 75, 87, 97.8 106.9 114.8
144, 20000. 3000000. 84, 107. 129, 148.8 167.0 L83.7

160.0 200. 1000000. 60.9 72.6 83.6 94.3 104.8 115.0

160.0 200. 3000000. 74.5 89.9 104.4 118.4 132.1 145.5
160.0 2000. 1000000. 73.5 84.9 94.0 101.2 107.3 112.6
160.0 2000. 3000000. 104.9 127.9 147.6 164.7 179.9 193.5
160.0 20000. 1000000. 67.9 81.9 93.4 103.0 111.3 118.3
160.0 20000. 3000000. 98.2 124.2 147.6 168.8 188.1 205.5

MAXIMUM LOAD IN TIE

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 200. 1000000. 48.7 61.3 73.7 86.0 98.1 110.2

112.0 200. 3000000. 50.1 63.1 75.8 88.4 101.0 113.4
112.0 2000. 1000000. 74.8 92.1 108.0 122.7 136.6 149.8
112.0 2000. 3000000. 81.3 101.4 119.9 137.1 153.5 169.2
112.0 20000. 1000000. 92.8 116.9 139.2 160.1 179.8 198.4
112.0 20000. 3000000. 105.4 134.5 161.9 187.9 212.9 236.7

128.0 200. 1000000. 50.0 62.6 74.8 86.9 98.9 110.9

128.0 200. 3000000. 52.3 65.5 78.4 91.0 103.6 116.1
128.0 2000. 1000000. 75.7 92.5 107.7 121.8 135.0 147.6
128.0 2000. 3000000. 84.6 104.9 123.5 140.8 157.2 172.7
128.0 20000. 1000000. 106.1 129.4 155.7 179.8 202.1 222.8
128.0 20000. 3000000. 111.0 140.2 167.5 193.1 217.5 240.7

144.0 200. 1000000. 50.6 62.9 74.9 86.7 98.5 110.1

144.0 200. 3000000. 54.0 67.3 80.2 92.9 105.5 117.9
144.0 2000. 1000000. 75.2 91.2 105.7 119.0 131.5 143.4
144.0 2000. 3000000. 86.5 106.7 125.1 142.0 158.0 173.1
144.0 20000. 1000000. 123.5 155.3 183.4 208.5 231.4 252.3
144.0 20000. 3000000. 114.5 143.4 169.9 194.7 218.0 240.0

160.0 200. 1000000. 50.4 62.4 74.1 85.6 96.9 108.2

160.0 200. 3000000. 55.2 68.5 81.4 94.0 106.5 118.8
160.0 2000. 1000000. 73.8 88.9 102.4 115.0 126.8 138.1
160.0 2000. 3000000. 87.5 107.2 125.0 141.4 156.8 171.2
160.0 20000. 1000000. 145.5 179.2 208.3 233.7 256.5 277.0
160.0 20000. 3000000. 118.3 152.2 183.2 211.7 238.1 262.6

FIG. 8—Wall results from studs designed for L/600.

MAXIMUM TOP DISPLACEMENT

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 200. 1000000. 0.294 0.362 0,429 0,495 0.560 0.625

112,0 200. 3000000. 0.303 0.374 0,444 0,512 0.579 0.647
112.0 2000. 1000000. 0.056 0.067 0,076 0,085 0.093 0.100
112.0 2000. 3000000. 0.061 0.073 0,085 0,095 0.105 0.114
112.0 20000. 1000000. 0.011 0,013 0.014 0.016 0.017 0.018
112.0 20000. 3000000. 0.012 0,014 0.016 0,018 0.020 0.022

128.0 200. 1000000. 0.295 0,362 0.428 0,492 0.555 0.618

128.0 200. 3000000. 0.310 0.381 0.451 0,519 0.586 0.652
128.0 2000. 1000000, 0.054 0.064 0.073 0,081 0.088 0.095
128,0 2000, 3000000. 0.060 0.072 0.083 0.094 0.103 0,112
128.0 20000, 1000000. 0.010 0.012 0.013 0.014 0.016 0,017
128,0 20000. 3000000. 0.011 0,013 0.015 0.017 0.019 0.020

144.0 200. 1000000. 0.293 0.358 0.421 0.483 0.545 0.605

144.0 200. 3000000. 0.314 0.386 0,455 0.522 0.589 0.655
144.0 2000. 1000000. 0,051 0.061 0,069 0.076 0.083 0.090
144.0 2000. 3000000. 0,059 0.071 0.082 0.091 0.100 0,109
144.0 20000. 1000000. 0,009 0.011 0.012 0.013 0.014 0,015
144.0 20000, 3000000. 0.011 0.013 0.015 0.016 0.018 0,019

160.0 200. 1000000. 0.288 0.351 0.411 0.471 0.529 0.587

160.0 200. 3000000. 0.317 0.388 0.456 0.523 0.588 0.653
160.0 2000. 1000000. 0,049 0.057 0.065 0.072 0.078 0.084
160.0 2000. 3000000. 0,058 0.069 0,080 0.089 0.097 0.105
160.0 20000. 1000000. 0.009 0.010 0,011 0.012 0.013 0,014
160.0 20000. 3000000. 0.010 0.012 0,014 0.015 0.017 0,018

FIG. i—Continued.

Figure 10 compares the flexure bond stress in the brick veneer as determined by ordinary
analysis and by computer analysis. As shown in this figure, the flexural bond stresses in the
brick veneer, as determined by the computer analysis, are lower than the stresses as determined
by the ordinary analysis. For a metal stud wall with a span-deflection ratio of 600 and a metal tie
spring constant of 2000, the stresses analyzed are as shown in Table 2.
Based upon Table 2 information, it can be seen that the stresses in the brick veneer as deter-
mined by the computer analysis are significantly less than the stresses determined by the ordi-
nary analysis and that fixing the base of the studs is very effective in reducing flexural bond
stress in the brick veneer.
A study of the computer results shows that as the tie stiffness is reduced, flexural bond stress
in the brick veneer is reduced. A 1:100 change in tie stiffness changes the stress roughly 1:2.
For low tie stiffness (200 lb/in.) the maximum top displacement becomes excessive; for high tie
stiffness (20 000 lbs/in.), the maximum load in ties becomes excessive.
When the top of the brick veneer is pinned, the beneficial effect of the spring ties is negated
and will result in the brick veneer receiving more load than if the ties had been completely rigid.
A study of the computer results also shows that as the modulus of elasticity of the brickwork is
reduced, flexural bond stress in the brick veneer is reduced. A 1:3 change in the modulus of
elasticity changes the stress less than 1:2. For a 1 000 000 psi modulus of elasticity, the maxi-
mum flexural bond stress is under 100 psi. For a 3 000 000 psi modulus of elasticity the value is
under 200 psi. The latter value falls to under 100 psi when the stud base is fixed.
Many BV/MS walls of commercial interest are not over 10 ft tall. The computer analysis
reveals that these walls can withstand a wind load up to 40 psf without cracking when designed
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
110 MASONRY

WALL RESULTS FOR STUDS DESIGNED FOR L / 6 0 0 ; BOTTOM OF STUDS FIXED

MAXIMUM STRESS IN MORTAR

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF itOPSF

112.0 200. 1000000. 22.6 28.9 35.3 41.8 48.3 54.8

112.0 200. 3000000. 24.3 31.2 38.0 44.9 51.9 58.8
112.0 2000. 1000000. 26.5 31.0 34.7 38.1 41.2 44.1
112.0 2000. 3000000. 35.4 42.5 49.0 55.1 61.0 66.7
112.0 20000. 1000000. 30.6 37.7 43.7 48.8 53.3 57.1
112.0 20000. 3000000. 39.4 49.9 59.3 67.8 75.5 82.6

128.0 200. 1000000. 29.0 36.9 44.8 52.7 60.6 68.6

128.0 200. 3000000. 32.5 41.6 50.5 59.4 68.3 77.3
128.0 2000. 1000000. 31.8 36.2 40.0 43.5 46.6 49.5
128.0 2000. 3000000. 45.7 54.5 62.2 69.2 75.8 81.9
128.0 20000. 1000000. 36.6 44.4 50.9 56.4 61.2 65.3
128.0 20000. 3000000. 49.0 61.4 72.5 82.4 91.4 99.5

144.0 200. 1000000. 34.4 43.7 52.9 62.2 71.5 80.7

144.0 200. 3000000. 41.0 52.1 63.2 74.1 85.1 96.1
144.0 2000. 1000000. 36.5 41.2 44.9 48.0 50.7 53.2
144.0 2000. 3000000. 55.1 65.0 73.3 80.6 87.2 93.2
144.0 20000. 1000000. 42.0 50.1 56.7 62.1 66.6 70.5
144.0 20000. 3000000. 58.9 73.3 85.9 97.0 106.8 115.6

160.0 200. 1000000. 39.4 49.3 59.1 69.2 79.4 89.6

160.0 200. 3000000. 50.5 63.3 76.1 88.8 101.5 114.1
160.0 2000. 1000000. 40.6 45.2 48.6 51.4 53.6 55.6
160.0 2000. 3000000. 63.7 74.0 82.7 90.2 97.1 103.3
160.0 20000. 1000000. 46.2 54.1 60.2 65.1 69.2 72.6
160.0 20000. 3000000. 68.1 83.8 97.5 109.3 119.6 128.8

MAXIMUM LOAD IN T I E
WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 200. 1000000. 41.8 53.8 65.8 77.8 89.8 101.7

112.0 200. 3000000. 42.9 55.3 67.7 80.0 92.3 104.6
112.0 2000. 1000000. 59.2 72.4 84.7 96.3 107.7 118.7
112.0 2000. 3000000. 66.1 81.2 95.3 108.7 121.6 134.1
112.0 20000. 1000000. 83.9 104.2 122.4 139.1 154.6 169.0
112.0 20000. 3000000. 89.4 113.2 135.4 156.3 176.1 194.8

128.0 200. 1000000. 42.3 54.2 66.0 77.8 89.5 101.3

128.0 200. 3000000. 44.3 56.7 69.0 81.3 93.6 105.9
128.0 2000. 1000000. 60.2 73.0 84.9 96.2 107.1 117.7
128.0 2000. 3000000. 69.1 84.3 98.4 111.7 124.4 136.7
128.0 20000. 1000000. 92.5 112.4 129.9 145.6 160.1 173.6
128.0 20000. 3000000. 94.1 118.0 140.2 160.9 180.4 198.9

144.0 200. 1000000. 42.4 54.0 65.5 76.9 88.3 99.7

144.0 200. 3000000. 45.3 57.6 69.9 82.1 94.3 106.5
144.0 2000. 1000000. 60.1 72.6 84.1 95.0 105.5 115.7
144.0 2000. 3000000. 70.7 85.8 99.6 112.6 125.0 136.9
144.0 20000. 1000000. 98.0 116.6 132.8 147.3 160.5 172.8
144.0 20000. 3000000. 96.9 120.5 142.2 162.3 181.1 198.9

160.0 200. 1000000. 42.0 53.2 64.3 75.4 86.4 97.4

160.0 200. 3000000. 45.9 58.1 70.3 82.3 94.4 106.4
160.0 2000. 1000000. 59.4 71.4 82.6 93.2 103.4 113.4
160.0 2000. 3000000. 71.3 86.1 99.6 112.2 124.2 135.7
160.0 20000. 1000000. 100.7 117.9 132.6 145.7 157.8 169.1
160.0 20000. 3000000. 104.5 126.3 145.4 162.5 179.2 196.0

FIG. 9—Wall results for studs designed for L/1445—bottom of studs fixed.

MAXIMUM TOP DISPLACEMENT

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 200. 1000000. 0.244 0.308 0.372 0.436 0.500 0.564

112.0 200. 3000000. 0.251 0.318 0,385 0.451 0.517 0.583
112.0 2000. 1000000. 0.044 0.052 0.059 0.065 0.071 0.077
112.0 2000. 3000000. 0.049 0.058 0.066 0.074 0.081 0.088
112.0 20000. 1000000. 0.009 0.011 0,012 0.013 0.014 0.015
112.0 20000. 3000000. 0.010 0.012 0,014 0.015 0.017 0.018

128.0 200. 1000000. 0.243 0,305 0,367 0.429 0.491 0.553

128.0 200. 3000000. 0.255 0.321 0,387 0.452 0.518 0.583
128.0 2000. 1000000. 0.043 0.050 0,057 0.063 0.068 0.074
128.0 2000. 3000000. 0.049 0.058 0.066 0.073 0,080 0.087
128.0 20000. 1000000. 0.008 0.010 0.011 0,012 0,013 0.014
128.0 20000. 3000000. 0.010 0.011 0.013 0,014 0.016 0.017

144.0 200. 1000000. 0.239 0.300 0.360 0,419 0.478 0.538

144.0 200. 3000000. 0.258 0.323 0.388 0,452 0.516 0.580
144.0 2000. 1000000. 0.041 0.048 0.054 0,060 0.066 0.071
144.0 2000. 3000000. 0.048 0.057 0.065 0,072 0.079 0.085
144.0 20000. 1000000. 0.008 0.009 0.010 0.011 0.012 0.013
144.0 20000. 3000000. 0.009 0.011 0.012 0.013 0.015 0.016

160.0 200. 1000000. 0.234 0.292 0.349 0.406 0.463 0.519

160.0 200. 3000000, 0.258 0.323 0.386 0.449 0.512 0.574
160.0 2000. 1000000. 0.039 0.046 0.052 0.057 0,063 0.068
160.0 2000. 3000000. 0.047 0.056 0.063 0.070 0,076 0.082
160.0 20000. 1000000. 0.007 0.008 0.009 0.010 0.011 0.011
160.0 20000. 3000000. 0.009 0.010 0.012 0.013 0.014 0.015

FIG. 9—Continued.

TABLE 2—Stresses for a metal stud wall with a span-deflection ratio of 600 and a
metal tie spring constant of 2000.

Ordinary Analysis, psi Computer Analysis, psi

Studs hinged top and bottom 57.9 to 232.4 46.7 to 193.5

Studs hinged at top and fixed
at bottom 44 to 145.1 26.5 to 103.3

with a metal stud span-deflection ratio of 600. At 8 ft tall, the flexure bond stresses in the brick
veneer approach the "allowable" stresses (37 to 48 psi) when the wind is not over 25 psf.
As shown in Fig. 11, the stress in the brick veneer on earlier typical 8 to 10-ft-tall BV/MS
walls designed with a span-deflection ratio of 360 and with corrugated wall ties with an assumed
spring constant that approaches 500 lb/in. is less than the stress required to crack the wall. In
contrast, by the ordinary analysis the flexure bond stress in some of these walls exceeds cracking
stress levels. This computer analysis demonstrates why the vast majority of the typical BV/MS
walls earlier constructed with span to deflection ratios of 360 and 600 have not cracked and are
apparently successful.
The computer analysis has shown that above 10 to 12 ft the designer has a choice of either
increasing the span-deflection ratio or fixing the base of the metal studs. This would be a matter
of comparative economics.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
112 MASONRY

I m e-J tJ\ r^ c4 n CM
so <7> m r-l VO CO CM CO CO m in (o
\0 •* en 0 0 in - 3 - ON VO m -* o

o \o o^
•* C^J O

(*i
m

m
CO

in
00 r-»

O
O

-3-
CM
^ vo

CM ( O
O

00 r-
rH

vfl -3- ^o •a' in CO 0 0 VO m CO <JN

T-i r-l

m <H o f-4 -3- in CJ (M -3' o to vO CO - 3 - vD (M

CM 00 m n n m ON CM GO o O i-H ON O
o in -3- CM CO vO in CM ON

^o f^ r* o vD O i-t CM -3- c^ ON CO O vO -3" P^

o CO -3- CM CO CO oo ON CM
in n 00 -3- m -3- o m -if i n •J- r-l oo

r-- o
r-l tH
o m

c^
in eg ON

00
m
-* CM CO

ON
o
m
c^

J-
CM

in
ON

vO
O

-3-
in en «n CO 00 »n in -3- ON m o r-

m
O in \D •T CT« 0 0 -a- r* 00 r*. i~(

m
^o VO ON 1^

-3- \o O m
en
r-l
cn
CM in ON
-3- CO
CM
00 i n
s -3- ON VO

O O O O o o o o o o o o o o o o
o o o o o o o o O O O O o o o o
o o o o o o o o o o o o o o o o
o o o o o o o o o o o o Q O O O
o o o o o o o o o o o o
IP o
^
o
r-l
o
CO
o
CO
o
^
o
r-l
o
CO
o
CO
o
W
o
iH
o
CO
o
CO
!

H
W
PQ
Ul
X
M
CO
i/i
U
o::
CU
(0
I
Z
e
tn 1
at^ o o o o o o o o
< tn
5
s:
EE O O O O O O O O
•s.
CM (M CM CM 00 00 00 00 -3' •^ -3 -T o o o o
GH r-l rH CM CM CM <M <r -a- vO vO vO >0

CJ Ul iV,

I
tu mr* fo-S' OcM ONQ o^r^ ONn (Ovo ^ m ?
t/1 • • • • • • • • • • • • • • • • Sr
CU cM(o ooro r-ir- r-co ^ m CMO incM cMtn 5
O r-lOO VOr-l CMON C O - * C M © r - i r * CMr-l COON *
" * * " * '^ ^ ^ ^ ' ^ r ^ , - l c M ^ r ^ r - I C M r H *J

u< r-(o CMON mcM com mcM ONCO O C O COON "Q

U) • . . . . . . . . • . . . . . . -^
p^ -3-00 CNjco coi-i O"-* COO <X T^ OP^ -TON *^
m or^ -3-0 r-iov r^co i-io Ovin CMO •-* t-^ S

bj vom r-ivc r^-T ^rO 0000 j-trn vocM r-l^-. C

CO . . . . . . . * . . . . . . , . (^
OJ -3'r-i mm -3'-* r-tON oco men cor-i in-* jj
o ONr*- cMON oco inr-i r-iON r^-T ^ o ONVO •*
C^ rH ,_| r4 >~i rH r H r H ^ ^ ^ ^ O

[u 0-3 rHvo vOvD oco r^fo co<r P^O OVO I

w
di -3'-3- r».cM .3-vo rnm r-ivo ror* m-3- cor^ ^
m oovo 000 Oir- coo ooo mc>j oov r^.* ^
Ut ooco cMcn vom OON mr-i CMVO r^ON ooov C5
« . . . . . . . . . . . . . . . . a
CL, r-ivo OOO cMr^ ONON Q ^ ONON m-^ p-.r^ ri.
O P ^ m OOP^ OOvO O o o ONp^ C M O ONOO -Tf-J " ^
CM r-l r-l r-l r-l r-l

u. CT\p^ o m rHvo cMm mp-» J-CM p*m ONON

tn ' • • • • • • • . • . . . • . •
(X, p«-vo oovo covo m N vom CMON CMC^ OO-3'
m m-3- vx>m vom^-oop^ P^VD 000 OOP^ I - H O
r-l r-l iH r-l

>. . . . . . , . . . . . . . . . .
OS 00 00 00 00 00 00 00 00
X O Z 00 00 OQ 00 00 00 00 00
a: z o 00 00 00 00 00 00 00 00
<
z
•< w
Cii<
00
0 0
00
0 0
00
0 0
00
0 0
00
0 0
00
0 0
00
0 0
00
0 0
M PU < z 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Q O H r-lt~l COCO r ^ r H COCO f-t r-i COCO CO,H COCO

vi a:
en ou
u (o

il 0

ss
0
N
0

CM
0 0

CO
CN|
00
N
0 0

00
CM
00
01
0 0

-a- -a-
-3- -a-
0 0
-a- - *
•a- •*
0 0

0
vO
0
vO
0 0

0
VO
0
vO
0

MAXIMUM STRESS IN MORTAR

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 500. 1000000. 46.1 55.7 64.3 72.2 79.6 86.5

112.0 500. 3000000. 50.7 62.0 72.3 81.7 90.5 98.8
112.0 1000. 1000000. 51.1 62.4 72.4 81.2 89.3 96.8
112.0 1000. 3000000. 57.7 72.0 84.9 96.6 107.4 117.4
112.0 1500. 1000000. 52,9 65.2 75.9 85.3 94.0 101.8
112.0 1500. 3000000. 60.6 76.4 90.9 104.0 116.1 127.5

128.0 500. 1000000. 59.4 71.3 81.7 91.0 99.8 108.0

128.0 500. 3000000. 68.3 83,6 97.2 109.6 121.2 132.1
128.0 1000. 1000000. 64.0 77,8 89.6 99.9 109.2 117.5
128.0 1000. 3000000. 75.9 94.8 111.7 126.9 140.9 153.8
128.0 1500. 1000000. 65.6 80.2 92.7 103.7 113.4 122.1
128.0 1500. 3000000. 78.5 99.1 117.7 134.6 150.2 164.6

144.0 500. 1000000. 72.2 86.1 98.1 108.8 118.5 127.5

144,0 500. 3000000. 87.0 106.6 123.8 139.5 154.0 167.6
144.0 1000. 1000000. 76.8 92.0 104.9 115.9 125.7 134.6
144.0 1000. 3000000. 95.2 118.4 139.1 157.7 174.7 190.4
144.0 1500. 1000000. 78.2 94.3 107.8 119.2 129.1 138.0
144.0 1500. 3000000. 98.1 123.0 145.3 165.5 183.8 200.9

160.0 500. 1000000. 83.9 98.9 111.4 122.4 132.6 142.1

160.0 500. 3000000. 106.7 130.2 150.8 169.3 186.5 202.4
160.0 1000. 1000000. 88.1 104.5 117.8 129.0 138.8 147.4
160.0 1000. 3000000. 114.8 142.2 166.2 187.5 206.8 224.4
160.0 1500. 1000000. 89.6 106.7 120.7 132.4 142.4 151.1
160.0 1500. 3000000. 117.3 146.3 172.0 194.9 215.6 234.3

MAXIMUM LOAD IN TIE

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 500. 1000000. 66,6 82.2 96,7 110.5 123.9 136,8

112.0 500. 3000000. 69,5 86.2 101,8 116.6 130.9 144,8
112.0 1000. 1000000, 76,0 94.0 110,6 126.1 140.8 154,9
112.0 1000. 3000000, 80,2 100.1 118.6 135.9 152.4 168,1
112.0 1500. 1000000. 81,2 100.9 118.9 135.6 151.5 166,5
112.0 1500. 3000000, 86,3 108.2 128.6 147.7 165.8 183,1

128.0 500. 1000000. 68,6 84.3 98.7 112.4 125,5 138,2

128.0 500. 3000000, 73,1 90.4 106.5 121.7 136,3 150,4
128.0 1000. 1000000. 78,0 96,0 112.3 127.4 141,8 155,4
128.0 1000. 3000000, 84,1 104,7 123.6 141.4 158,2 174,2
128.0 1500. 1000000. 83,3 102,8 120.4 136.7 152,0 166,5
128.0 1500. 3000000. 90,4 113,0 133.8 153.3 171,6 189,1

144.0 500. 1000000. 69,5 84,8 98.9 112.2 124.9 137,2

144.0 500. 3000000, 75,6 93,3 109.7 125.1 139.8 153,9
144.0 1000. 1000000. 78,8 96,2 112,0 126.5 140.2 153.3
144.0 1000. 3000000. 86,9 107.7 126,8 144.5 161.3 177.3
144.0 1500. 1000000. 84,1 102.9 119,9 135.4 150,0 163.8
144.0 1500. 3000000. 93,4 116.1 137,0 156.4 174,7 191.9

160.0 500. 1000000. 69,5 84,3 97,8 110.6 122,7 134.5

160.0 500. 3000000. 77.4 95,2 111,5 126.8 141,5 155.5
160.0 1000, 1000000. 78.5 95,2 110,2 124.1 137,1 149.5
160.0 1000. 3000000. 88.8 109,5 128,4 145.9 162,4 178.0
160.0 1500, 1000000. 83.7 101,6 117,8 132.5 146,3 159.3
160.0 1500. 3000000, 95.4 118,0 138.6 157.6 175,4 192.3
FIG. 11—WaH results for studs designed h/360—bottom of studs pinned. Spring constants from 500 to
1500 by
Copyright lb/in.
ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
114 MASONRY

MAXIMUM TOP DISPLACEMENT

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF AOPSF

112.0 500. 1000000. 0.181 0.217 0.251 0.282 0.312 0,340

112.0 500. 3000000. 0.189 0.228 0.265 0.298 0.331 0,362
112.0 1000. 1000000. 0.111 0.133 0.153 0.171 0.188 0.205
112.0 1000. 3000000. 0.117 0.142 0.164 0.185 0.204 0,223
112.0 1500. 1000000. 0.083 0.100 0.115 0.128 0.141 0,153
112.0 1500. 3000000. 0.088 0.107 0.124 0.139 0.154 0.168

128.0 500. 1000000. 0.180 0.215 0.248 0.278 0.307 0.334

128.0 500. 3000000. 0.191 0.231 0.268 0.302 0.335 0.366
128.0 1000. 1000000. 0.109 0.130 0.149 0,167 0.183 0.198
128.0 1000. 3000000. 0.117 0.142 0.164 0.185 0.204 0.223
128.0 1500. 1000000. 0.081 0.097 0.111 0.124 0.135 0.146
128.0 1500. 3000000. 0.087 0.106 0.123 0.138 0.153 0,166

14^.0 500. 1000000. 0.177 0.211 0.242 0.271 0.298 0,324

lltU.O 500. 3000000. 0.192 0.233 0.269 0,304 0.336 0.367
144.0 1000. 1000000. 0.106 0.126 0.144 0.160 0.176 0.190
144.0 1000. 3000000. 0.116 0.141 0.163 0.184 0.203 0.221
144.0 1500. 1000000. 0.078 0.093 0.106 0.118 0.129 0.140
144.0 1500. 3000000. 0.086 0.105 0.121 0,137 0.151 0.164

160.0 500. 1000000. 0.172 0.205 0.235 0.262 0.288 0.312

160,0 500. 3000000. 0.192 0.232 0.269 0,302 0.334 0.364
160.0 1000. 1000000. 0.102 0.121 0.138 0,154 0.168 0.181
160.0 1000. 3000000. 0.116 0.140 0.161 0,181 0.200 0.217
160.0 1500. 1000000. 0.075 0.089 0.102 0.113 0,123 0.132
160.0 1500. 3000000. 0.085 0.103 0.119 0.134 0,148 0.160

FIG. 11—Continued.

Discussion and Recommendations

From the foregoing analysis it has been shown that if careful attention is paid to the stud
fixity, to the tie stiffness and tie force, and to the wall movement at the top soft joint, it is
possible to reduce flexural bond stresses in BV/MS construction to within tolerable limits.
In retrospect, favorable inherent boundary conditions in as-built BV/MS walls may have
actually reduced flexural bond stresses in the brick veneer below the stress levels predicted by
the ordinary analysis. As typically built, metal studs have some partial fixity to an unknown
degree. It has been shown that the lower the tie spring constant, the lower the flexural bond
stress and total tie force. Possibly the earlier practice of using corrugated wall ties may have had
some beneficial effect. Since the mid 1950s the need for a soft joint under the shelf angle to
accommodate vertical wall movement has been recognized. Until the Clemson tests, its benefi-
cial influence on stresses in the brick veneer was not recognized. Given these three conditions,
as-built BV/MS walls have behaved better than the ordinary analysis would suggest.
Many failures of BV/MS construction known to the authors are frequently due to poor work-
manship practices. A 4-in. wall is most unforgiving to construction installation practices that
increase water penetration. As all 4-in. brick wythe systems will leak to some degree even under
the best of commercial workmanship practice, a proper flashing and weephole system is imper-
ative.
While proper workmanship is necessary to the success of BV/MS construction, the authors
would emphasize an equal need for better engineered designs. The following design recommen-
dations are made to enhance structural performance.

WALL RESULTS FOR STUDS DESIGNED FOR L / 6 0 0 ; BOTTOM FIXED

MAXIMUM STRESS IN MORTAR

WALL H I . SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 500. 1000000. 23.7 29.3 34.8 40.4 45.8 51.3

112.0 500. 3000000. 27.7 34.5 41.1 47.7 54.2 60.7
112.0 1000. 1000000. 24.6 29.5 34.2 38.9 43.4 47.9
112.0 1000. 3000000. 31.4 38.2 44.8 51,1 57.3 63.4
112.0 1500. 1000000. 25.7 30.1 33.9 37.9 41.8 45.6
112.0 1500. 3000000. 33.6 40.7 47.2 53.4 59.4 65.3

128.0 500. 1000000. 28.8 35.3 41.7 48.2 54.6 61.0

128.0 500. 3000000. 36.5 45.2 53.6 62.0 70.3 78.6
128.0 1000. 1000000. 29.9 35.0 39.7 44.0 48.3 52.8
128.0 1000. 3000000. 41.3 49.6 57.2 64.3 71.9 79.3
128.0 1500. 1000000. 30.8 35.5 39.7 43.4 47.0 50.4
128.0 1500. 3000000. 44.0 52.6 60.2 67.2 73.9 80,3

144.0 500. 1000000. 33.4 40.0 46.4 52.7 59.5 66,5

144.0 500. 3000000. 46.2 56.2 65.9 75.3 84.7 94,4
144.0 1000. 1000000. 34.0 38.8 42.9 47.1 51.5 55,8
144.0 1000. 3000000. 50.6 60.3 69.1 77.5 85.7 93.6
144.0 1500. 1000000, 35.3 40.0 43.8 47.0 50.0 52,7
144.0 1500. 3000000, 53.4 62.8 71.0 78.7 86,2 93,2

160.0 500. 1000000. 36.4 42.5 49.1 55.7 62,3 68.7

160.0 500. 3000000. 54.8 66.3 77.3 88.2 98.9 109.4
160.0 1000. 1000000. 37.0 41.5 45.5 49.2 52.7 56.1
160.0 1000. 3000000. 59.4 69.8 78.9 87.4 95.4 103,7
160.0 1500. 1000000. 39.0 43.3 46.7 49.4 51.8 53.9
160.0 1500. 3000000. 62.0 72.2 81.0 88.8 96.1 102.9

MAXIMUM LOAD IN T I E
WALL H T . SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 500. 1000000. 46,7 58.6 70.3 81.9 93.5 105.0

112,0 500. 3000000. 49,4 62.0 74.5 86.7 99.0 111.2
112,0 1000. 1000000. 52.3 64.5 76.2 87.6 98.9 110.0
112,0 1000. 3000000. 56.9 70.4 83.3 95.8 108.1 120.3
112,0 1500. 1000000. 56.3 68.9 80.9 92.4 103.6 114.6
112,0 1500. 3000000. 62.2 76.5 90,0 102.9 115.5 127.8

128.0 500. 1000000. 47.4 59.0 70,3 81.5 92.7 103.7

128.0 500. 3000000. 51.5 64.2 76.6 88.8 100.9 112.9
128.0 1000. 1000000. 53.1 65.0 76.3 87.2 98.0 108.6
128.0 1000. 3000000. 59.6 73.1 86.0 98.4 110.6 122.5
128.0 1500. 1000000. 57.2 69.5 81.0 92.1 102.8 113.4
128.0 1500. 3000000. 65.1 79.5 92.9 105.7 118.1 130.2

144.0 500. 1000000. 47.3 58.5 69.4 80.2 90.8 101.4

144.0 500. 3000000. 52.9 65.5 77.7 89.7 101.6 113.4
144.0 1000, 1000000. 53.1 64.5 75.4 85.9 96.2 106.3
144.0 1000, 3000000. 61.2 74.6 87.2 99.4 111.2 122.8
144.0 1500. 1000000. 57.1 69.0 80.2 90.8 101.2 111.3
144.0 1500. 3000000. 66.7 81.0 94.2 106.7 118.8 130.5

160.0 500. 1000000. 46.8 57.5 68.0 78.3 88.5 98,6

160.0 500. 3000000. 53.7 66.0 78.0 89.7 101.3 112.8
160,0 1000. 1000000. 52.5 63.5 74.0 84.1 94.1 103.9
160,0 1000. 3000000. 61.9 75.0 87.3 99.1 110.6 121.8
160,0 1500. 1000000. 56.5 68.0 78.8 89.1 99.2 109,0
160.0 1500. 3000000, 67.3 81.3 94.2 106.3 118.0 129.3

FIG. \2—Wall results for studs designed for L/1445—bottom of studs fixed. Spring constants from 500
to 1500 lb/in.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
116 MASONRY

MAXIMUM TOP DISPLACEMENT

WALL HT. SPRING E MASONRY 15PSF 20PSF 25PSF 30PSF 35PSF 40PSF

112.0 500. 1000000. 0.117 0.143 0.168 0.193 0.217 0.241

112.0 500. 3000000. 0.125 0.152 0.179 0.206 0.232 0.258
,112.0 1000. 1000000. 0.071 0.085 0.097 0.109 0.121 0.133
112.0 1000. 3000000. 0.078 0.093 0.107 0.121 0.134 0.147
112.0 1500. 1000000. 0.054 0.063 0.072 0.080 0.088 0.096
112.0 1500. 3000000. 0.059 0.070 0.081 0.090 0.100 0.109

128.0 500. 1000000. 0.116 0.141 0.164 0.188 0.211 0.234

128.0 500. 3000000. 0.127 0.154 0.181 0.207 0.233 0.258
128.0 1000. 1000000. 0.070 0.082 0.094 0.106 0.117 0.128
128.0 1000. 3000000. 0.078 0.093 0.107 0.121 0.134 0.147
128.0 1500. 1000000. 0.052 0.061 0.070 0.078 0.085 0.092
128.0 1500. 3000000. 0.059 0.070 0.080 0.090 0.099 0.108

144.0 500. 1000000. 0.113 0.137 0.159 0.182 0.203 0.225

144.0 500. 3000000. 0.128 0.155 0.181 0.206 0.231 0.256
144.0 1000. 1000000. 0.068 0.080 0.091 0.102 0.113 0.123
144.0 1000. 3000000. 0.078 0.093 0.107 0.120 0.132 0.144
144.0 1500. 1000000. 0.050 0.059 0.067 0.075 0.082 0.089
144.0 1500. 3000000. 0.059 0.070 0.079 0.088 0.097 0.105

160.0 500. 1000000. 0.110 0.132 0.154 0.175 0.196 0.216

160.0 500. 3000000. 0.128 0.154 0.179 0.203 0.227 0.251
160.0 1000. 1000000. 0.065 0.077 0.088 0.098 0.108 0.118
160.0 1000. 3000000. 0.077 0.092 0.105 0.117 0.129 0.141
160.0 1500. 1000000. 0.048 0.057 0.064 0.072 0.079 0.085
160.0 1500. 3000000. 0.058 0.068 0.078 0.086 0.095 0.103

FIG. 12—Continued.

1. Stud fixity—Fixing both ends of the stud while having dramatic influence on wall stress is
probably not practical because of the need for the joint at the top of the stud to accommodate
vertical building movements. However, fixing the base should be given serious consideration by
the designer as an alternate to increasing the span-deflection ratio of the metal studs, especially
for walls over 10 ft tall.
2. Engineered ties—Based on the studies made, ideal spring tie constants appear to be be-
tween 200 and 2000 lbs/in. Figure 12 has been prepared for the constant values of 500, 1000,
and 1500 lb/in. As shown in the figure, for metal studs fixed at the bottom to yield an effective
span-deflection ratio of 1445, the stress in the brick veneer in walls up to 13 ft, 4 in. is under the
cracking stress. An ideal tie to embrace the conditions studied is one having a 1000 lb/in. spring
constant with a capacity of 175 lb.
3. Top joint displacement—The maximum top joint displacement is limited by the deforma-
tional limit of the material used to seal the joint. On the assumption that for a Vs-in. joint a
20% shear deformation is tolerable, a maximum top displacement is 0.25 in. Such a 'A in.
displacement will be controlled by the spring constant selected as determined from Fig. 12.

Summary
This paper has shown that excessive flexural bond stresses associated with brick-veneer/
metal-stud walls as determined by ordinary strength of materials analysis are significantly low-
ered by taking into account the effect of the wall top soft joint, the tie flexibility, the shorter stud
length, and the stud fixity. The complexities of the actual wall boundary condition have necessi-
tated the use of a frame network analysis (computer) for solution. The results indicate that for
walls under 10 ft, flexural bond stress levels in the brick veneer can be not only below cracking

limits but also can be within "allowable" values as traditional built with corrugated ties. For
walls between 10 and 12 ft, the authors recommend engineering the tie stiffness or fixing the
stud base or both as design conditions demand. For taller walls less than three stories, fixing
both ends at the stud would permit designs to 20 ft.

References
(/] "Performance Evaluation of Brick Veneer with Steel Stud Backup," Clemson University, Department
of Civil Engineering, Clemson, SC, 23 April 1982.
[2] "National Testing Program, Progress Report No. 1 Small Scale Specimen Testing," Structural Clay
Products Research Foundation, Geneva, IL, October 1964.
[3] "Building Code Requirements for Engineered Brick Masonry," Structural Clay Products Institute,
McLean, VA, August 1969.

Analysis and Test of a Torsional

Sensitive C-Shaped Prefabricated
Brick Spandrel Panel^

REFERENCE: Arnold, R. C , Dondanville, L. J., Krogstad, N. V., and Monk, C. B., Jr., "Anal-
yib and Te*t of a Torsional Sensitive C-Sliaped Prefabricated Brick Spandrel Panel," Masonry:
Materials, Design, Construction, and Maintenance, ASTMSTP992, H. A. Harris, Ed., Ameri-
can Society for Testing and Materials, Philadelphia, 1988, pp. 118-144.

ABSTRACT: Recognizing the inherent torsional flexibility of an open C-section, the authors ana-
lyzed in detail the dead and live load behavior of the section under review. The section consisted of
a sloping sill, a vertical spandrel, and a horizontal soffit configuration fabricated as a continuous
bent plate. The three planes when joined together form an open C-section. To explore the stress
patterns in this shaped plate, the authors modeled the analysis for a hand solution. This was then
compared to a more rigorous solution using a flnite-element computer approach.
The hand approach determined the principal axes of inertia, the shear centers, and the princi-
pal states of stress by Mohr's circle. Comparison is shown with graphic plots of principal stresses
from the finite-element analysis. The latter technique makes a detailed investigation possible,
including plots of both principal tensile and shearing stresses. The modeling of the complex metal
embedments at each reaction location was also possible by the finite-element solution.
Full-scale testing of the panel was made utilizing dead weight from brick loading platforms that
exerted gravity and wind loading on the spandrel panel through a series of compound lever
frames. Contrary to results of brittle torsional failure expected from the analytical work, the ex-
perimental behavior demonstrated a capacity to a load factor of five without cracking or collapse.

KEY WORDS: prefabricated brick panels, reinforced structural masonry, torsional sensitive, C-
shaped cross section, full-scale testing, combined wind and gravity forces, unique compound lever
loading system

Prefabricated brick spandrel panels, about 27 feet long, had been fabricated to replace sets of
three shorter panels of equal total length for a facade replacement program. A single, longer
panel had the advantages of fewer total panels, fewer attachment anchors, and fewer vertical
joints. A design goal was to replace the facade from the exterior as fast as possible with mini-
mum disturbance to the on-going activities of the owner's business. While spandrel panels of
this length are not unique, it is believed that brick panels of this length have been rare. During
the course of erection one panel sustained failure. The mode of failure suggested torsional twist
near one end. When erected, the panels were lifted at their center of gravity as simple beams
near their ends without the benefit of lateral wind connections at their third points when in-
stalled. Such lateral constraints would restrain any twisting moments.
The panel geometry is described in Fig. 1 together with the cross-sectional properties. Inher-

'Engineer II, senior engineer, engineer II, and senior consultant, respectively, Wiss, Janney, Elstner As-
sociates, Inc., Northbrook, IL 60062.
^Metric conversion for units in this paper: in. = 2.54 cm; 1 ft = 30.48 cm; 1 psi = 6894.0 N/m^; 1 pcf =
157.0 N/m'; 1 lb = 0.4536 kg.

118

19.5'

AREA = 264.48 m^
Ix = 75,658.8 m"
ly = 8,918.4 m"
Ixy = 9,199.3 m^

BED JOINT REINFORCING

AT 16' O.C. {*9 WIRE)

-TOP WIND SUPPORT

GRAVITY SUPPORT (ON ENDS ONLY)

lO
eg
lO

7.706'
CENTROID

MASONRY PROPERTIES
MASONRYPRlSMf^- 7300psi
MORTAR COMPRESSIVE f'c>S000psJ
REINFORCING GROUT f'g- 3600psi
SHEAR
CENTER BRICK COMPRESSIVE f'(,= 12,200p8i
CO
oo ABSORPTION: 24 - h SUBMURSION 4.4%
H- • r.. o » 5 - h BOIL 5.3%
IRA - 3.8 grams

-BOTTOM WIND SUPPORT(ON ENDS ONLY)

3.47' 7.46* LONGITUDINAL REINFORCING

(^5 BAR)
FIG. 1—Spandrel cross section showing dimensions, steel, and structural properties.

ently this configuration as an open C-section suggests torsional sensitivity, whereas its relatively
low span-depth ratio suggests flexural strength. It is well recognized that prefabricated panels
are frequently subjected to loading and boundary conditions more severe during handling when
fabricated, transported, and erected than when finally installed. Such loading conditions are
difficult to define rationally. The failed panel's condition was complicated by air voids that
surrounded the longitudinal steel due to fabrication procedures. At the failed end a significant
void was found in the outside, bottom steel rod. In retrospect, it is still unknown what precise
mechanism caused failure. Nevertheless, the failure incident caused a full-scale investigation to
assure adequacy of the panels as fabricated and installed.
A typical panel as fabricated, including representative steel grout void conditions, was se-
lected for test. Generally the voids in question seldom exceeded a brick thickness in length,
although the failed panel did have a void of several feet. Theoretically, the steel bond contact
area that existed despite voids was capable of sustaining the flexural loading. Prior to the full-
scale test described here,fivefull-scale tests were performed using lead weights to load the panel
through its center of gravity. Generally, these tests withstood a vertical load factor of three
without failure. However, these vertical load tests did not simulate to a sufficient degree the
lateral loading torsional moment as required by code. It was decided to carry one panel to fail-
ure, loading both vertically and laterally by simulated gravity and wind forces. Not only was the
lateral wind loading distributed over the surface of the panel but also the gravity loading as well.
Such body loading is in sharp contrast to vertical gravity loading concentrated at the center of
gravity using lead weights only in the prior testing. The latter, while relatively simpler to per-
form, failed to apply the proper ratio of torsional moment to flexural moment as the loading to
failure was increased.

Code Review
The wind loading was based on the 1981 edition of the Building Official Conference of Amer-
ica (BOCA) code as required by the local authority. As will be shown, appropriate normal forces
were applied to the sloping sill, the vertical fascia, and the horizontal soffit as interpreted by the
authors and as verified by consultation with BOCA. With respect to torsion about the shear
center, the torsional moment was increased by a factor of 1.81 when surface loadings were dis-
tributed as compared to concentrated test loads at the center of gravity. For an open C-section
this difference could be significant, at least theoretically.
Two torsional issues are raised by the "Recommended Practice for Engineered Brick Ma-
sonry" of the Brick Institute of America (BIA). The first concerns the recommendation for
shear reinforcement and the second is the calculation of the torsional shearing stresses them-
selves.
As shown in Fig. 1, the panels were originally designed using bed joint reinforcement as hori-
zontal stirrups. BIA under Section 4.8.5.2 expressively omits the use of wire reinforcement for
this purpose. Further, under BIA Section 4.8.5.1(5) all the shear is to be taken by the stirrups
when unreinforced criteria are exceeded. For the case under study, the allowable maximum
unreinforced shear of 50 psi was exceeded, necessitating reliance on the horizontal Vs in.
rounds for possible stirrup steel action providing the maximum shear stress did not exceed 120
psi. It will be shown later that this value is also exceeded. However, the top rod failed the limits
of stirrup spacing by BIA Section 4.8.5.5. Thus, BIA criteria limited the participation of steel
inherently available in the design, namely the use of the bed joint reinforcement and the top
steel bar for resisting shear.
It must be pointed out that the BIA recommended practice is directed primarily at flexural
shear, not torsional shear. For helpful insight one must turn to the American Concrete Institute
code: ACI 318, Section 11.6, dealing with combined flexural shear and torsional shear. First,
ACI 11.6.1.1 limits the effective flange overhang to three times the flange thickness. Unlike

BIA, ACI 11.6.6 does allow the torsional resistance of the masonry material to be utilized (see
ACl equation 11-22). However, calculation showed that, when the torsional resistance of the
masonry was added to that of the steel, the combined strength is less than that of the applied
factored torsional moment. Further, by ACI 11.6.8, stirrup spacing for torsion is very restric-
tive, not to exceed 12 in. Also, in the commentary on ACI 11.6.7.3, it is pointed out that "both
longitudinal and closed transverse reinforcement are required to resist diagonal tension stresses
due to torsion, and if one or the other types of reinforcement are not provided, the other will be
relatively ineffective. The stirrups must be closed, since inclined cracking due to torsion may
appear on all faces of a member." Thus the ACI criteria, which utilizes ultimate strength de-
sign, showed excessive shearing stresses and restricted steel participation to spacings less than
12 in. Further, the closed stirrup requirement would support the use of bed joint reinforcement
but not centerline horizontal steel to avoid face cracking as cautioned by the commentary.
By both BIA and ACI (American Concrete Institute) design codes the torsional shear capacity
was underdesigned. This occasioned the need for a more detailed study.

Strength of Materials Analysis

Initially, the design was reviewed by the ordinary strength of material analysis. The shear
center was determined by using the basic definition as the locus at which loading must be ap-
plied to cause the torsional twist to vanish. Figure 2 shows the neighborhood of the plane of
failure when the panel was lifted as a simple beam over a 24 ft, 2 in. span subjected to its own
dead weight. An analysis was made at the failed section by ordinary strength of materials.
Several features of the analysis are pointed out. The cross section is assumed unreinforced.
This is a very conservative assumption based on the uncertainties of the steel grouting voids.
The calculation of the polar moment of inertia is based on the sum of the values for each thin
rectangular element as known from basic mechanics. No reduction is made for overhanging
flanges as required by ACI. The calculation of torsional shear (88 psi) and flexural tension (37
psi) is made by classic strength of materials formulae. Flexural shear less than 10 psi is ignored.
As the calculated stresses are within 10% of the BIA allowables, it was believed that these con-
siderations were not the cause of the field failure during erection lifting. Using Mohr's circle of
stress, principal tensile stresses of 108 psi were determined. While this value approached the
known modulus of rupture value (110 psi) for the bond of brick prisms, such a value occurs on a
plane significantly inclined to the orientation of the bed and head joints. Given the interlocked
arrangement of mortar and brick, critical inclined principal tensile failure was judged an un-
likely event. It appeared that simple dead load lifting did not cause failure even if unreinforced.
It should be noted, however, that this analysis did not take into account any lateral loading due
to handling forces from fabrication, transportation, and erection. Such events are unknown in
magnitude and position. Therefore a rational design investigation is not possible. However,
based on this analysis, gravity alone is an unlikely cause of failure.
To use the ordinary strength of material theory for the as-built wind and gravity analysis had
some difficulty. The original boundary conditions were so detailed as to cause simple beam
action for any vertical loading but to cause continuous beam action over four supports for any
horizontal loading. At the third point reactions, vertical slots allowed only vertical motion, but
not horizontal. At the ends, both movements were prevented. Originally, top and bottom hori-
zontal restraints were designed for all four reactions: two end reactions and two wind reactions.
However, the location of the bottom inward projecting soffit with shallow clearance beneath the
primary steel spandrel beam prevented field installation of the two bottom horizontal restraints
at the wind reactions. Thus the brick spandrel acted as a bent plate supported horizontally
along the top by four supports positioned at the third points and along the bottom by two sup-
ports at the ends along the bottom edge. Further, the design was complicated by the manner in

FIG. 2—Closeup offailed panel at plane of failure.

which the steel support mechanisms were embedded into the brickwork to avoid point concen-
tration at all reactions.
To handle these as-built support conditions, a finite-element analysis was used.

Finite-Element Analysis
In an attempt to better understand the torsional behavior of the panel, a finite element
method (FEM) analysis of a typical panel was performed. Both handling and "as-built" condi-
tions were considered. Loadings consisted of gravity and wind forces. Analysis of the model was
performed using a proprietary FEM processor based on SAP IV, the Structural Analysis Pro-
gram for Static and Dynamic Response of Linear Systems developed at the University of Cali-
fornia, Berkeley. The problem was executed on a Prime super minicomputer system.

Details of Model
A model of one half of a symmetric panel was constructed using a thin plate and shell quadri-
lateral element to represent the masonry. A total of 1200 elements were used to model the soffit.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
ARNOLD ET AL. ON BRICK SPANDREL PANEL 123

fascia, and sill components of the "C"-shaped panel. Three-dimensional beam elements were
incorporated in the model to represent the steel plate and channel elements embedded in the
masonry at support locations. Beam elements were also used to model portions of the support-
ing framework. Material properties used for the masonry plate elements include: E =
3 000 000 psi, Poisson's ratio = 0.25, and a unit weight of 132 pcf. For the steel members, a
modulus of 29 000 000 psi, Poisson's ratio = 0.30, and a unit weight of 490 pcf were used.
Support conditions for the analysis included two distinct cases. An initial analysis with the
panel supported at lifting points near each end of the panel was performed to determine han-
dling stresses. A second analysis with the panel supported by beam elements representing mem-
bers of the building's structural framework was performed to determine response after erection.
In this "as-built" model, a support providing both vertical and horizontal reactions was pro-
vided near the end of the panel ("gravity support"). A second support providing only a horizon-
tal reaction represents the panels "wind support" near the one-third point of the panel. In the
actual construction, this connection uses a slotted hole to eliminate vertical load transfer.
Loading of the panel for both the handling and erected conditions included gravity loads. In
addition, analysis of the "as-built" model included a uniformly distributed wind load as re-
quired by the BOCA Basic Building Code/1981, which governed the initial design of this build-
ing. After discussion with BOCA officials concerning interpretation of the requirements, the
loadings shown in Fig. 3 were utilized.

y^^yy
1
FIG. 3—Orientation of positive wind loading for analysis and testing.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
124 MASONRY

Results
Results of the analysis described above were reviewed to determine the relative response of the
structure to the dead load and dead load plus/minus wind load. Contour plots of the maximum
principal stress on the faces of the brick panels are attached as Figs. 5A to 7B (stresses in these
plots are in psi). Each group of plots shows either an inside or outside face of the panel. Stresses
printed in large letters are the highest and lowest stresses on the plot and are located near the H
(high) and L (low) printed on the plot. Stresses in the longitudinal direction of the panel for the
dead load only case were also plotted as Figs. 4A and 4B to compare with hand calculations
described earlier.
In all of these plots, high stresses in the immediate location of the support were considered as
locations of critical stress, realizing that the magnitudes of stress directly at the point of support
may be artificially high. No attempt was made to refine the model in this area or to determine
the convergence of the solution. Stresses under consideration were primarily at midspan or a
number of elements away from the support locations.

-104

SLOPE

--J3-.
•q.7B7

ia ifl-
FASCIA

5 9 .[34

SOFFIT

GRAVITY WIND <i_ (SYM)

SUPPORT

FIG. 4—(A) Longitudinal stress distribution on insideface ofpanel due to dead load only. (B) Longitudi-
nal stress distribution on outside face of panel due to dead load only.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
ARNOLD ET AL ON BRICK SPANDREL PANEL 125

85.55 -9S.21

GRAVITY WIND (t (SYM)

SUPPORT

Discussion
Correlation of the model with hand calculations appears relatively good. Longitudinal
stresses shown in Figs. 4A and 4B indicate tension at midspan near the soffit/fascia intersection
of 60 (inside face) to 70 psi (outside face). The "A" figures represent stresses on the inside face
of the panels whereas the "B" figures represent stresses on the outside face of the panels. At the
top of the sloping panel near midspan, compression of approximately 93 psi is indicated on the
outside face. These compare with hand calculations of approximately 73 psi at the extreme
edges of the section when considering simple bending of the panel. Torsional effects can be
noted in the reduction of tension at the free edge of the soffit relative to the fascia/soffit inter-
section and in the stress gradient from the inside to the outside face at the top of the sloping
panel at midspan.
Principal stresses shown on Figs. 5A and 5B for dead load only should be compared to Figs.
6A and 6B for dead load plus wind load and compared to Figs. 7A and 7B for dead load minus
wind load. The differences between the longitudinal and principal stresses are indicative of the
effect of the torsion effect. Significant differences exist in stress levels between the inside face
and the outside face: each must be examined to determine the location of maximum stress.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
126 MASONRY

SLOPE

FASCIA

&a^05^

SOFFIT

GRAVITY WIND <l (SYM)

SUPPORT

FIG. 5—(A) Principal stress distribution on inside face of panel due to dead load only. (B) Principal
stress dbtribution on outside face of panel due to dead load only.

108.S
-42.52

SLOPE

-L-21 . 11

-es^^^
sfl--"^ eflb=
FASCIA

70. 1
Hm
-£a

SOFFIT

GRAVITY WIND q (SYM)

SUPPORT

201 . 8
-102 .5
110
180
170
160
-20

.B(C^^?^. ^ 2 - ^ ^
SOFFIT

GRAVITY WIND (fc (SYM)

SUPPORT

FIG. b—Principal stress distribution on inside face of panel due to dead load plus wind load. (B) Princi-
pal stress distribution on outside face of panel due to dead load plus wind load.

14R.5
- 4 9 . 18
120

^77//. " • ' • " -

->^^-7-
'"'^sej.Ub
U J 1/ / / {T

SOFFIT
7R- --. , - arf"

GRAVITY WIND q (SYM)

SUPPORT

168.S
-64.46

GRAVITY WIND t (SYM)

SUPPORT

FIG. 7—Principal stress distribution on inside face of panel due to dead load minus wind load. (B) Prin-
cipal stress distribution on outside of panel due to dead load minus wind load.

Rq . 86

SLOPE

zmr—?^-B^
FASCIA

72.75
^a L-12.rU

SOFFIT

GRAVITY WIND (g^ (SYM)

SUPPORT

Principal stresses are substantially increased when wind torsion effect is included. In the neigh-
borhood of the gravity support, the maximum principal stresses exceed the BIA limit of 120 psi.
It is this observation plus the concerns discussed in the code review discussion earlier that led to
the decision to perform a full-scale test.

Fall-Scale Test Program

Loading Mechanism
The loading was applied through vertical truss bents located at twelve stations along the total
length of 26 ft, 10 in. Each bent applied load to a ten brick course region. The bents were braced
in pairs to act as a coupled stable structure forming twelve frames, six inside and six outside the
spandrel. See Figs. 8 and 9 for the loading plan layout and the loading bent elevations. The
wind loading was uniformly applied through three sets of compound levers acting normal to the
sill, fascia, and soffit surfaces. The gravity loading was similarly applied acting vertically to the
three surfaces. The loadings were distributed through a surface grillage at all locations except
the vertical gravity force on the fascia, which was necessarily an edge load along its top bound-
ary. Grillage pads in contact with the brick surface were grouted with gypsum to minimize stress
concentrations. The frame fabrication was made out of wood with the levers made out of wood
and steel links connecting to wooden grillage. The wood loading platforms and the com-
pound levers were connected by steel wires. Figures 10 through 14 are illustrative of the loading
mechanism.

Loading Procedure
The loading mechanism was originally designed to be done in the field. Events cause the
program to be done in a laboratory where better instrumentation was possible. The gravity
loading was devised so that the weight of two 5-lb bricks on each of the 72 loading platforms
constituted a loading increment on the spandrel panel surface. The lever system would neces-
sarily follow the specimen as it deflected under load, avoiding any problems of the specimen
creeping away from the loading mechanism. The system was designed to apply a load factor of
six. At a load factor of three, the levers reached their allowable travel; the test was stopped and
the travel clearances increased. The test was continued until a load factor of five was achieved,
at which lever clearances were again exceeded. At the load factor of three and five, the loads
were held for 24 h to measure creep and then were fully unloaded to measure recovery after an
additional 24 h.

Instrumentation
Twenty-eight movement readings were taken on the test specimen itself: two vertical and two
horizontal at each of the following seven locations:
1. Two at the end reaction supports.
2. Two at the wind reaction supports.
3. Two between the reaction and wind supports.
4. One at the panel center.
Six further additional movement readings were taken at the end reactions and wind supports as
follows:
1. Two vertical at the reaction supports.
2. Two horizontal at the reaction supports.
3. Two horizontal at the wind supports.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
ARNOLD ET AL ON BRICK SPANDREL PANEL 133

<
_i 6.
Q-

_ _ 01
Q* a:
oa: > 1 ^
(-2 I
i

FIG. 10—Overall view of panel with loading bents in position.

These 34 movements were monitored by LVDTs (linear variable differential transducers) for
automatic electronic collection and data processing. From these readings midspan deflection
and rotations could be determined relative to the restraints at the supports. It is to be noted
again that only a top horizontal restraint existed at the wind support.
In addition to these movement monitors, the loads at three bents were checked by electrical
strain gauges. Special steel connecting links were machined to mount strain gauges on the links
at the six delivered load locations: the wind and gravity loads on the sill, fascia, and soffit sur-
faces. This required 36 strain-gauge measuring channels. These gauge reading loads were com-
pared to the primary loads delivered from the brick gravity platforms. During the test an auto-
matic computation was made to compare the measured load to the expected load from the brick
gravity platforms. This was read out as an actual load factor. The planned values ranged from
1.0 to 5.0 in 0.25 increments during the second test. Thus 20 readings were taken at each of 70
channels (34 movements plus 36 strain values) for a total of 1400. For the first test, which was
stopped at a load factor of 3.0, 840 readings were taken. Without automatic data acquisition,
such small incremental values would not be practical to do.

Test Results and Analysis

Figure 15 shows the center midspan vertical deflection as a function of the load factor and
compares results from Test No. 1, Test No. 2, and the computer model. Figure 16 shows similar
data for the relative midspan torsional rotation. Also shown are the 24-h creep deflection or
rotations under sustained load and the 24-h deflection or rotational recovery upon load re-
moval. Figures 17 and 18 show the individual vertical deflection readings under the soffit for
each test. It is apparent that the outside edge deflected less than the inside, indicating substan-
tial torsional twist.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
136 MASONRY

FIG. 11—View of typical end reaction frame.

FIG. 12—Loading system as seen from the left end.

FIG. 13—Closeup of grillage and loading fixtures on sloping sill.

FIG. 14—Closeup of bricks being applied to the loading platforms.

0 'lO

z
o
0.35 -
oliJ
_l
u.
tu
o 0.30
z
<
Q.
w 0.25 -
Q

r 0.20 15 HOUR
RECOVERY

0.15 -
<
o
c 0.10 -
liJ
>
tu
o
< 0.05
UJ
>
<
0.00
2 3
LOAD FACTOR
TEST1 TEST2 O COMPUTER MODEL

FIG. 15—Plot of panel midspan deflection versus load factors.

0.0010
0.0000
-0.0010
-0.0020
-0.0030
-0.0040
2 4 HOUR
-0.0050
z -0.0060 RECOVERY
o
I - -0.0070 15 HOUR
< RECOVERY
I - -0.0080
o -0.0090
cr
z -0.0100
<
a. -O.011O
CO -O.012O
Q
-O.013O
-O.014O
-0.0150
-0.0160
-0.0170
-0.0180
-0.0190
-O.02OO
2 3

LOAD FACTOR
TEST1 TEST2 O COMPUTER MODEL

FIG. 16—Plot of panel midspan rotation versus load factors.

z
o
o
UJ

111
•
z
<
Q.
W
O

_J
<
o
OC
UJ
>
lU
>
I-
<
-J
UJ

FIG. 17—Plot of Test 1 vertical deflection versus load factors: front, rear, average.

0.40

0.35
O
UJ

Ui
o
z SUSTAINED
< 0.25
a.
tt>
a
i
J 0.20
<
u
S 0.15
>
UJ
> 0.10
I-
•<
-J
u
K 0.05

0.00 +
0

FIG. 18—Plot of Test 2 vertical deflection versus load factors: front, rear, average.

TABLE 1—Load Test 1: Table of measured load and percentage of theoretical load
by location on loading frame.

Location on Loading Frame

Load Frame
Factor Number A B C D E F

1 1 • 136(117%) * 138(110%) 74(101%) •

2 • 112(97%) » 116(92%) 82(112%) •
3 • 114 (98%) * 111 (88%) 71 (97%) •
2 1 208(102%) 288(124%) 223(99%) 223(88%) 144(99%) 118(98%)
2 230(113%) 218(94%) 234(104%) 249(99%) 149 (102%) 115 (96%)
3 194(95%) 225(97%) 214(95%) 208(83%) 130(89%) 105(88%)
3 1 442(108%) 408(117%) 452(100%) 466 (123%) 209 (97%) 255(106%)
2 456 (112%) 395 (114%) 491 (109%) 414 (110%) 260 (120%) 256 (107%)
3 410 (100%) 358 (103%) 441 (98%) 300 (79%) 188 (87%) 239 (100%)

*No load applied at load Factor 1.

TABLE 2—Load Test 2: Table of measured load and percentage of theoretical load
by location on loading frame.

Location on Loading Frame

Load Frame
Factor Number A B C D
1 1 * 117(101%) « 101(80%) 75(103%) *
2 • 111 (96%) * 94 (75%) 91 (125%) *
3 • 3 3 (28%) * 121 (96%) 64 (88%) *
2 1 190 (93%) 236 (102%) 217 (96%) 225 (89%) 141 (97%) 113 (94%)
2 203 (100%) 224 (97%) 220 (98%) 223 (88%) 210 (144%) 31 (26%)
3 177(87%) 223(96%) 214(95%) 228(90%) 126(86%) 112(93%)
3 1 384 (94%) 354 (102%) 454 (101%) 395 (104%) 207 (96%) 245 (102%)
2 410 (100%) 349 (100%) 472 (105%) 373 (99%) 291 (135%) 151 (63%)
3 379(93%) 337(97%) 436(97%) 332(88%) 184(85%) 246(103%)
4 1 591 (97%) 464 (100%) 703 (104%) 565 (112%) 275 (94%) 313 (87%)
2 636 (104%) 486 (105%) 788 (117%) 672 (133%) 369 (126%) 198 (55%)
3 565(92%) 457(98%) 677(100%) 440(87%) 244(84%) 322(89%)
5 1 824(101%) 762(131%) 742(82%) 531(84%) 352 (96%) 491 (102%)
2 ** 491 (85%) *» •* 474 (130%) 497 (104%)
3 821 (101%) 597 (97%) 917 (102%) 514 (82%) 371 (102%) 143 (30%)

•No load applied at load Factor 1.

**No reading due to debonded strain gages.

Tables 1 and 2 compare the measured loads from the strain gauge readings with the theoreti-
cal values expected.
The deflection/rotational plots (Figs. 15 and 16) show considerable departures from the com-
puter model. This may be understood as showing the effect of the cracked section behavior of
the actual steel reinforced panel over the uncracked, unreinforced behavior as assumed in the
computer model.
Since the span-depth ratio is low, about 5.5, the spandrel panel is expected to be stiff ap-
proaching a thin deep beam. At service loads (load factor of one), the span deflection ratio is
estimated to be more than 4000. This estimate is based on the assumption that the change in
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
144 MASONRY

deflection between a load factor of one to two is the same as from zero to one. Such an assump-
tion is required because the actual deflection of the specimen under its own gravity loading is
necessarily unknown: it is conservative to use the calculation made. Even at a load factor of five,
the vertical midspan deflection was only about 4/lOths of an inch or a span-deflection ratio
of 725.
When similar analysis is made of the rotations at service loads (load factor of one), the rota-
tion is about 0.001 rad. The horizontal movement of the tip of the sloping sill relative to the
juncture of fascia and soffit is estimated to be l/20th of an inch. At a load factor of five, this
movement becomes about 1 in. It is apparent from these results that, relatively, the spandrel
panel displays greater flexibility in torsion than in flexure as is to be expected of a deep, thin,
open C-section. This observation becomes very significant when it is noted that the flexural span
is approximately three times the torsional span, the wind supports being at the third points
where the bottom horizontal tie was never installed.
Despite continuous inspection during loading and a thorough check of all specimen surfaces
after the test, no visible cracking was noted of the brick and mortar materials. The total absence
of cracking at the higher load factors was unexpected.

Investigation Summaiy
From the code review and the finite-element analysis, it was expected that the panel would
fail in excessive principal stresses due to torsion. Even at a load factor of five, no brittle cracking
was observed. The joint action between the steel reinforcing and the brick masonry proved' to
have a greater capacity than either the BIA or ACI Codes predicted. In fact, the ultimate behav-
ior was surprisingly ductile-like. The authors conclude that both code requirements and analyti-
cal results are conservative. While the results proved to show satisfactory behavior for the prob-
lem at hand, it does suggest that code criteria may be more conservative than needed.

Test of Model Masonry Single Pier

Under Dynamic Shaking and Quasistatic
Cyclic Loading

REFERENCE: Chen, H.-L. and Shah, S. P., "Test of Model Masoniy Single Pier Under Dy-
namic Shaking and Quasistatic Cyclic Loading," Masonry: Materials, Design, Construction, and
Maintenance, ASTM STP 992, H. A. Harris, Ed., American Society for Testing and Materials,
Philadelphia. 1988, pp. 145-165.

ABSTRACT: In order to improve seismic design of masonry structures, research is needed in the
area of masonry response to seismic loading. One of the purposes of this program is to investigate
the difference of the structural response of masonry single pier when in-plane shear force is ap-
plied dynamically and slowly. In addition, modeling techniques for brick masonry were studied.
The experimental results are reported for the structural behavior of two brick masonry piers
under dynamic shaking and slowly applied cyclic loading. Each specimen was an unreinforced
model pier having a height-to-width ratio of 1 with fixed-end conditions. The model was con-
structed of one fourth scale brick and mortar composed of fine sand, lime, and cement. Compres-
sive tests of model mortar cubes, prisms, and one square panel were done to understand the mod-
eling techniques of brick masonry. It was found that the piers under dynamic shaking are
characterized by localized failure and abrupt stiffness degradation. However, the peak strength
was lower under dynamic shaking. This tentative, preliminary conclusion should be verified with
further tests.

KEY WORDS: modeling, masonry, bricks, mortar, prisms, single pier, square panel, cyclic load-
ing, dynamic loading, shaking, compressive test, scale factor, shear

Although masonry is one of man's oldest and most common building materials, it has proba-
bly remained the least understood. The structural behavior in particular is not well known, and
very little work has been done in studying the dynamic behavior of masonry structure.
Historically, masonry has shown poor resistance against earthquakes. After an extensive re-
view of the literature, it was concluded that shear walls penetrated by numerous window open-
ings, which were the components of multistory masonry buildings, were most frequently dam-
aged in past earthquakes [4,5,12]. These structural components can be identified as the piers
and the spandrel beams of a shear wall [4].
A testing fixture was designed in order to study the pier behavior under dynamic shaking and
statically applied cyclic lateral loads (Fig. 1). Since experimental programs on large-scale speci-
mens are expensive and difficult to conduct, model masonry piers having a height-to-width ratio
of 1 were chosen. Each specimen was an unreinforced model pier with fixed-end conditions.
One-fourth scale model brick and model mortar composed of fine sand, lime, and cement were
used for model construction.
In this paper the experimental investigation is reported for the structural behavior of model

'Research assistant. Department of Civil Engineering, Northwestern University, Evanston, IL 60201.

^Professor, Department of Civil Engineering, and Director, Center for Concrete and Geomaterials,
Northwestern University, Evanston, IL 60201.

145

Copyright by ASTM
Copyright'^^ 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
146 MASONRY

outer frame
^.JDI

U shaped aluminum channel

lumped mass 1 thick aluminum plate
steel rod \ SPB-16 shaft clamps

ACC (1) n° ix/nrr-f^y ^shaking table

/ / //
FIG. 1—Test setup.

masonry single pier under dynamic sliaking and slow-rate cyclic loading. The variable included
in the program is the rate of loading. The results are presented in the form of hysteresis loops,
envelope curves, and stiffness degradation. Simple analytical models are developed in order to
check the shaking facilities and to understand the behavior of the pier under dynamic shaking.
The modeling technique was checked by comparing the static behavior of the model pier with a
prototjrpe pier.

Objectives and Scope

In order to improve seismic design of masonry structures, research is needed in the area of
masonry response to seismic loading. One of the purposes of this program was to investigate the
difference of the structural response of masonry single piers under d)mamic shaking and slow-
rate cyclic loading. To our knowledge, this difference is not yet well understood, although some
theoretical as well as experimental studies have been undertaken [3,5,7-9,11,13,14].
Modeling techniques for brick masonry have been studied by some researchers [1,10]. It is
recognized that the influences of scale factor and material properties are the main difficulties in
extrapolating the model performance to the actual case. The behavior of approximately
'A-scale specimens was chosen for examination. In order to compare the material properties,
compressive tests of model mortar cubes, prisms, and of a square masonry panel were done. A
more detailed examination of scale effects will be possible in the future using the data obtained
in this research. The scope was limited to model brick masonry units. Other masonry units,
such as model concrete blocks and reinforced masonry assemblages, could form part of the
basis for continued studies in the future.
The applied loadings were induced as horizontally in-plane shear forces with approximately
276 kPa (40 psi) vertically bearing stress at the fixed ends. Dynamic analysis was performed,
based on linear assumption, to calculate the response of the model pier under dynamic shaking
and to check the loading facilities. The calculations compared with the experimental results
showed that the dynamic shaking setup was satisfactory and that the dynamic response of the
model pier can be predicted provided the stiffness remains constant during loading.

Test Specimens

Material Properties
The specimens used to determine the material properties are shown in Fig. 2, and Table 1
shows the mechanical properties of the materials used in the construction of the piers. The
'/4-scale (50.8 by 25.4 by 14.3 mm) model bricks were made by Belden Brick Co. and were
donated by Ramm Brick Inc. for support of this research. The brick was made from shale and
was classified as Grade SW according to ASTM Specification for Facing Brick (Solid Masonry
Units Made from Clay or Shale) (C 216-81). The test results following ASTM Method of Sam-
pling and Testing Brick and Structural Clay Tile (C 67-81) are shown in Table 2A. The area of
the holes is about 12.5% of the gross area.
The mortar was simulated as Type N, that is. Cement: Lime: Sand = 1:1:6 by volume. Port-
land Type I cement and Type S hydrated lime were used. Finer sand was used in constructing
the model mortar. The sand gradation is shown in Table 2B. Compressive strength of 50.8-mm
(2-in.) cubes for model mortar is also listed in Table 1. Three samples of cubes were taken at the
same time when constructing the pier specimens. The sampling, curing, and testing of cubes
followed ASTM Test Method for Compressive Strength of Hydraulic Cement Mortars (Using
2-in. or 50-mm Cube Specimens) (C 109-77).
Six prisms for uniaxial compression tests and one square panel for diagonal tension tests were
constructed. Three of the six prisms were three-stack bond (48.4 by 50.8 by 25.4 mm) prisms.

1/4 s c a l e model b r i c k

jSSSi
square pane

<toc*)

WW.

diagonal t e n s i l e lest
and bliv If p i e r

FIG. 1—Specimens.

TABLE 1—Material properties of model masonry.'

Mortar 3-Stacks 6-Stacks

Type N, Prism, Prism,
Specimen, age 28 Days 28 Days 31 Days Square Panel, 61 Days

Compressive 4.96 47.82 47.23 Ultimate A, cm^ <T„* = 0.734

strength, 5.58 44.81 43.78 load, kN P/-J2A,
MPa 5.37 36.41 40.68 MPa

Average, MPa 5.31 43.03 43.90 8.01 40.0 1.04

Codes ASTM C 270-84' ASTM E 447-84

NOTE: 1 in. = 25.4 mm, 1 lb = 4.448N, 1 ksi = 6.895 MPa, 1 lb/in. = 0.175 N/mm.
'Based on gross area.
'Procht, M. M., 1931.
'ASTM Specifications for Mortar for Unit Masonry (C 270-84).

TABLE 2A—Material properties of model brick.'

Initial Rate Compressive

Saturation of Absorption, Strength,
Absorption Coefficient grams/min. 194 cm^ MPa Code

2.2% (24 h immersion) 0.88 2.1 138.20 ASTM C 216-81,

Grade SW
2.5% (5 h boiling)

"The values listed are the average of three samples.

TABLE 2B—Mortar sand gradation.

Passing %, Passing %,
Sieve No. Model Mortar ASTM C 144"

30 100 40 to 75
50 70 10 to 35
100 20 2 to 15
200 0

"ASTM Specification for Aggregate for Masonry Mortar

(C 144-84).

6-STACK BOND MODEL PRISMS ( h / t - 3 . 8 7 )

I
i8
&5

— MODEL PRISMS
A PROTOTYPE PRISM
[McNary Sc A b r a m s ]

I I I I I 1 P" — I 1 1 1
0.002 0.004 0.006 0.008 0.01 0.012
AXIAL STRAIN

FIG. 3—Stress-strain curves for model prisms.

and the other three were six-stack bond (98.4 by 50.8 by 25.4 mm) prisms. These prisms were
constructed, prepared, and basically tested according to ASTM Test Methods for Compressive
Strength of Masonry Prisms (E 447-84). The thickness of mortar joint was approximately 2.8 +
0.3 mm. All the compression tests were performed on an MTS loading machine at a loading rate
of about 4448 N/min (1000 Ib/min). The stress-strain curves for the three tests of six-stack bond
prisms are shown in Fig. 3. The modulus of elasticity for the model masonry was obtained from
these results. It is noted that the vertical cracks were usually formed before peak load and the
prisms always failed due to the splitting of bricks. The stress-strain curves of the model prisms
are compared with a test result of a prototjrpe prism from Ref 6.
The values of compressive strength of mortar, brick, prism, and square panel of larger scale
specimens tested by other investigators are compared with the model specimens in Table 3. A
direct comparison with the prototype and the model is not possible because of the variations in
mortar strengths, brick strength, and test parameter involved in the prism test and square panel
tests. However, considering these inherent variations, it can be said that the model prism
strength and the model diagonal tension strength are reasonable when compared to the typical
prototype values.
The prisms and the square panel were cured under the same normal atmospheric conditions
as the piers. The ages of the specimens are listed on Table 1.

Model Piers
The model pier specimen is composed of two parts: top and bottom flanges and a square
panel at the center. The weight of the pier was 30.7 N (6.9 lb). The overall dimensions of the
model pier are shown in Fig. 2. The thickness of mortar joint is about 2.8 + 0.3 mm. The
thickness of the pier is 25.4 mm. The piers were constructed vertically by an experienced mason
employing careful leveling techniques. Water was added to the mortar until it became work-
able. Two unreinforced ungrouted model piers were constructed at the same time. One pier was
tested one month later. The other one was tested four months after construction.

Test Equipment and Setup

The shaking table used can generate a horizontal motion of single degree of freedom. The
drive unit is an MB electromagnetic exciter model C5HE. It has a sinusoidal frequency range of
2500 Hz, a maximum force of 1779 N (400 lb), and a maximum displacement of ±12.7 mm
(±0.5 in.). The excitation instrument used to produce the vibration signal is a WAVETEK
Model 185 sweep function generator. The data acquisition system includes an IBM PC and
ISAAC 2000 data acquisition system.
There were two test setups: one for dynamic shaking and the other for slow-rate cyclic loading
(Fig. 1):
1. Dynamic shaking—The bottom flange of pier was bolted to the shaking table. The top
flange was bolted with a 25.4-mm-(l-in.)-thick aluminum plate. Prior to the bolting process,
epoxy was placed on the surfaces between the aluminum plate and the top flange and between
the table and the bottom flange. The 1-in.-thick aluminum plate was bolted with two shaft
clamps. The two clamps were fixed to a 1-in.-diameter steel rod. There was lumped mass on the
steel rod, providing 1423-N (320-lb) vertical load on the pier. The rod passed through two
Thomson Bearing Rollers (SPB-16), which only allow the rod to move freely along the X-direc-
tion (the direction of motion of the shaking table). The rollers were bolted with a U-shaped
aluminum channel at the top, and the channel was fixed to the outer frame. The shaking table
and the outer frame applied fixed-fixed end conditions for the pier.
Figure 1 also shows the instrumentation for measurements. Shaking table displacement was
measured by LVDT(l), and the displacement at the top of the pier was measured by LVDT(2).
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
CHEN AND SHAH ON MODEL MASONRY 151

X ^^
in Z 00
X
Ul
1
f*^ ri
TT
fS in
00

(N
00
0^
O^
1 o
& <s (S
X
s o
d
Ou 1 i
s
«> toj:* o
s
o. e S
•« Si
'S z
s 'K
•1a 2 t^ S, 1/5
*.
a. -; fj' ri
S
o
o C, "^

1
1
CO
U
>J
n
<
H
X

^^ I/)
6

I
It
I ll
X

X
a
t/5

The LVDT(3) measured the diagonal displacement of the specimen. The working range for
LVDT(l) and (2) was ±12.7 mm (0.5 in.) and for LVDT(3) was ±2.54 mm (0.1 in.). The
accelerometer ACC(l) measured the acceleration of the table. The acceleration at the top of the
pier was measured by ACC(2).
2. Slow-rate cyclic loading—There is only one change of the testing setup than that for the
dynamic shaking. Four shaft collars (steel clamps) were used to fix the steel rod against the
bearing rollers. Therefore, the top of the pier was not allowed to move and the outer frame was
working as a reaction frame at this moment. The horizontal in-plane force was applied from the
bottom of the pier by displacing the table along the X-direction.
One load cell with capacity ±4448 N (1000 lb) was placed between the exciter and the shak-
ing table in order to measure the horizontal inputting force. The LVDT(1), (2), and (3) were at
the same places as described in the dynamic shaking, and a LVDT(4) was added for measuring
the vertical displacement of the pier.

Testing Procedure
Two testing procedures are described below:
1. Dynamic shaking test: The five stages (SI to S5) during this test are reported in Table 4.
At each stage, the motion history of the table was prescribed by the function generator. Sinusoi-
dal cycles of motion at a specified displacement amplitude of the table were set through the
entire test. The measured displacement of table was slightly different at different frequencies.
The duration between each stage was about 15 min to 1 h. After each stage, the pier was visually
inspected and the crack pattern was identified and photographed.
2. Slow-rate cyclic loading test: There were 18 stages (Rl to R18) of loading. The pier was
subjected to a series of increasing displacement amplitude controlled in-plane shear forces.
Each stage of loading consisted of two sinusoidal cycles of loading at a specified exciter displace-
ment amplitude. The specified amplitude was gradually increased stage by stage. The full se-
quence of loading was applied at a frequency of 0.01 Hz. Each stage had a duration of 5 to
20 min.
The test was terminated at Stage R18 because of the capacity of the exciter. The shear
strength of the pier at this stage was still in the ascending part. The test of the same pier by
dynamic shaking was continued the following day in order to obtain more information about the
model.
All of the tests were carried out under an essentially constant vertical load of 1423 N (320 lb).
The additional vertical force developed during the cyclic displacing will be discussed in the
presentation of test results.

TABLE 4—Testing procedures.

Cycles Sampling Initial

Loading Frequency, per Rate, Stiffness,
Pier, age Stage Hz Stage points/s N/mm

No. 1,30 days SI 0.5 5 200

S2 1 5 200 1064
S3 7 5 350 1456
S4 3 20 300 780
S5 4 20 350 986
No. 2, 120 days Rl ~ R18 0.01 2 2 834

Test Results

Mode of Failure
The mode of failure under dynamic loading is shown in Fig. 4. Flexural mode of failure was
observed for both piers. This is characterized by horizontal cracks going through the toes of the
pier. Shear slips along the horizontal cracks were observed during the dynamic shaking test and
the permanent deformation of the measured relative displacement was mainly due to the slips.
The pier was considered to fail when the horizontal cracks were completely formed, and further
increase in inertia force only produced more slips. It is noted that the failure under dynamic
shaking is characterized by localized cracking. Such failure has already been reported by other
researchers [16].

Pier No. 1
The time histories of the relative lateral displacement is plotted in Fig. 5. The relative lateral
displacement was computed from the difference between the lateral deflection at the top of the
pier and the table displacement. For the dynamic shaking test, the acceleration at the top pro-
duces the inertia force necessary to displace the specimen. The acceleration measurements are
also shown in Fig. 5.
The average shear stress (the inertia force divided by the net area) versus the relative displace-
ment for Pier No. 1 is plotted in Fig. 6. In Stage SI, the shaking frequency was low (0.5 Hz), and
the pier with its heavy lumped weight (1423 N) was moving as a rigid body. The measured top
acceleration is about zero at Stage SI. Prior to Stage S3 there were no cracks. At Stage S3, a
frequency of 7 Hz was applied and failure occurred. Two main flexural cracks and shear slips of
the specimen were observed. A permanent deflection occurred as seen in Fig. 5, Stage S3. The
LVDTs were zeroed again after Stage S3. At Stage S4, no further damage of the pier was found

FIG. 4—Mode of failure.

STAGE S2 (1.0 Hz)

^^^j^I^^^^''' ^-^•^^"'vjijyi/v^'TABLE ACC. 0.1 g

TOP ACC. ,0.1 g

\0Ui^s^^ ^•'^'^^'4li^'

STAGE S3 (7.0 Hz)

TIME (SEC)

STAGE S4 (3.0 Hz)

- TABLE ACC.
lO.ls

t°-l9

^llWWl/WlWii»'
• RELATIVE DISPL.
jO.lr

-I 1
mm1 1 —1 r

TIME (SEC)

FIG. 5—Time histories of test results.

-
h
z
o
z
o

i
ffl
>
UJ
o
a:
o
li.
N
I

n
\
UJ

«
-

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1
O O g O O O Q O O O O O O Q O O O O O O O O
I I I t l l l t l V T

(Ddl) SSSaiS aV3HS 30Va3AV (Dd>i) ssaais av3Hs aovasAv

•a
e

E •^
O
u Lb
2
ur
O

" ^ " ^ ^ ^ ^ ^r?- ° ^a.

o o
m
o o o o o o o o o N o n o * o lOo «>o No CO
N w 10 •* fo N

(od>l) SS3dlS yv3HS 30Va3AV (t>d>i) ss3y±s av3Hs IOVWBAV

and the hysteretic loop went back to the origin at rest. During Stage S3 loading, the stiffness of
the pier degraded 27% (as indicated by the difference in the observed initial stiffness during
Stages S4 and S2). At Stage SS, the pier had further slips and crushing, and splitting at the top
right and left corners was observed. No cracks appeared in the diagonal direction and the pier
became unstable under shaking. Therefore, tests were stopped after this stage. It is interesting
to note that even after severe damage during Stage S3, Pier No. 1 could still sustain a 3-Hz,
0.22-g shaking at Stage S4 without any further damage. As shown in Fig. 6, the damaged pier
behaved as a rigid body. It vibrated around the origin under the inertia forces and returned
back to the origin at rest.

Pier No. 2
The envelope (the maximum load of each stage) of the hysteresis is shown in Fig. 7. The
individual hysteretic loop of Stage R5, R13, and R18 is also shown in Fig. 7. The average shear
stress is calculated as the applied horizontal force divided by the net area. Each stage contains
two cycles. The pier behaved linearly up to R5. At R13 the nonlinearity was observed because of
cracking. The plots of shear stress versus vertical displacement are shown in Fig. 8. The nonlin-
earity of the vertical displacement shows the opening of the flexural cracks.
The reading of the LVDT(4) (vertical displacement) was recorded when the lumped load was
put on the pier. There was concern that the reaction frame was not rigid enough and the con-
stancy of the vertical load was not maintained. From the measured displacements we found that
the maximum change of the average of vertical force is about —222 N (—50 lb), 15% of total
vertical load), and we noted that the reading of vertical LVDT went back to the same reading
after each stage. Furthermore, the maximum rotation of the reaction frame (Stage R18) is only
0.0075 rad. This influence of the fixed-end condition is negligible. Therefore, the rigidity of the
reaction frame is confirmed and the vertical load can be assumed approximately constant dur-
ing the tests.
The dynamic shaking of Pier No. 2 was started with small increments of acceleration, but no
new crack was found. The complete opening of a new horizontal crack (other than the cracks at
R18) was found after a 4-Hz, 0.25-g excitation. It is noted that under dynamic shaking the
cracks may be very unstable. Even though the increment of acceleration was small, the displace-
ment jump was very noticeable. There was no diagonal crack. The measured vertical displace-
ment was due to the opening of horizontal cracks at the toe.

Discmsion of Test Results

Static Response Comparison -with a Prototype

The experimental results of the model pier under slowly applied cyclic loading is compared
with the test results of a prototype pier, HC6R-11-2 [3]. The specimen HCBR-11-2 is an unrein-
forced partial grouted brick masonry single pier with fixed-ended conditions and height-to-
width ratio of 1. For comparison purposes, the stiffness of the model pier, K„ was multiplied by
a scale factor, F, according to similitude requirements as follows:

Fr = -^ = KJr (1)

where Kp is the stiffness of prototype pier and K^ is the theoretical initial stiffness ratio of
prototjrpe pier to model pier due to difference of dimension and elastic modulus. The calculated
values are listed in Table 5. Since the mortar compressive strength and the bearing stress were
found to be important factors influencing the behavior of the pier [5], we chose these two

PIER N0.2, STATICALLY CYCLIC LOADING

300.0

200.0 - n " R18

0 °
i t p ° R13
in 100,0-
D
R5
^
0.0

V)
tlJ
O-100.0
cfi
a° R13
-200.0 R18

-300.0
-2.0 -1.0 0.0 1.0 2.0
RELATIVE DISPLACEMENT ( m m )
n 2 CYCLES PER STAGE

PIER N0.2, STATICALLY CYCLIC LOADING

300.0

-300.0
-2.0 -1.0 0.0 1.0 2.0 3.0
RELATIVE DISPLACEMENT(1 DIVISION = 1mm)

FIG. 7—Test results under statically cyclic loading (Pier No. 2).

PIER N 0 . 2 , STAGE R18

300.0

-300.0
-0.6 0.6
VERTICAL DISPLACEMENT ( m m )

FIG. 8—The opening of the flexural cracks.

parameters and assumed a proportional relation as a first approximation for the current
comparison,

fr = "MrOBr (2)

where a Mr is the ratio of the compressive strength of mortar between prototype and model and
a Br is the ratio of the initial bearing stress of prototype pier to that of model pier. The compari-
son of load-deflection diagram, stiffness degradation diagram, and equivalent damping ratio
diagram are shown in Fig. 9. The comparison shows an adequate correspondence between pro-
totype and model. However, because the model pier was still loaded on the ascending part, the
comparison is only meant to illustrate the trend of the response. The validity of the model can-
not be completely established until the overall loading behavior is simulated. Moreover, unlike
the HCBR-11-2 specimen, there were no diagonal cracks in the model pier. This may be because
of the relatively lower bearing stress. The model pier experienced about 276 kPa (40 psi) bearing
stress during the test, but the stress on HCBR specimen increased from 730-kPa (106-psi) to
2793-kPa (405-psi) bearing stress when the ultimate load was reached. It is noted that during
the first few loading stages, it is reasonable to assume the bearing stress remained as initial
bearing stress for the HCBR specimen. This assumption was used in the comparison.

Dynamic Response
In order to understand the dynamic response of the model masonry and to check the dynamic
shaking setup, the following analytic models were developed. As shown in Fig. 10, a Single

- a

SOU ir> m

1/1 m
d o -^
I
>>
faj! 00 •»!•

I
m
U
I
'II SO 00

s
n ^ l/J ».
(N ^ O^
< & o
<S

t*)

C5
^1

HYSTERETIC ENVELOPE

0.0
0.0 0.8 1.2 1.6 2.0
RELATIVE DISPLACEMENT (mm)

STIFFNESS DEGRADATION DIAGRAM

0.8 1.2 1.6 2.0

RELATIVE DISPLACEMENT (mm)

DAMPING CHARACTERISTICS DIAGRAM

3.0
2.8
2.6
2.4
2.2
MODEL PIER
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.8 1.2 1.6 2.0
RELATIVE DISPLACEMENT (mm)

FIG. 9—Static results comparison with prototype.

v(t) v(t)
1 M
1
! -^
1
1
K //
/ rW /
/r /
/ '
y'^^
/i^ /i^^ /:ii."'
SDOF
Vg(t)

v(t) v(t)

1 m

/'c K
/ n
/
KT M
1
r
-ja^ mj I
.J
CT
2DOF

FIG. lO—SDOF and 2D0F.

Degree of Freedom (SDOF) model was used for the analysis of the response under dynamic
shaking. The equation of motion is

mv + cv + Kv — —mvgit) = P„ sin cof (3)

For an underdamped system, the steady state response can be shown as

v{t) = — D sin(a)t — <^) (4)

K
where
m — the lumped mass,
c = the damping coefficient,
K = the stiffness,
(J) = the shaking frequency,
Po — the amplitude of shaking,
D = the dynamic magnification factor, and
^ = the phase angle.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
162 MASONRY

Assuming that the system is linear, we let K be equal to the initial stiffness of the pier, damping
ratio equal to 5%, and the loading function equal to the table excitation. These data are shown
in Table 6. It was also noted that the dynamic magnification factor was not influenced signifi-
cantly by the assumed damping ratio at the currently applied loading frequency.
The free vibration response was simulated by a 2 Degree of Freedom (2D0F) system as shown
in Fig. 10. This was necessary since the shaking table did not stop right after the force vibration.
It was assumed that the shaking table had the damping coefficient CT and spring constant KT-
The equations of motion are as follows:

rtirvj + CjVj + KjVj — cv — Kv = 0 (5)

m(v + vr) + cv + A'v = 0 (6)

where mr is the mass of the shaking table. The solution of the equations of motion is of the form

y = A sin(cot — </>) (7)

We end up with the frequency equation of the system as

(-mro)2 + CTOI + KT)(-moi^ + co) + K) - (coi + K) moP- = 0 (8)

The spring constant of the shaking table was measured while the damping ratio was assumed
equal to 5%. Table 6 shows the input data and the calculated frequency of the first mode. Since
the response of the pier is of main interest, we use the measured table acceleration to calculate
the time history of the relative displacement of the pier. The comparisons with the experimental
results are shown in Fig. 11. It was noted that in Stage S2 and Stage S4 the pier behaved linearly
under dynamic shaking and the calculated results are in good agreement with the experimental
results. But in Stage S3, as a result of damage, the stiffness of the pier varied during loading
(Fig. 6). Thus the assumption of a constant stiffness could not predict the measured response.
For further investigation of the dynamic behavior of this stage, a model constructed with a
nonlinear spring is suggested. However, more experiments should be conducted to achieve this
aim.

Difference Between Dynamic Response and Static Response

By comparing the hysteretic behavior of the model pier under dynamic shaking (Fig. 6) with
the behavior of the pier under static loading (Fig. 7), it is seen that the conventional quasistatic
cyclic test results may not be able to represent the dynamic properties of the masonry pier. The
use of an envelope curve constructed from statically cyclic tests may lead to an unconservative
estimation of the ultimate strength of the masonry pier when dynamic response is concerned. By
comparing Stage S3 (Pier No.l) with Stage R18 (Pier No.2), and assuming Pier No.l and Pier
No.2 are identical, it can be noted that the ultimate load of the pier under dynamic shaking was
118 lb (816 N; 7 Hz, 0.37 g). This value is lower than the maximum load (230 lb; 1586 N)
applied statically during Stage R18. When Pier No.2 was subjected to dynamic shaking after
the static test, it failed at a 4 Hz, 0.25 g (80 lb; 552 N) excitation.
It should be noted that a frequency of 7 Hz applied to the model pier during Stage S3 corre-
sponds to a frequency of 7/4 = 1.75 Hz applied to a full-scale pier. Thus it is likely that ma-
sonry structures subjected to dynamic loading during a strong earthquake may exhibit a behav-
ior similar to the model pier tested here. Additional model and full-scale dynamic tests should
be conducted to verify this preliminary tentative conclusion.

o^ o ^
"/•B^
V r-' -q^

O
Q
fM (N (N
l/> lO I/)
z rJ <N r i

s^^

n o> r^
t^ rn 1/^
o -^ <s

c3
vO vO rt
O Ov 0^
^H O N i-H

< O -u;#
Q
(/5

TT .o o
s-
^ I/) 00

o
Z

MEASURED
+ CALCULATED SDOF
o CALCULATED 2D0F

5.2 5.4 5.6 5.8

CALCULATED

SDOF 200F

0.8 1.2
TIME (SEC)

MEASURED
+ CALCULATED SDOF
« CALCULATED 2D0F

-T 1 1 r
6.2 6.6
TIME (SEC)

FIG. 11—Calculation of dynamic response.

Summary
As a preliminary study, the possibility of using the model brick masonry to study the static
behavior as well as dynamic behavior of single pier was explored. The difference of the masonry
response under static loading and dynamic loading was observed, and simple analytical models
were developed in order to check the shaking facilities and to understand the behavior of the
pier under dynamic shaking. The strength of model prisms and square panels were found in
good correlation with the prototype prisms and square panels. The possibility of using model
piers to reproduce the behavior of prototype piers was examined by considering a scale factor.
The model piers under dynamic shaking were characterized by localized flexural failure. How-
ever, the peak strength was lower under dynamic shaking.

Acknowledgment
Acknowledgment is made to C. Ostrander from Masonry Advisory Council for his valuable
suggestions. Ramm Brick Inc. is gratefully acknowledged for providing the model brick, and we
thank J. Gilstrom from Ramm for providing us with very useful information. R. Nudd from
Bricklayers Local 21 of Illinois is gratefully acknowledged for his construction of the model
masonry specimens. The authors also want to acknowledge M. L. Wang (currently at the Uni-
versity of New Mexico) for his helpful suggestions and his contributions to the experimental
setup that he made while he was research associate at Northwestern University.

References
[/] Benjamin, J. R. and Williams, H. A., "The Behavior of One-Story Brick Shear Walls," Proceedings
o/ASCE, Journal of Structural Division, Vol. 84, No. ST4, July 1958.
[2] Brown, R. H., "Prediction of Brick Masonry Prism Strength from Reduced Constraint Brick Tests,"
Masonry: Past and Present, ASTM STP 589, American Society for Testing and Materials, Philadel-
phia, 1975, pp. 171-194.
[3] Chen, S. J., Hidalgo, P. A., Moyes, R. L., Clough, R. W., and McNiven, H. D., "Cyclic Loading
Tests of Masonry Single Piers, Vol. 2—Height to Width Ratio of 1," EERC 78-28, Earthquake Engi-
neering Research Center, Berkeley, CA, December 1978.
[4] Mayes, R. L. and Clough, R. W., "A Literature Survey—Compressive, Tensile, Bond and Shear
Strength of Masonry," EERC 75-15, Earthquake Engineering Research Center, Berkeley, CA, 1975.
[5] Mayes, R. L. and Clough, R. W., "State-of-the-Art in Seismic Shear Strength Masonry—An Evalua-
tion and Review," EERC 75-21, Earthquake Engineering Research Center, Berkeley, CA, 1975.
[6] McNary, W. S. and Abrams, D. P., "Mechanics of Masonry in Compression," Journal of Structural
Engineering, American Society of Civil Engineers, New York, April, 1985.
[ 7] Mengu, Y. and McNiven, H. D., "A Mathematical Model of Masonry for Predicting Its Linear Seis-
mic Response Characteristics," EERC 79-04, Earthquake Engineering Research Center, Berkeley,
CA, 1979.
[8] Priestley, M. T. N. and Bridgeman, D. O., "Seismic Resistance of Brick Masonry Walls," Bulletin of
the New Zealand National Society for Earthquake Engineering, Vol. 7, No. 4, 1974.
[9] Qamaruddin, M., Arya, A. S., and Chandra, B., "Dynamic Response of Multi-Storied Brick Build-
ings," f'artA^uafce^/i^meermg and Structural Dynamics, Vol. 13, March-April 1985, pp. 135-150.
[10] Samarasinghe, W., Page, A. W., and Hendry, A. W., "Behaviorof Brick Masonry Shear Walls," The
Structural Engineer. Vol. 59B, No. 3, September 1981.
[//] Stafford Smith, B., Carter, C , and Choudhury, J. R., "The Diagonal Tensile Strength of Brick-
work," The Structural Engineer, Vol. 48, No. 6, June 1970.
[12] Stephen, R. M., Hollings, J. P., Bouwkamp, J. G., and Jurukovski, D., "Dynamic Properties of an
Eleven Story Masonry Building," EERC 75-20, Earthquake Engineering Research Center, Berkeley,
CA, 1975.
[13] Sucuoglu, H., Mengi, Y., and McNiven, H. D., "A Mathematical Model for the Response of Masonry
Walls to Dynamic Excitations," EERC 82/24, Earthquake Engineering Research Center, Berkeley,
LCA, 1982.
[14] Williams, D. W., "Seismic Behaviorof Reinforced Masonry Shear Walls," Ph.D. thesis. University of
Canterbury, Christchurch, New Zealand, 1971.
[15] Yokel, F. Y. and Fattal, S. G., "Failure Hypothesis for Masonry Shear Walls," Journal of the Struc-
tural Division, Vol. 102, No. ST3, American Society of Civil Engineers, New York, March 1976.
[16] Gulkan, P., Mayes, R. L., and Clough, R. W., "Shakingof Single-Story Masonry Houses", Vol. 1, 2
Copyrightand 3, EERC
by ASTM Int'l79/23-79/25, Earthquake
(all rights reserved); Engineering
Mon May Research
19 10:24:50 Center, Berkeley, CA, 1979.
EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
Construction

Statistical Primer for Brick Masonry

REFERENCE: Grimm, C. T., "StaUstical Primer for Brick Masonry," Masonry: Materials, De-
sign, Construction, and Maintenance, ASTM STP 992, H. A. Harris, Ed., American Society for
Testing and Materials, Philadelphia, 1988, pp. 169-192.

ABSTRACT: Methods of sampling and statistical data reduction applicable to brick masonry are
reviewed. Techniques for plotting histograms and probability density functions for normal and log
normal distributions are discussed. Criteria for establishing sample size and specification require-
ments are enumerated. Binary operations with variables are illustrated. The concept of structural
reliability is introduced. Seventeen log-normal graphs of various brick masonry properties are
included.
KEY WORDS: absorption, brick, strength (compressive) (flexural), cracks, expansion (freezing)
(moisture) (thermal), joint thickness variation, masonry, modulus of elasticity, mortar, sampling,
statistics

Sampling
The purpose of materials testing is to formulate generalizations about the characteristics of
those materials. To determine absolutely the mean strength of a large quantity of material, it
would be necessary to test all of the material, which may be impractical, virtually impossible, or
even self-defeating. Accordingly, some of the material may be tested from which results infer-
ences may be drawn about the strength of all the material. Valid tests of a sample give results
which are certain only for the sample, but those results permit conclusions of varying certitude
about the material from which the sample was taken [11\.
The material from which a sample is taken is called a lot, that is, a quantity of material
which, insofar as is practical, consists of a single type, grade, class, size, finish (texture), and
composition produced by a single source by the same process and under practically the same
conditions. The size of a lot is sometimes standardized; for example, ASTM Method of Sam-
pling and Testing Brick and Structural Clay Tile (C 67-83) [2] establishes 250,000 brick as a lot.
A specimen is an individual piece of material from a lot. A sample is a number of specimens.
Valid conclusions about a lot can be drawn by testing a sample only if the sample is representa-
tive of the lot. For that reason the method of selecting specimens must be unbiased. Bias is
avoided by selecting a random sample, that is, every specimen of the sample has an equal
chance of being selected in every trial. A sample selected haphazardly, without a conscious
plan, is not a random sample. It is a haphazard sample. It is virtually impossible to draw a
sample at random by exercise of human judgement alone. The proper use of an artificial or
mechanical device of selecting a random sample is necessary. Specimens may be selected by
assigning a number to each unit or small group of units in the lot and using a table of random
numbers [7/] or an electronic random number generator to select a number of specimens. For
any given set of conditions there are usually several possible sampling plans, all valid, but differ-
ing in speed, simplicity, and cost.

'Consulting architectural engineer and senior lecturer in architectural engineering. University of Texas at
Austin, Austin, TX.

169
Copyright by ASTM
Copyright'^^ 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
170 MASONRY

The degree of certitude with which conclusions are drawn from test results increases with the
number of randomly selected specimens in the sample. The required number of specimens de-
pends on: (1) the variability in the characteristic to be tested; (2) the permissible difference
between the tested mean value and the mean value of the lot; and (3) the degree of confidence
with which one wishes to state what that difference may be.

Histogram
Having collected a random sample and performed the tests, the first step in data analysis is
often the preparation of a histogram, that is, a plot of the frequency of occurrence versus se-
lected class intervals (i) of the random variable. The number of intervals can be any convenient
number, usually not less than six and increasing with the amount of data. Consider the 25 data
points on the tensile strength of a mortar in psi given in Table 1.
Note in Table 1 that the minimum value is 496 and the maximum if 598. The range (R) is the
difference between the maximum and minimum. If there are 7 class intervals, there are about
R/7 or 15 units per class interval, that is, / = 15.
The frequency distribution is given in Table 2. Note for example that the class interval is from
and including 495 to but not including 510. In Table 2 the number 525 occurs in the 525 to 540
class interval rather than in the 510 to 525 class interval. The histogram is shown in Fig. 1.

Mean and Variance

The sample mean (x) is defined as the sum of the individual data (Lx) divided by the number
of data points (n), x = (Lx)/n. x is an estimate of the lot mean i/i).
The residual is the difference between the sample mean (x) and an individual value (jc), that
is, (x — x).
The sample variance (s^) is the quotient of the sum of squares of the residuals and the number
of data («) less one. s^ is an estimate of the lot variance (o^).

x=n
s^ = (n- 1 ) - '' E
E ((x - x)2 (1)
jc=0

TABLE \—Sample data.

518, 560, 538, 577, 544, 525, 574, 528,

534, 538, 554, 598. 579, 524, 544, 555,
567, 550, 570, 542, 556, 496. 541, 562,
511

TABLE 2—Frequency distribution.

Interval Data Frequency

495 to 510 496 1

510 to 525 511, 518, 524 3
525 to 540 525, 528, 534, 538, 538 5
540 to 555 541, 542, 544, 544, 550, 554 6
555 to 570 555, 556, 560, 562, 567, 570 6
570 to 585 574, 577, 579 3
585 to 600 598 1

X = 547.40
s = 23.54
V= 4.30 %
U n= 25.
Z
UJ i = 15.00
3
o

lO O lO O in o in o in
CM in s. 00 o
lO •*
lO lO in in CO CD

FIG. 1—Histogram.

The sample standard deviation (s) is the square root of the variance (s^). It is an estimate of
the standard deviation of the lot (a).
The sample coefficient of variation (v) is the sample standard deviation divided by the sample
mean, v = s/x or expressed as a percentage, v% = 100 s/x. T5rpical coefficients of variation (v)
for some samples of materials are given in Table 3.
From the data in Table 1, n = 25, x = 547.4, s^ = 554.13, s = 23.54, and v = 4.3%.

Normal Distribution
As the total number of observations approaches infinity and the class interval approaches
zero, the histogram approaches a continuous curve, that is, a frequency distribution curve re-
ferred to as & probability density function. The integration of the equation for that curve be-
tween specified limits is a probability function. Although there are several types of such func-
tions, an axiom of probability theory, the central limit theorem, states: "Under very general
conditions, as the number of variables in the sum becomes large, the distribution of the sum of

TABLE 3—Typical coefficients of variation in some masonry properties, v.

Material See Figure No. Low Normal High

BRICK
24-h cold water absorption 5 <4.9 4.9 to 7.8 >7.8
Saturation coefficient 7 <2.9 2.9 to 5.3 >5.3
Initial rate of absorption, suction 8 <13.1 13.1 to 21.3 >21.3
Compressive strength 10 <7.0 7.0 to 12.0 >12.0

BRICK MASONRY
Compressive strength" 12 <3.4 3.4 to 5.7 >5.7
Mortar head joint thickness 18 <17.0 17.0 to 21.0 >21.0
Mortar bed joint thickness 19 <11.9 11.9 to 15.5 >15.5

'Laboratory specimens.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
172 MASONRY

random variables will approach the normal distribution." The normal distribution function is
also known as the normal probability density function.

f(x) = (27r)-"2 exp - (x - xf/2s^ (2)

where
fix) = the normal probability density function for a sample, that is, the height of an ordinate
erected at a distance x — x from the mean.
It is convenient to express the deviation of x from the mean (jE) in terms of i, so that z —
(x — x)/s. Thus, the density function for the standard normal curve is

/(z) = (27r)-'^2 gxp _ (^2/2) (3)

where
/(z) = the normal probability density function in terms of z.
At the mean, z = 0 and/(z) = (27r)~''^ or 0.3989. In terms of a histogram the computed
frequency is nif{z)/s. Although the data in Fig. 1 are not symmetrically arranged, we will as-
sume by virtue of the central limit theorem that the data would be normally distributed given
more data. Accordingly, for the histogram in Fig. 1 at ;ic = 518, x — x = 518 — 547.4 or
- 2 9 . 4 , z = (x - x)/s or -29.4/23.54 or -1.2489, from Eq 3,/(z) = 0.1829, and the fre-
quency is 25 X 15 X 0.1829/23.54 or 2.91. The computed normal frequency at the midpoint of
each interval in the histogram in Fig. 1 is given in Table 4. With those data the normal curve
may be drawn on the histogram. The accuracy of such a plot increases as the number of class
intervals increases.
The probability that x will fall between xx and x^ can be determined by expressing x in terms
of z and integrating Eq 3 between z\ and zi.

P{z) ^ f(z)dz = (27r)-''2 -<^''2'dz (4)

TABLE 4—Normal curve fit.

Deviation at Interval
Midpoint fromjc

Observed Number of Computed

Intervals Frequency, Interval Residual, Standard Frequency,
(i) Table 2 Data Midpoint X — X Deviations, z Ordinate

(1) (2) (3) (4) = [(1) -1- (2)1/2 (5) = (4) - J (6) = (5)/i (7)

495-510 1 502.5 -44.9 -1.91 1.03

510-525 3 517.5 -29.9 -1.27 2.84
525-540 5 532.5 -14.9 -0.63 5.21
540-555 6 547.5 0.1 -0.004 6.36
555-570 6 562.5 15.1 0.64 5.18
570-585 3 577.5 30.1 1.28 2.80
585-600 1 592.5 45.1 1.92 1.01

where
P(z) = the normal probability function, that is, the area under the normal curve between
— 00 and z (see Table 5).
Equation 4 cannot be integrated in closed form and must be evaluated by numerical integra-
tion. Tabular values of i'(z) between — oo and z may be found in Table 5. The following polyno-
mial approximation may be used for P(z) between — oo and z [/].

P(z) = 1 - 0.5(1 + 0.049867347z

+ 0.0211410061 z^ + 0.0032776263z3

-I- O.OOOOaSOOSez" + 0.0000488906z5

-I- 0.000005383Z)-' + 1.5 X 10"' (5)

Thus, for example, from the data in Table 1 the probability that the tensile strength of the
mortar sample will be greater than 533 psi is determined as follows

z = (jc - x)/s = (533 - 547.4)/23.54 = -0.61

From Eq 5, P(z) = 0.2710 or 1 - P{z) = 0.7290, which compares with 0.2291 + 0.5 or 0.7291
from Table 5. Similarly, the probability that the tensile strength of a specimen lies between plus
or minus one standard deviation of the sample mean, that is, between 523.96 and 571.04 psi, is
0.6826, that is, from Table 5 where z = 1, P(z) = 0.8413 and between z = - 1 and z = 1,
P(z) = 2(0.8413 — 0.5) or 0.6826 (see Fig. 2). The shape of the normal curve is wider and
flatter as the standard deviation increases, that is, as the data become more variable.
It is often necessary to know the value of z required for a given probability, P{.z), in which case
refer to Table 5 or the following polynomial approximation, where P{z) is the portion of the area
under the normal curve between — oo and z [/]. Where P{z) > 0.5, z = F. Where P{z) < 0.5,
z = -F.

F~c + 4.5 X 10-" - (2.515517 + 0.802853c + 0.0103280^)

X (1 + 1.432788c -I- 0.189269c2 -|- 0.002308c3)-i (6)

where
P{z) > 0.5: setc = {ln[l - P(.z)]-^y'\ and
P(z) < 0.5: setc = {\n[P(z)]-^y'\
For example, if tests on representative samples from the national brick production give a
mean compressive strength of 10 437 psi with a standard deviation of 3636 psi, what are the
limits of the 20 percentiles of the sample; that is, 20% of the sample has a compressive strength
lower than what value? What are the strength limits for the next higher 20 percentiles?

P(z) z(Eq6) zs X + zs

0.2 -0.8419 -3061 7376

0.4 -0.2534 -921 9516
0.6 0.2534 921 11358
0.8 0.8419 3061 13495

^ d
d o o o o o o o d o d d d d
li

o -^ r^ G ^ o r^ f^ ' ^ i/) 2> o t* fs

00 -^ ^ O 00 ^
1 o ro r- ^ -^ 00
(/> I/) sO ^O ^ r ^ r ^ r - - « o o o o 00
« of"
o a"^
n 6
d dd d d dd d dd d d dd

O^ i/> •^ r*) 00
o t^ t^ Ô -^ Q uôoa^r^•^ r^ôo^r
(N vO O ' ^ 00 r - - r ^ r ^ o o o o oqooooa«
o lO i/> ^ ^ O
d dddd G <6 <d G G G G G Ci
^

ONvo^o^ors rrt rr -^ -^ in • ^ o r s * - '

i/)i/^^so\o r*-i^t^oqoo OOQOOOON
6 •=* ^ <s S G G ^ d G S ^ Ci ci c> G
3 N

E .o O^^r^ooo ootN'^roCT^ ^o^TTin

O^O'OOvOrO 0OP^<^<N0O fO'^'^'-i
^~> *n O r^ t^ 0 ^ t ^ Q < N lOf^^*-"
l O i o i / i ^ O ' ^ h-r^r^oooo oqoqooa^
G G> c> G <d <i d G G S <:> d <i <>
3

i^ sd
^
lOl/)l/)^«P
• <^ S lO
lO ^ W ^ -J 00 o a* — O <N <N (
'"< l O ^ < ^ P~- •»-; " - J t^
'O^ 'f^
^f^'^
« ^ C^
w " fS
qp O^ lO i
tî w* . - w '
O p OOOOGOi
eo 3
c c
G <S <d c$ d d d d S dd d d d
•3 '5
O (U
o. c
J! O
•a
a
so ^S
O 3
r^
O
u5 00 t-- <s
00 O O 00
S a, « d d <6 <6 <6 d <6 d d d do
S 00 00 ch
«M K|
dddd
I °1
M
u b?
tt] =£
.J
^2: o
S OOTHIOOO
_t^rîO(N
UOTTfSa^M
ootN-^f^^ ôoogo^
•^>^00^

< O'ÔOfS'iO CT^rOÔ^(N ^'ÔQQ

u d i/ii/>i/)Ovo ^r^r^h-oq oooqoqo>
a
<s
d d d d d Gd d d S dddd
>
u
S o 00 <N r^ ^
u S 8 5;::2S § 1 1 1
3 d o o o o o o o o o o o o o o
*C
«>
«
S s i0^ff^^-l/) i-tlOaOOOlO T-'^''^""
•o d
u
a .ooooo o o o o o d d d d
•a
c
«
• * -

w
U
J:
*^
(4 N d d d d d d d d d d «-4-;«
N

a^î/>f*^'or- r-rô^so (N-^'^^\O orûorôo

^^ ^^ ^^ ^^ ^^ ^^ ^^ G^ ^^ ^^ ^^ ^^ ^^ ^^ ^f^ ^^
oooooo ooooo ooooo ooooo

fsrocNO^ • ^ ^ 9 2 ^ ^ i/)^h-oooo ^O'O^^'^'

fiû^ôôr*- ooooooô^ 2'2^2'^2* o > 5 ' ^ ^ ' ^
Os O^ ON O^ ^ O^ ^CT*O^ O^ O^CT*CTÔ^CT*CT^CT*CT^CT^CT*O^
dooddd <6 cS d G ci d <£ ci <:i <6 <S <:6 <6 G (6

fS 00 l/l vO f^ ^
Ov ^ (N ^ ON I/)
(N -^ i/i ^ -O t^
o^ o^ o^ o^ o^ o^
o d o o o d d d d d d d d d d d a o o do

ON ^o I/) go ^ o
r- o ^ o 00 i/> m ^ -^ ^ 00 ^ ^ ON lO ON <N -^ NO r^
- ^ vo r^ h- 00
<N ^ 10 -S ^ r- ~i TT 00 w n 00 O N O N O^
ON ^ ^ ON ON ON X^^ 00
•J 00 5; 2' tf- >^' ^' \^' <^' S O;
^^^
V^' >^i w <
dddddd d
^* d
O^ d
O^ dO^ dO** d d d d d
ddddd

(TV (N TT vO I ^
\o ^ ^ '
. ^ ro
IN . *10w ) I/) ^ 1 r^ 06 bo < .^ ^SP^ > 1^ r^
- . . . » o02'2'2'2'
^CT^O^0^
ON ^ ^ I ^ ON ON I QN ^^ O^ 5 * 5 * QN O^CT^^^CT*ON
ONCT*ON ON ^^CT*ON ^^CT*^^ ON
dddddd ddddd ooooo
ddddd

^ rN<N wj
10 •^^
^ »-^
'^ uu rn gw
00 n 00 uj
uo -^-^ r^
i^ wj 0-1 w> t^ "^ oo CN "^ NO i^
IT) OOONONr-ro
00 O' ON r- ro O N f O r - Q ( N î/l'OtÔO OOON^ÔN
rn^-i/iNor^ r ^ o o o o o N ^ ^ ^ ^ ^ ^ ÔN2<CTSC*
ON Ch O N O N O N ON ON ON ^ O^ ON O N O N O N O N ON ON ON ON ON
d G d G d d G G G G O <0 G G d G G G G G G

'OO'^rNiTftN oo^^^io f^r^oor--fi oo^TrNph-

rnrôoQONOr^ oofotôcN •î/)Norôo ooff^^CT>ON
rsfO'^mNDr-' rôoooÔN OÔNONS'S' ONON5'$>a>
ON ON Cr* ^ O N O N ^ ^ ^ ^ ON O NCT*O N O ^ O N ON ON ON ON ON
d <6 <S <6 (6 d <6 G <6 G G <S G <6 G G O d d G G

ra r - TT (T) NO vp
pNi l o r-- t ^ uo rN( -w . , ^ « . .
(NfOTTi/jNor-- r^ooQOooi
ON ^ ON O^ ON ON ON ON ON ^ O"*
d d d d d d d d d d d d d d d d

r-i/)fo^ONON OONO*^NOO Qi/^NOi/^rs r-^roi/jr-

OTTNONO'V'^ r--pN»\ooN(N î/)Norôo ooÔNa>ON
fsm'vu^NOf tôooooooN a.oN^'S'S' 2*2'2'S'^
ON O N O N O NCT*O N ON ON ON ON ON <^ O NCT*O N ^> î O NCT*ff*O N
dddddd ddddd ddddd ddddd

rJ M (N TT -H PT) <N -H ^ n 00 00 fo I tf ^- r- O ro to t*^

ONn wm
w j i 10
/ j -lO
^ j^
- ' —'-^
' r- (N vO ON ^ r
r~*r^>iju^^^ r^n uuo
^j^^r^OO O O C T ' ^ O N O ^
^ ( *^5 •^^ m m ^ r r^-^ t^r^o00o «
« o 00
oO^N -ffs».*ON^
- 2 . * O N ^ 2 >
^-^ f^ C^ ^ ^ ONCT»^ ON 0> 0^ ON V^ ^ ^ ON ^ ON ^ ^ ^ ^
dddddd ddddd ddddd ddddd

•^lONor^oqoN o^Nf^'^ i/jsOf^^oqoN ©•^raf^-^

- J ^ ^ ^ ^ : ^ (N(NfNr4<N <N<N(N<N(S cnrOf^rorO

f i x ) , HZ)

FIG. 2—Areas under the normal probability curve.

It is concluded that 20% of the national brick production sample has a compressive strength
less than 7376 psi, that 20% is in the range of 7376 to 9516 psi, 20% is in the range of 9516 to
13 495 psi, and 20% have a compressive strength greater than 13 495 psi. It is important to
realize that those conclusions are valid for the sample and not necessarily for the lot. Those
conclusions are valid for the lot only if x = ji and s = a, that is, that the sample has the same
distribution as the lot. That probably is not exactly true.

Sample Size
The distribution of the sample approaches that of the lot as the number of specimens
increases. The deviation of the sample mean from the lot mean expressed as a percentage of the
sample mean is the sampling error, [t = 100 (^ — x)/S]. The number of specimens required for
a sample is a function of: (1) the allowable sampling error; (2) the coefficient of variation of the
sample; and (3) the confidence required that the allowable sampling error will not be ex-
ceeded [3].

rir = {kv/e,y (7)

where
Hr = required sample size to the next higher whole number,
k = a. factor corresponding to a probability that the sampling error will not exceed e„ (see
Table 6),
V — coefficient of variation, and
«„ = allowable sampling error, %.
The probability that the sampling error will not exceed e^ is, of course, set at a low value.
Approximate values of k for various levels of confidence are given in Table 6. An allowable
sampling error (e„) is often set at 5%; for example, the true mean compressive strength of the
brick in the lot may range from 7600 to 8400 psi, when the sample mean strength is 8000 psi.
The use of Eq 7 requires an estimate of v, in which case Table 3 may be helpful.
Suppose we select a sample from a lot of what we believe is good quality-controlled brick to
estimate the 24-h cold water absorption of the lot within 5% of the sample mean with a confi-
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GRIMM ON STATISTICAL PRIMER 177

TABLE 6--k Factors for various probabilities of excessive

sampling error.

Approximate
Probability
Confidence of Excessive Chance of
Level, % Error Excessive Error k

99.9999 0.000001 1 in 1 000 000 4.892

99.999 0.00001 1 in 100 000 4.417
99.99 0.0001 1 in 10 000 3.891
99.9 0.001 1 in 1000 3.291
99.7 0.003 3 in 1000 3.000
99.0 0.01 1 in 100 2.576
95.5 0.045 45 in 1000 2.000
95.0 0.05 5 in 100 1.960
90.0 0.10 linlO 1.645
70.00 0.20 lin5 1.282
80.0 0.30 3 in 10 1.037
60.0 0.40 2 in 5 0.842
50.0 0.50 lin2 0.674
40.0 0.60 3 in 5 0.524
30.00 0.70 3 in 10 0.385
20.0 0.80 4 in 5 0.253
10.0 0.90 linlO 0.127

dence level of 95%; that is, the probability is 0.05 that the sampling error will be exceeded.
From Table 6,fc= 1.96. From Table 3, v = 6% for the coefficient of variation in the 24-h cold
water absorption of good quality-controlled brick. £„ = 5%. Accordingly, from Eq 7, « =
(1.96 X (s/sy or 5.5 or 6 to the next larger whole number.
Suppose further that when we test the 6 brick, we find that x = 20% and v = 10%. Appar-
ently, what we thought was good quality brick is only fair brick by reference to Table 3. We may
now determine the sampling error (e) from a rearrangement of Eq 7.

kvn~ (8)

or
6 = 1.96 X 10(6)-'''2 = 8%

Therefore, the mean 24-h water absorption of the lot might be between 18.4 and 21.6%. Since
that range is intolerable (that is, 8% > 5%), more testing is necessary. We now set v = 10 and
from Eq 7, B = (1.96 X 10/5)^ or 16 to the next higher whole number. Accordingly, we must
test 10 more brick, that is, 16 — 6 = 10.
If a national survey of the compressive strength of brick, consisting of 31 samples, indicates a
mean strength of 10 437 psi with a standard deviation of 3636 psi, we can be 95% confident that
the true national average compressive strength lies between 9157 psi and 11 717 psi, that is, k =
1.96, V = 100 X 3636/10 437 == 34.84%, andn = 31. Therefore, e == 12.26%. Lower value =
O.OlxdOO - «) or 9157, and upper value = 0.01^(100 -I- e) or 11 717.

Specification Versos Requirement

To be reasonably confident that the mean strength of a lot will exceed a specified mean
strength, the required mean strength must be somewhat higher than the specified mean
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
178 MASONRY

strength, that is, if the mean strength of the brick on a project equals the specified mean
strength, half the brick is below the specified mean strength, if the strength is normally distrib-
uted. The required mean strength will be greater for brick having high variance in strength than
for brick having low-strength variance. Also, the required mean strength will be higher if the
mean strength is determined from a small sample than from a large one. The required mean
strength is determined as follows:

Xr = lOOjC.dOO - rVM-'^2)-l (9)

where
Xr = required mean value,
jCj = specified mean value,
V = coefficient of variation of the sample, %
t = factor depending on confidence that jCj < x^ (See Table 7), and
ra = number of specimens in sample.
For example, if you wish to be 90% confident that the mean strength of a lot of brick will
exceed a specified mean compressive strength of 8000 psi, what mean strength would be re-
quired of a sample of eight bricks which had a coefficient of variation of 12%?
From Table 7, t = 1.415 andx, = 100 X 8000(100 - 1.415 X 12(8)-'^2]-i ^jj^ = g s ^ p^j
Similarly, the maximum required mean value can be determined as follows
Xr ™ax = 100S,(100 + tv«-'^2)-i (10)

If one wishes to be 99% confident that the mean 5-h boiling water absorption of a lot of brick
will not exceed 17%, should brick be accepted on which the manufacturer offered test data
indicating a mean 5-h boiling water absorption of 10% with coefficient of variation of 25%
based on a sample of five bricks?

Xr max = 100 X 17[100 + 3.747 X 25(5r'/2]-'

Xr max ^ 1 2 % > 1 0 % (ok)

Under those conditions, brick having a 5-h boiling water absorption of less than 12% should
be accepted.

Maximum and Minimum Properties

Given the mean (jc), the coefficient of variation (v), and the number of specimens tested (n), it
can be stated with a given level of confidence (Pc) that a specified portion of a normally distrib-
uted lot (Pp) will exceed a given minimum value (x/) or will not exceed a given maximum value
M [12].

X, = x[l - (Kv/IQO)] (11)

x* = j?[l + (Kv/lOO)] (12)

where
xi — lower limit of x,
x/, = upper limit of x, and

V = coefficient of variation, %.

Zp = K- z j n - ' + O.SKHn - l ) - ' ] ' ' ^ (13)

where
Zp = number of standard deviations from the mean below which the values of ;c lie;
Pc = confidence probability; decimal and
Zc = the number of standard deviations from the mean below which the confidence probabil-
ity lies.
Procedure:
Given xi or x/,, x, and v, determine K from Eq 11 or 12.
Given Pc, determine z<. from Eq 6 or Table 5.
Given K, Zc, and n, determine Zp from Eq 13.
Given Zp, determine Pp from Eq 5 or Table 5.
For example, a representative sample of nine mortar specimens has a mean sand/cement
ratio of 3.04 with a coefficient of variation of 8.72%. We can be 99% confident what percent of
the mortar has a sand/cement ratio of at least 3 (that is, 3 or greater).
From Eq 6, where P^ = 0.99, z^ = 2.33.
From Eq 11, where x = 3.04, v - 8.72, andxi = 3, K = 0.1509.
From Eq 13, where K = 0.1509, n = 9, and Ze = 2.33, Zp = -0.6307.
From Eq 5, where Zp = -0.6307, Pp = 0.264.

We can be 99% confident that at least 26.4% of the mortar sample has a sand/cement ratio
greater than 3.

Blnaiy Operations
The mean and the standard deviation of the sum, difference, product, or quotient of two
normally distributed, independent variables may be determined as follows [10].

Mx+i- = Mx + h (14)

"x+y — [ol + o'yr' (15)

l^x-y = Mx ~• /*y (16)

Ox-y = [ol + ajr' (17)

/!). = ti/ty (18)

ax, = [(/txa,)2 + (^^oJ^ + (a,a,)2]'^2 (19)

IHx/y) = ^X/MV (20)

a(x/,) = Mr'[(Mxa,)2 + (^lyO:,n'H^i^ + <^r"^ (21)

V^? = [0.5(4^^ - 2aiy'^V'^ (22)

a i f = [fe - 0.5(4^^ - lair^^'^ (23)

r--i/it-i^(N r-'^i/jO'J^ oi/irjr-t---

l O f S ^ O r O 00^^/>^/l•^ OuD'-<I~--tr
â;ooô i^'^f^r)--^ ôc)oâ^
r^' O^* uS V V f*i r^' f*i r<S ro* r*^ fO rrj <S ( S

^i/)i^r-i/> noo^'^'^ «^OTr<s

(NO^TT'O •^0^0^(N^ ^00i/>fSO
ooo>i/)!^r^ •^ovooQor^ t^^vosO'O'O
^^d-^rrori ro(N<N<s<N rsri(NfN<N

O r o (N \ 0 • ^
oooot^r- . _ _ _ . .
r4Trp^<N<N (N<N<N<N(N (N<S<N(N(N

sO <N (S <N (N

00 ^O 00 r*l O
r^ 00 (*! n t^
'^ O 00 so in -^ ^ Tp r^i f^ ro <*) CO f^ ro f^

^ ^ 00 ^ O ^ *0 0^ ro O^ o ro O 00 so
r- ^ r-- ^ fs O 0^ 00 00 r^ r^ t^ r^ so so
r^ OCTvO^ 0> a^ 00 00 00 00 00 00 00 00 00
^ôoo ooooo ooooo
<
CO a^ ^ <r) (N o o^ 00 r- so
(N'ôbîo i/>TrTfTr'^
TT Tf TT -^ ^-i/)i/îoi/>io
^ r*)
n r ^f^
r ofi
* ^f*!
t^>oi/)mi/) i/î/)ioi/)i/>
l/^
o d o o o d o o d d d d d d d

uOO^t^i-il^ i / ) r o < N ^ Q OO>a^0000

(Noor^t^so \0\o^'^« ^ lo IT) tn iO
rO(Nr4(N<N (N<NfSr«»<N P^r4(N<N(N
d dddd ddddd dddd d

•-"rsdîo ôh-ooô i-Hfsr^ûo

o>oooooooo oooqoqr^t^ r^r^r-^r^t^ ^-o^ou^

fSrsrifStN (N(N<N(N<N (N<N(N(N<S <N(N(N(N

IT) in in tn in m m in "^ '^ t^ t-- \0 \0 in <N^l/)(N

(N<N(N<N(N fS(NfNfS<N r4<N(N<N<N *N(N<NfS

Q 0 ^ r o * . 0 O ' ^ 0 ^ ' ^ 0 ^(NQ0iO(N •^

^^og^oo got:^^^ lOi/jttTfrr (N
^^'-HOC;
fS(N<N(N<N
O O O O O
(N(N<N<S<N
p o o o o
(NfSp4<N<N
o 11
(NfS-H»-H

oooa^c^ ooi^m^

r^rooooi/) ro•-ÔâoÔ tO^r^'-HO ^^ÔO^rÎ

r o r ^ r ^ ( N r 4 < N ( S ^ - H W ^ ^ ^ H ^ H ^ O O N Q O O O
flfir'ir'ifl f^c^f^fOfO rofOror'ir's rop^fNr^

^ v^ »,0 1.0 \ 0 " "• " " " " " — '"^ in in -^ -^ •>-> 00 m r^
0 0 0 0 00
o o o o o oo"ooo o o o o o o o o o

lO TT Tf fO r^
rO r o f 5 f*5 f*5 (S(N(N^'-H ^ » - « o o o a>r-^^OTr
1/) LO I/) lO l/l i/>uot/)i/)io in in in ^n in m in in m
o o o o o o o o o o o o o o o o o o o

oor^t^r^r- r - ^ ^ o ^ ^ -sO>o^vO»o lO^-^r^

i/5iOiOiOi/) i/>i/)i/)iOi/l I0i0i0*0»0 ifti/^iOU^
<N(NfN(N<N <N(N<NirS(N *NfS(N<NrS <N(NCS<N
o o o o o o* d o d d o o o o o o o o o

vOr^OOOO Q Q O _

A unit of heat flow through a wall is one btu/h — ft^ — °F. The rate through Material A has a
mean value of 0.13 units with a standard deviation of 0.04 units. The mean rate through Mate-
rial B is 1.13 units with a standard deviation of 0.1 units. What is the mean flow through both
materials combined in parallel and what is the standard deviation of the mean flow? From Eqs
14 and 15

X = 0.13 + 1.13 = 1.26 units

s = (0.04^ + 0.12)1/2 ^ 0.1077 units

The gross area of a hollow masonry unit is 88.64 in.^ with a standard deviation of 4 in.2. The
net area of the unit has a mean value of 45 in.^ with a standard deviation of 2 in.2. What is the
mean cell area of the unit and its standard deviation? From Eqs 16 and 17

X = 88.64 - 45.0 = 43.64 in.^,

i = (42 + 22)''2 = 4.47 in.2

A pier has a mean width of 7.625 in. and a mean thickness of 11.6 in. The standard deviation
in width and_thickness is 0.17 in. What is the mean area of the pier and the standard deviation
in the area? A = 7.625 X 11.6 = 88.45 in.^. From Eq 18, s = [(7.625 X 0.17)2 + ( H Q ^ X
0.17)2 + (o.i7)'']>/2 o „ = 2.2839 in.
If that short pier at failure carries a mean axial load of 30 kips, having a standard deviation of
5 kips, what is the mean unit compressive strength of the pier and what is its standard devia-
tion? S = P/A = 30/88.45 or 0.339 ksi. From Eq 21, s = 88.45"' [(30 X 2.28)2 + (gg 45 x
5)2]i'2[88.452 ^ 2.282]-'/2 oj J = 0.057 ksi. What is the allowable stress, if we wish to be 95%
confident that stress will not exceed strength? From Eq 6 where P(z) = 0.95, z = 1.6458.
Therefore P/A = 0.339 - (1.6458 X 0.D57) or P/A = 0.245 ksi. This is correct only if jc = u.

ReUabU»7
Consider Fig. 3.f(y) is the frequency distribution curve for strength of a structural member,
and/(x) is the frequency distribution curve for the stress in the same member. Wheii stress
equals strength, failure occurs. Reliability (R) is the probability that strength will exceed stress.
R may be determined from Eq 5 or Table 5 where [10]:

z = (jcr- x,)(sl + sj)-"^ (24)

where
Xr = mean strength, psi,
X, = mean stress, psi,
ij = standard deviation in stress, psi, and
Sr = standard deviation in strength, psi.
For example, the mean flexural strength of a particular type of masonry is 135 psi with a
standard deviation of 33 psi. The flexural stress in the masonry is 25 psi with a standard devia-
tion of 9 psi. What is the reliability of the system? z = (135 - 25)(332 + 92)-''2 or 2 = 3.22.
From Table 5 or Eq 5, /? — 0.9994. There is one chance in about 1667 that failure will occur.
What is an acceptable level of probability of being killed by structural failure? The appropri-
ate value of R varies with the type of material, load, and structure.

FIG. 3—Applied stress versus strength.

Log-Nonnal Distribution
The frequency distribution is often not symmetrically arranged but is skewed to the right or
left. When that happens the mean, mode, and median are not coincident as they are in normally
distributed data. The mode is the value of the observation that occurs most frequently. It is at
the peak value in a frequency distribution. The median in a set of observations is the middle
observation when the observations are arranged in order of magnitude, that is, half the data are
greater than and less than the median.
Many types of data are not symmetrically (normally) distributed. For example, if the mean
price of a house is $100,000, a house selected at random is much more likely to cost $200,000
that it is to cost zero. In such cases when the coefficient of variation exceeds 30%, it is often
helpful to use the log-normal distribution, that is, the natural logarithm of the variable is nor-
mally distributed. The log-normal density function is determined as follows in which v is ex-
pressed as a decimal:

f(x) = (2ir)-<'-5(;8jc)-i e x p [ - (Injc - a)^!^] (25)

a ^ ln[x(l + v2)-»-5] (26)

V = s/x (27)

/3 = [ln(l + v^)]"-5 (28)

where
z = (Inx - a)0~^:

f{z) = (2ir)-'>-5/3-' e x p [ - (zV2) - 0z - a] (29)

y = nifiz) (30)

where
y — the histogram ordinate at any value of jc,
n = number of data in the histogram.

I = histogram interval, and

V — coefficient of variation, decimal.
For a given probability that x will not be exceeded, values of 2 are determined from Eq 6 or
Table 8.
A national survey of 30 samples of brick were tested for initial rate of absorption (suction),
grams of H2O per 30 in.^/min. The mean value was 23.7 with a standard deviation of 31.3. The
test data were distributed as indicated in Table 9.
From those data, x = 23.7, s = 31.3, v = 31.3/23.7 or 1.32 or 132%, a. - l.bbOl, and |3 =
1.00473. This histogram in Fig. 9 is drawn with intervals of 6 for 31 samples (that is, / = 6 and

TABLE &—Probability that x will not

be exceeded.

Probability, % z

1 -2.3267
5 -1.645
10 -1.28
20 -0.842
25 -0.674
30 -0.524
33.3 -0.4317
40 -0.253
Mode -0
SO (median) 0
Mean 0/2
60 0.253
66.7 0.4317
70 0.524
75 0.674
80 0.842
90 1.28
95 1.645
99 2.3267
99.5 2.575
99.9 3.10

TABLE 9—Log-normal fit of brick suction.

Computed
Interval Observed Interval Frequency
(0 Frequency Midpoint (Ordinate)

(1) (2) (3) (4) (5)

0to6 7 3 7.3
6 to 12 5 9 7.4
12 to 18 6 15 4.9
18 to 24 4 21 3.3
24 to 30 1 27 2.2
30 to 36 3 33 1.6
36 to 42 0 39 1.2
42 to 48 2 45 0.9
48 to 54 2 51 0.7

X = 7.15 % LOW< 4 9 %
s = 4.13 % NORMAL = 4.9 - 7.8 %
LOG NORMAL V = 58 % HIGH > 7 8 *
n = 31.
i = 2.14 %
MODE = 4.65%
MEDIAN = 6.20 %

i
24 hr. COLD WATER ABSORPTION, %

FIG. 4—Twenty-four-hour cold water absorption of brick [4].

X = 12.35 % GOOD < 6.7 %

s = 10.65% FAIR= 6.7-13.0
V = 86. % POOR > 13.0%
n = 31.
i = 5.54%
MODE = 5.36 %
MEDIAN = 9.35 %

COEFFICIENT OF VARIATION, *

FIG. 5—Coefficient of variation of 24-h cold water absorption of brick [4].

n = 31). Since z = (ln:c — a)i8"', x = exp(/3z + a). For example where P(z) = 80%, from
Table 8, z = 0.842, from Eq 29,/(z) = 0.3938(1.00473)"' exp[-(0.842V2) - (1.00473 X
0.842) - 2.6607] or/(z) = 0.008325, and from Eq 30, y = 0.008325 X 31 X 6 or3; = 1.55 and
X = exp [(0.842 X 1.00473) + 2.6607] or x = 33.3. The ordinates to the distribution curve are
recorded in Table 9.

Application
The principles of statistics as discussed above have been applied to some published test data
on masonry. Histograms and log-normal probability density functions have been plotted in
Figs. 4 thru 21. In those figures, "high" values are in the upper 33 percentile, low values are in
the lower 33 percentile, and normal values are in the middle third of the distribution.

X = 9.3% LOW < 6.6%

s = 5.0% NORMAL = 6.6 - 10.0 %
V = 53.6 % HIGH >10.0%
n = 31.
i = 3.5%
14 MODE = 6.4%

12 MEDIAN = 8.2%
>-
u
10 / ^
7 8
3
2 6
4
2
i ^ LOG NORMAL

y §^^ 1^^""^'- L_
1 r •—t—

5hr. BOILING WATER ABSORPTION, %

FIG. 6—Five-hour boiling water absorption of brick [4].

GOOD < 2.9%

FAIR =2.9-5.3%
POOR > 5.3%

COEFFICIENT OF VARIATION , X

FIG. 7—Coefficient of variation in the mean saturation coefficient of brick [4].

x = 19.6% GOOD < 13.1 %

s = 12.0 % fiMR=13.1 -21.3%
V = 61. % POOR > 21.3%
n = 31.
i = 5.6 %
MODE = 12.2 %
MEDIAN = 16.7 %

COEFFICIENT OF VARIATION, »

FIG. 8—Coefficient of variation in the mean initial rate of absorption of brick [4).

x= 23,74 LOW< 9.31

s= 31.27 NORMAL = 9.31 22.13
V = 132. % HIGH>2213

-LOG NORMAL n = 31.

i = 6.00
MODE = 5.25
MEDIAN = 14.36

o
UJ

GRAMS/30 SQ. IN.- MIN.

FIG. 9—Initial rate of absorption of brick [4].

GOOD<7 %

FAIR = 7 - 1 2 ^

POOR > 12 %

COEFFICIENT OF VARIATION, %

FIG. 10—Coefficient of variation in brick compressive strength [4].

s = 31.32 NORMAL = 124 -150

V= 22.42% HIGH > 150

= 12.80
MODE = 129.76
MEDIAN = 136.29

I
• * c \ j o < c c p ^ e M O a ) < p
f^' o w in 00 ,- Tf i< oj cj
^ C 0 5 l 0 t D ( 0 0 5 O r ; ( r )

FLEXURAL STRENGTH, psi.

FIG. 1 i—Flexural strength of brick masonry wall panels Type S mortar, inspected workmanship, psi [9].
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
188 MASONRY

GOOD < 3.4 %

LOG NORMAL
FAIR = 3.4-5.7 ^
POOR > 5.7 %

COEFFICIENT OF VARIATION, »

FIG. \2—Coefficient of variation in compressive strength of laboratory-prepared brick masonry

prisms [4].
LOW< 0.39
NORMAL= 0.39-0.51
HIGH > 0.51

RATIO OF STRESS AT FIRST CRACK

TO STRENGTH

FIG. 13—Ratio of stress at first crack to ultimate strength of brick masonry in compression [5].

LOW < 61.21

NORMAL = 61.21 -11741
HIGH > 117.41

STRAIN, in./in. X 10"^ AT FIRST CRACK

FIG. 14—Brick masonry prisms; longitudinal strain at first crack: three brick strengths with three mor-
tar types [5].
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GRIMM ON STATISTICAL PRIMER 189

)c= 640.1 LOW < 506.6

s = 250.3 NORMAL = 506.6 - 701.6
V = 39.1 % HIGH > 701.6
n = 294.

11
80 i = 133.4
MODE = 517.1

60 MEDIAN = 596.1
1,

i
20
1
J 1
\
LOG NORMAL

CO p ^ CD eg 00 Cy
^, ifi CO ir, ifi •^ •.•,-'lo O
-vb '-•./
cN iri u4
oi
cvjinroWtncDcytooofijinco

RATIO OF MODULUS OF ELASTICITY

TO COMPRESSIVE STRENGTH

FIG. IS—Ratio of modulus of elasticity to compressive strength of brick masonry prisms, dimension-
less [6].

x= 0.0176% LOW< 0.0097%

s = 0.0149% NORMAL= 0.0097-0.0184%
V = 85. % HIGH > 00184 %
n = 62.
LOG NORMAL i = 0.0063 %
MODE = O0078 %
MEDIAN = O0134 %

FREEZING EXPANSION %

FIG. 16—Freezing expansion of brick [7].

LOW< 2.7
NORMAL = 2 . 7 - 3 . e
HIGH > 3.68

COEFFICIENT OF THERMAL EXPANSION

10"'in./in,- 'F

FIG. n—Coefficient of thermal expansion of brick (10 ' in./in. °F) [7].

X= 0.0237 LOW < 0.0152

s= 0.0158 NORMAL= 0.0152-0.0256
V = 66.7 % HIGH>0.0256
n= 27.
i = 0.0056
LOG NORMAL MODE = 0.0137
MEDIAN = 0.0190

U
Z

LENGTH CHANGE, »

FIG. \&—Brick moisture expansion at 17 months [13].

GOOD<17%
FAIR =17-21.%
POOR > 21%

LOG NORMAL

COEFFICIENT OF VARIATION, %

FIG. 19—Coefficient of variation in head joint thickness in brick masonry [8].

X = 14.3% GOOD < 11.9%

s= 4.6% FAIR = 11.9-15 5%
LOG NORMAL V = 32. % POOR >15.5%
n = 24.
i = 3.8%
MODE = 12.4%
MEDIAN = 13.6%

i'
COEFFICIENT OF VARIATION, %

FIG. 20—Coefficient of variation in bed joint thickness in brick masonry [8].

X = 0 2 3 in. GOOD < O.ISin.

s = 0.14 in. FAIR = 0.15-0.25
LOG NORMAL y v = 62.8% POOR > 0.25 in.

8r / n=24.
i = 0.069 in

6 MODE = 0.14 in.

1
V MEDIANJ = n19 in
>-
u 4
z
UJ
13
\\v
2
i
2

0 J
'A T- (M CM S- n ^ in
c o o b Y^ d d d
in

DE ^wTION, in.

FIG. 1\—Deviation of masonry head joint from vertical in 6-ft height of brick wall [8].

References
[/] Abromourtz, M. and Stegum, I. A., Handbook of Mathematical Functions, NBS Applied Science
Series 55, Superintendent of Documents, U.S. Government Printing Office, Washington, DC, Nov.
1964.
[2] ASTM Method of Sampling and Testing Briclt and Structural Clay Tile (C 67-84), 1985 Annual Book
of ASTM Standards, ASTM, Philadelphia, PA, 1985, pp. 49-58.
[3] ASTM Recommended Practice for Choice of Sample Size to Estimate the Average Quality of a Lot or
Process (E122-72), 1985 Annual Book ofASTM Standards, American Society for Testing and Materi-
als, Philadelphia, PA, 1972.
[4] National Testing Program, Progress Report No. 1, Small Scale Specimen Testing, Brick Institute of
America, Hemdon, VA, Oct. 1964.
[5] Grimm, C. T. and Fok, C.-P., "Brick Masonry Compressive Strength at First Crack," Masonry Inter-
national, University of Edinburgh, Edinburgh, Scotland, Vol. 1, No. 2, July 1984, pp. 18-23.
[6] Grimm, C. T., "Elastic Modulus of Clay Brick Masonry," Proceedings of International Symposium
on Reinforced and Prestressed Masonry, University of Edinburgh, Scotland, 26-28 Aug. 1984,
pp. 225-252.
[7] Grimm, C. T., "Probabilistic Design of Expansion Joints in Brick Cladding," Proceedings: 4th Cana-
dian Masonry Symposium, University of New Brunswick, Freducton, NB, Canada, June 1986,
pp. 553-568.
[8] Grimm, C. T., "Some Brick Masonry Workmanship Statistics," Journal of the Construction Division,
American Society of Civil Engineers, New York, NY, March 1988.
[9] Gross, J. G., Dickers, R. D. andGrogan, J. C , Recommend Practice for Engineered Brick Masonry,
Brick Institute of America, Reston, VA, Nov. 1969, pp. 257-262.
1/0] Haugen, E. B., Probabilistic Approaches to Design, John Wiley & Sons, Inc., New York, NY, 1968,
p. 123.
[//] Natrella, M. G., Experimental Statistics, NBS Handbook 91, Superintendent of Documents, U.S.
Government Printing Office, Washington, DC, 1966.
[12] Hald, A., Statistical Theory with Engineering Applications, John Wiley & Sons, New York, NY, 1965,
pp. 303-316.
[13] Young, J. E. and Brownell, W. E., "Moisture Expansion of Clay Products," Journal of The American
Ceramic Society, Columbus, OH, 1 Dec. 1959, Vol. 42, No. 12, pp. 571-581.

An Investigation of an Extended Plastic

Life Mortar

REFERENCE: Matthys, J. H., Houston, J. T., and Dehghani, A., "An Invettlgatloii of an Ex-
tended Plastic Lite Mortar," Masonry: Materials, Design, Construction, and Maintenance,
ASTM STP 992, H. A. Harris, Ed., American Society for Testing and Materials, Philadelphia,
1988, pp. 193-219.

ABSTRACT; The use of an extended plastic life mortar which can be stored for a prolonged
period of time without losing its desirable properties can be beneficial to the masonry industry in
several ways.
This paper presents the findings of a two-phase study related to an extended plastic life mortar
(ready mix mortar) using a single brand of set-reducing additive. Phase 1 was optimization of a
final mix in terms of setting characteristics and compressive strength. Mixes were made which
examined the variables of dosage rates, cement content, lime content, and fly ash content. Each
mix was characterized by compressive strength, air content, water retention, and plasticity. Sev-
eral mix designs were generated for achieving particular strength and setting characteristics.
Phase 2 compared one of the generated extended plastic life mortars to a standard Portland
cement lime Type N mortar as defined in ASTM Specification for Mortar for Unit Masonry (C
270) with respect to mechanical properties. The properties examined included: compressive
strength, bond strength, freeze thaw, corrosion, shrinkage, efflorescence, and compressive
strength and modulus of elasticity of masonry prisms.

KEY WORDS: mortar, ready mixed, retarded strength, freeze-thaw resistance, corrosion,
shrinkage, efflorescence

The commercialization of extended plastic life (EPL) mortars as a ready-mixed product avail-
able to the masonry industry is a relatively new concept in the United States. At this time several
chemical admixtures have been developed and are being promoted for use in producing ready-
mixed EPL mortars. The marketing of such products in the United States has developed quite
rapidly and has been spurred on by such factors as prior usage in non-U.S. construction mar-
kets and desires of inventors, entrepreneurs, and major companies to be first in the market
place for obvious reasons. The incentive of the potential user is to achieve a promised or implied
net savings over the conventional job-mixed mortars.
Agencies and individuals who participate in the development of specifications and standards
for the use of construction materials in the United States are traditionally conservative in their
approach to that task, and that stance is certainly proper in these writers' points of view. Public
safety and long-term performance issues as well as the more common and obvious short-term
physical properties of new products must be adequately assessed prior to public use.
In the case of EPL mortars the authors have found little published research data from U.S.
sources; foreign sources were not reviewed. In addition, there is currently no recognized U.S.
specification or standards specifically applicable to the evaluation and use of EPL mortars. The

'Professor, Civil Engineering Dept., University of Texas at Arlington, Arlington, TX 76013.

^Consulting engineer, Constructive Engineering, 1711 S. Parkway Blvd., Mesquite, TX 75149.
^Student, Civil Engineering Dept., University of Texas at Arlington, Arlington, TX 76013.

193

current, generally referenced U.S. specification for mortar for unit masonry is ASTM C 270,
and EPL mortars do not comply with regard to duration of use provisions and in some cases may
exceed the air content limitations. Consequently, there is a definite need for both thorough
research and specification development to address the subject of EPL mortars for use in ma-
sonry construction in the United States. The market testing for such products has produced
sufficient results to motivate both the technology and marketing interests in the masonry indus-
try to work together in producing a cost-effective product with dependable short-term and long-
term performance properties.
There are at least six basic parameter study areas related to the proper evaluation of construc-
tion materials composed of filler aggregates bound together in a matrix containing portland
cement and other related binder components.
1. Physical and chemical properties, variability, and compatibility of each binder compo-
nent.
2. Compatibility of the binder components with the aggregate filler of the product matrix.
3. The effects of proportioning the binder components upon the performance of both the
plastic and hardened phase of the product matrix.
4. The influence of the batching and mixing cycle and plastic phase history upon short-term
and long-term performance of the product matrix.
5. The influence of environmental factors, especially temperature, upon the plastic phase
and in-place performance of the product matrix.
6. The long-term performance of the product matrix with emphasis on chemical and physical
stability in those environmental conditions where the product will actually be used.
Other areas of needed study for new products relate to considerations of constructability, that
is, the influence of construction variables upon the performance of the in-place product. The
emphasis of this study was to evaluate selected materials parameters only.

Phase 1: Early Screening Tests of EPL Mortars

Research on EPL mortars conducted at the Construction Research Center at The University
of Texas at Arlington has been selectively reported in this paper. One experimental chemical
admixture proposed for use in producing EPL mortar for unit masonry construction was evalu-
ated with respect to selected performance criteria. Preliminary testing using a wide variety of
controlled test parameters was conducted to provide an opportunity to develop testing proce-
dures and to become familiar with the general characteristics of EPL mortars (see Figs. 1-8).
This screening work, termed Phase 1, provided an opportunity to make several general observa-
tions concerning data trends for the parameters studied. This work is considered very prelimi-
nary by the authors and will not be reported in detail since the process of optimizing testing
procedures was not deemed to be complete and only a minimum of replicate testing was com-
pleted at this writing date. However, the authors have made a number of significant observa-
tions regarding EPL mortar performance which may be helpful in the planning of future-related
research by these authors and others. These observations and data presented here are not
known to be characteristic of commercially available EPL mortars now in use in the U.S. nor
are such relationships implied by these authors. As previously stated, the research discussed in
this paper is limited to one experimental admixture used to produce a variety of EPL mortar
mixes under laboratory-controlled conditions. A summary of general observations made to date
is as follows:
1. The mix proportions and binder types had a significant effect upon the mortar setting
characteristics and upon the retempering procedures required to maintain plasticity beyond
about 4 h and up to 72 h of age. (A 5-min mixing cycle was standardized for all work done to
date.)
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
MATTHYS ET AL ON EXTENDED LIFE MORTAR 195

FIG. 1—Mortar batch mixing.

FIG. 2—Flow test.

FIG. 3—Water retention test.

FIG. 4—Penetrometer test.

FIG. 5—Masonry unit suction effect on mortar.

FIG. 6—Two-inch by two-inch cube compression specimen.

FIG. 7—Containerized EL mortar.

FIG. S—Compression cube specimens.

2. Early screening tests suggest that EPL mortars may be significantly subject to the effects
of specimen size, shape, and curing regimen.
3. Strength retrogression from 7 to 28-day air-cured tests was observed in a significant num-
ber of EPL mortar mixes, especially when the mixtures were kept plastic for 24 h or more. It
should be noted that compressive strength retrogression did not occur in any moist-cured speci-
mens.
4. Almost all EPL mortar mixes had air contents exceeding 12%, with many exceeding 20%.
5. For all combinations of binders (cement only, cement plus fly ash, and cement plus lime)
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
MATTHYS ET AL. ON EXTENDED LIFE MORTAR 199

and the use of EPL admixtures in varying dosages, a severe reduction in compressive strength
was noted when comparing cubes made immediately after initial mixing and those made after
maintaining mortar plasticity for 24 to 72 h. Strength reductions typically ranged from 50 to
90+% for 28-day tests.
6. Increasing cement content and varying EPL admixture dosage were only marginally suc-
cessful in improving strengths of mortars aged 24 h or more.
7. Increasing fly ash content from 20 to 50% of cementitious material weight produced a
significant reduction in strengths of cubes molded immediately after initial mixing, but much
less effect was seen in strengths of EPL mortars kept plastic for 24 to 72 h.
8. In general, it is felt that EPL mortar testing is quite sensitive to testing procedure, both in
procedure design and operator technique. The influence of the binder systems employed, in-
cluding the chemical admixtures coupled with the uncharacteristically long set time extension
for Portland cement-based binders, has resulted in a hardened mortar product which would be
expected to be "testing techniques sensitive." As a result, the authors recommend that those
responsible for test and standards development should give due consideration to the potential of
this product to display significant sensitivity to field production and field construction variables
as well as laboratory test procedures.
Tables 1 and 2 have been included in this paper for general reference only in order to illus-
trate the relative magnitude of cube strengths for selected variations in mortar mix proportions.

Phase 2: Comparisons of Extended Plastic Life Mortar to a Conventional Type N Mortar

Based on Phase 1 study, several final types of mortar mixes were generated based on setting
times and 5.08-cm cube compression strength values. These mixes, when quantified on com-
pression strength requirements (initial mixing specimens only), would be classified as Type N,
Type S, or Type M mortars. With regard to relative mix economy, these EPL mortars varied
from 12% less to 11% more expensive than a standard Type N proportion specification port-
land cement lime mortar.
Since the majority of all mortars used in masonry construction would probably be Type N, it
was felt that a comparison of some general physical properties between a Type N EPL mortar
and a Type N ASTM C 270 portland cement lime mortar would be of interest. It was also de-
cided to examine a few properties of assemblages constructed with these mortars. Mortar prop-

TABLE \—Mortar bate!i properties'• (mixes with cement and sand omly).

= C.F., Cement, Water, Sand, EPL Water

Mix No. kg/m^ kg kg kg Admixture % Air Retention Flow W/C

EPM5A 0.98 2.67 1.70 14.20 170 25.8 85 127 0.63

EPM5B 0.98 2.67 1.52 14.20 283 27.9 86 127 0.57
EPM5C 0.98 2.67 1.50 14.20 425 25.6 84 134 0.56
EPM6A 1.21 3.31 1.67 14.20 170 23.8 86 129 0.51
EPM6B 1.21 3.31 1.40 14.20 283 27.6 91 129 0.42
EPM6C 1.21 3.31 1.0 14.20 425 25.2 80 129 0.30
EPM7A 1.44 3.93 1.73 14.20 170 22.9 86 133 0.44
EPM7B 1.44 3.93 1.69 14.20 283 25.3 90 130 0.43
EPM7C 1.44 3.93 1.0 14.20 425 22.6 73 134 0.25

"Batch weights (kg saturated surface dry (SSD)] flow = % (ASTM Specification for Flow Table for Use
in Tests of Hydraulic Content C 230-80); air = % by volumetric method (ASTM Test Method for Air
Content of Hydraulic Cement Mortar C 185-80); EPL admixture = dosage rate (g/45.3 kg total cementi-
tious material); water retention = % (ASTM Methods for Physical Testing of Quicklime, Hydrated Lime,
and Limestone C 110); W/C = water cement ratio (by weight).
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
200 MASONRY

00 -^ r^ 00 ON n t-- I/) t o TT I/) i/>

o ^ - ^ f^ ^ o •-« m • ^ TT a - t-'
ro ID -^ « ON r - ^ Tj-
f*5 O^ 1^ ' ^ ^ 'I* O ^
vO <N —

f^ t^ o in
IPS
-^ r - TT lO
^ TT lO I/) m ro 00 r^
X 00 O ^ Tf ON 00 ^
00 ^ ^
00 n

iTi <Z> rr CT^ O^ O '-H f^ r^ o «/)

s
-_ 00 ^ ro
ON fO ^ ^
1 ^ O^ <S i/>
- ^ t ^ TJ- (T)
(N <N fO !-^
o\ vo r^ rr

^c r^ TT u^ O ^O O 00 • uo f s ^
[^ ^ ro O I 00 *-H ^£
m ID f^ O
00 r*> f ) " ^

lO a^ n f*! 00 - ^ ^ lO n i-H f l 00
^ TT Q r^
-H ON (N O
fo fo r-^ to O (^ O - ^
^O 0 0 ^ l/>
^ - t ^ lO to
00 ^ ^
c

0^ f j "^ ^
^O O f ^ TT
^ <N m

r^ (^ ^o ^ sO O^ O 00 1/^ ^H 00 a>
•^ rn O^ yf
00 lO ^ r*)
^ 00 rs vo ^ r^ LO 00
t ^ 00 o • »
CM ^ r^ -H 1/5 O
- ^ T H (N
^ m <N •-"

2* ^ -^ r- r o r o r-- t ^ <N (N TT t~-

ON 4/) f O CO rr -^ m m - ^ O^ i/> f^ o
t~» r^ -^ m m 0 0 I/) ^ CS ID 00 r o

a
•a
X! j = J3 J : J : j :
u
J: J : .s £ J : f
O - ^ 0 0 <S
<N r r l ~
ss
H3
•-«
c o
05 .1 05 .1
•a a>
< " < S 2
OL.
a
CL Q
ON ^ B
o
z

erties evaluated included freeze-thaw resistance, corrosion performance, shrinkage behavior,

and efflorescence characteristics. Assemblage properties examined included compressive
strength, compression stiffness, flexural bond strength, and assemblage shrinkage. The tests
involving extended plastic life mortar used specimens constructed using three different suspen-
sion time intervals: initial mixing, 6 to 7-h suspension, 24-h suspension except as noted.

Materials
Masonry Units—Two types of clay brick were selected for construction of assemblages as
shown in Fig. 9. The net cross-sectional area was approximately 78% of the gross area, classify-
ing them as solid units. Initial rate of absorption tests and compression tests were conducted on
the units according to ASTM Method of Sampling and Testing Brick and Structural Clay Tile
(C 67). Test results and unit dimensions are given in Table 3.
Type I brick was a medium compressive strength and medium suction unit. Type II brick was
a high compressive strength and low suction unit.

79-0.25"
0.25"

( a ) Typ« I Brick

2.0" -1.5"

(X) c:)
.50"

-7.50"
(b) Type H Brick

FIG. 9—Masonry units.

TABLE 3—Physical properties of masonry units.

Net Gross Initial Rate Compressive

Height, Width, Length, Area, Area, of Absorption, Strength,
Type cm cm cm cm' cm' g/min/193 cm' kPa

I" 5.71 8.89 19.05 132 169 12.3 72 983

II 5.41 8.89 19.05 134 169 2.32 145 794

"Average of five tests.

Mortar

The conventional Type N portland cement lime mortar used was based on the proportion
specification. Water was added to produce a flow of 130 ± 5%. All conventional mortar was
used within a 2V2-h period and exhibited average 28-day moist-cured cube compressive
strength of 7949 kPa.

Assemblage Tests
Flexural Bond Test—Two stack bonded prisms five units high, fully bedded, with concave
tooled mortar joints were built by a qualified mason (Fig. 10). Prisms were air cured in the
laboratory environment for testing at 3, 7, and 28 days using the bond wrench. Figures 11 and
12 show the flexural bond strength with age for prisms constructed with Type I and II units,
respectively.
For the medium absorption (Type I brick) prisms, the extended plastic life mortar specimens
for initial mixing had 3-day bond strengths comparable to that of the conventional Type N
mortar, a 40 and 50% higher strength after 7 and 28 days, respectively. The 7-h suspension
mortar specimens had a 50% lower 3-day strength but about the same 7 and 28-day strength as

-7.5"

<z:>CD .^^So^
Joint No. I

12.75"
3/8"* / Joint No 7
\
1 3 . 2 5"
Joint Mo 3

Joint No t

( a ) Tjpe I B r i c k ( b ) Type n Brick

FIG. 10—Prism specimens for bond and compression test.

I
•H
Z
^
t
rt
O
r
u
9 rt o ka
a ID b a
b •M 3 0
0 J^ 0 S
£ •^ s
e •*
Z w r . «N

« La* U> U•
& (0 O 03
><4J jiJ 4J
E-
b b U
^u X0 ZO Z0
s
•3
e
a i-:i h^ i 4
f
to
^g^ S

(S 4 a O

(TSd) n^Susjas axTsusx lejnxaxi

I
r

• <

•S

(Tsd) U38uaJ3S STXSuai t«-tnxaxi

conventional T3rpe N mortar. The retempered 24-h extended mortar suspension specimens ex-
hibited 3-day, 7-day, and 28-day strengths about 60% to 70% less than that for standard Type
N mortar.
For the low absorption (Type II brick) prisms both mortars exhibited about 7% lower flex-
ural strength compared to the medium absorption specimens. The comparison of the two mor-
tars at 3, 7, and 28 days are basically the same as for the Type I brick specimens.

Compression Prisms
As for the flexural bond study, two stack bonded prisms were constructed for examining
compression strength of masonry assemblages for the two brick types at 3, 7, and 28 day ages.
The specimens were capped and tested according to ASTM Test Methods for Compressive
Strength of Masonry Prisms (E 447) (see Figs. 13 and 14). All specimens were air cured in the
laboratory.

FIG. 14—Compression prism testing.

Test results for medium absorption (Type I brick) prisms are given in Fig. 15. Three-day
strength and 28-day strength for initial mixing for extended plastic life mortar was equivalent to
the Type N standard. The prisms for 7-h and 24-h suspension periods, exhibiting lower strength
at all ages, gave 28-day strength at 20% and 40% less than those of standard mortar, respec-
tively.
Test results for the low absorption (Type II brick) prisms are given in Fig. 16 and exhibited a
lower overall strength as compared to Type I brick. With the exception of the 7-h suspension
period strength at 28 days, the values are basically the same as for Type I brick.
Modulus of Elasticity in Compression—To determine the initial modulus of elasticity in com-
pression, the two brick types were used as in the flexural bond and compression test to build
stack bonded prisms using standard Type N portland cement lime mortar and the extended
plastic life mortar at 0, 7, and 24-h suspension periods. For each test two prisms were con-
structed and tested after 28 days of lab curing. To measure strains, a transducer was placed at
the center line on both sides of the prism's wide face with a 20.32-cm gage length. Test data
curves are shown in Figs. 17 and 18. Modulus values are tabulated in Table 4. The prisms with
the higher strength units were stiffer.
The extended plastic life mortar selected for comparison purposes was the most economical
of the four final mixes generated in Phase 1. It contained only portland cement, admixture, and
sand. This mortar showed good workability and, according to the mason, was easy to work with.
It remained quite plastic for 24 h while maintaining workability. Without retempering, this mix
hardened in 35 h. The EPL mortar initial flow was in the order of 125%. Flows at 6-h suspen-
sion were about 105%. Flows at 24-h suspension were about 55%. All suspended mortar was
retempered to approximately the same initial flow prior to construction of specimens. The mor-
tar set time in assemblages was approximately 30 min for the high IRA brick (Type I) and V/i h
for low IRA (initial rate of absorption) brick (Type II). The mortar had an average air content of
21.3% and water retentivity value of 78.6%. The average 28-day moist-cured strength of mortar
was 6338, 7239, and 4343 kPa for 5.08-cm cubes made initially, 6 h, and 24 h after initial
mixing, respectively. The drop in mortar strength for 24-h suspension was associated with re-
tempering.
It should be noted that although both mortars were mixed to approximately a 130% flow, the
extended plastic life mortar showed excellent workability at flows as low as 80%. The color of
cubes made from the extended plastic mortar was slightly darker than the standard mortar
cubes. However, in assemblages, both mortars approached a similar shade of color after a few
days.
Masonry Assemblage Shrinkage—To examine the shrinkage of mortar when used in masonry
assemblages, brass gage plugs were epoxyed to two faces of the 28-day compression test stacked
bonded specimens. The gage length was then measured at time intervals of 1, 4, 7, 14, 21, and
28 days using a dial micrometer conforming to ASTM C 246. The length changes were then

TABLE 4—Tabulated values of modulus of elasticity.

Modulus of Elasticity,
Avg of Two Samples, kPa

Mortar Mix Type I Brick Type II Brick

Type N 9 653 000 14 755 300

EPL mortar
Initial mix 11 514 650 14 410 550
After 7 h 8 274 000 13 721 050
After 24 h 393 000 468 860

I
41
00

(xsd) qa3u9J3s aA-:ss9jdni03

9
I
a
V
s

I
a
a

(isd) q33u3j:is SAissajdmo^

(isd) ssaaas

O
O

o
o

m ^
o £;
o 1
d >'^ E? N ^

^u S (•9

s •a
•w J.
0,

d
o
o u
e 1
e u
•«4
>»
0
•s.
PI
o
o 1

O
d

(S
1
001
O ^^
d i

o
3

^ O c w
^
-^ =
^'^
w
lA un Its
•T o jn
• - *

m CN ^:M S
<•
(TS d ) sssai S

calculated using the initial one-day observation length as the reference. The test data is shown
in Figs. 19 and 20. Specimens were air cured in the laboratory.
As shown, the Type I brick prisms made with initial mixed extended plastic life mortar had
less shrinkage than that of Type N portland cement lime mortar. The 7-h suspended mortar had
the least shrinkage while the 24-h suspended mortar had the most. The results were basically
the same for the Type II brick prisms, except that the amount of shrinkage was about 50%
higher. The low absorption brick masonry assemblages exhibited similar shrinkage to that of
the mortar shrinkage specimens. The high absorption brick masonry assemblages exhibited a
smaller shrinkage than that of the mortar shrinkage specimens.
Freeze Thaw Resistance—To examine the resistance of the mortars to alternate freezing and
thawing, 5.08-cm mortar cubes were made for the conventional Type N portland cement lime
mortar and the comparison extended plastic life mortar. (Here only initially mixed cubes were
evaluated.) Cubes were air cured for 28 days and their dry weights recorded. The samples were
then subjected to alternate freezing and thawing cycles in accordance with ASTM C 67.
After nine cycles of freezing and thawing, it appeared the conventional standard mortar had
lost a considerable amount of material from the portion of the cube in contact with water (see
Fig. 21). The extended plastic life mortar specimens did not appear to have lost any weight. All
cubes were dried and weighed. The data are given in Table 5.
The conventional mortar cubes, having exceeded the 3% weight limit of ASTM C 67, were
considered to have failed. The extended plastic life mortar cubes had not lost any weight; there-
fore, the freezing and thawing procedure was continued. At the end of 50 cycles, these cubes
were still undamaged (see Fig. 22).
Efflorescence Test—A decision was made to examine the conventional and extended plastic
life mortar with regard to efflorescence. Standard 5.08-cm cubes were made for conventional
Type N mortar and the extended plastic life mortar for 0 and 24-h suspension times. For each
mix, six cubes were made, air cured in laboratory for 28 days, and divided into two three-cube
sets. At the end of the curing period they were dried and tested according to ASTM C 67. One
set of each specimen group was placed in the drying room, and the second set was placed in
distilled water for seven days. Both sets were dried in a drying oven for 24 h and examined for
changes due to efflorescence.
The results showed that the cubes made from standard Tjrpe N and the extended plastic
mortar (0 suspension time and 24-h suspension time) had effloresced on the cube portion in the
water. Figure 23 shows the extended life mortar cubes; Fig. 24 shows the standard Type N
mortar cubes.
Shrinkage Characteristics—To compare the shrinkage characteristic of the mortars, two
mortar specimens for Type N standard mortar and for the extended plastic life mortar (0, 7-h,
and 24-h suspension times) were cast and tested according to ASTM Test Method for Length
Change of Hardened Cement Mortar and Concrete (C 157) and ASTM Specification for Appa-
ratus for Use in Measurement of Length Change of Hardened Cement Paste, Mortar, and Con-
crete (C 490). The specimens were mortar prisms 2.54 cm^ and approximately 28.57 cm in
length. Figures 25 and 26 show the specimens along with the reference bar. The samples were
cured in a moist room for 48 h, then cured in lime-saturated water until they reached an age of
28 days. An initial observation of length was made after seven days of curing, this length to serve
as the reference length. The second observation of length was made at the end of the 28-day
curing period. The specimens were then stored in the laboratory environment with more length
observations being made after total air storage periods of 4, 7,14, and 28 days, 8 weeks, and 16
weeks. The results are shown in Fig. 27.
After 28 days of moist curing there was a slight increase in length of extended plastic life
mortar samples for 0 and 7-h suspension times. The 24-h extended mortar and standard Type N
mortar showed a decrease. In the air-curing period, all samples decreased in length; most of the
shrinkage occurred in the first 14 days. Very little change occurred after 28 days in air with
basically no change after 56 days.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
212 MASONRY

e
o

Q
I

go
a
IB

w 0
a 3
0 1
•rt
e s ^
z 1—( p»
*
0)
Q.
. u. b.
u
s 9 S
H>< w w iJ
b b u
"O 0 0 0
U X Z Z
«
•«
e
s
eg J ij ij 0)
QO
>
Oi
4PI

CO eu a<
u ww

!
9 4 0 ^
i

o o o o o o

FIG. 21—Freeze thaw—9 cycles, standard Type N mortar.

TABLE 5—Freeze thaw test results.'

Mix Sample Initial Weight After Weight Avg,

No. No. Weight, g Cycles, g° Loss, %

Standard
Type N 1 243 236 2.9
2 243 235 3.3
3 241 233 3.3 3.5
4 242 236 2.5
5 242 233 3.7
6 241 228 5.4

EPL 1 218 218 0

2 217 217 0
3 217 217 0
4 218 218 0
5 217 217 0
6 225 225 0

"9 cycles for Standard Type N mortar, 50 cycles for EPL.

After 16 weeks the total percent length change of Type N standard mortar specimens was
0.147%. The extended plastic mortar specimen for 0, 7, and 24-h suspension times measured
0.120, 0.140, and 0.143%, respectively.
Corrosion—An attempt was made to examine the corrosion resistance characteristics of an
ASTM Type N mortar and the extended plastic life mortar for 0 suspension time. Mortar sam-
ples were made using two different specimen sizes, namely 5.08 cm wide by 5.08 cm high by
30.48 cm long and 10.16 cm wide by 10.16 cm high by 30.48 cm long (see Figs. 28 and 29). A
No. 3 steel rod was placed in the center of each specimen for either an embedded condition or an

FIG. 22—Freeze thaw—50 cycles. EPL mortar.

FIG. 23—Efflorescence—EPL mortar.

FIG. 24—Efflorescence—standard Type N mortar.

FIG. 25—Mortar shrinkage bars.

FIG. 26—Shrinkage measurements.

exposed condition as shown in Fig. 30. Three surface bar conditions were examined: zinc
plated, cadmium plated, and regular steel without surface treatment. Two specimens per mor-
tar type per bar surface coating per exposure condition were constructed. The No. 3 rod was
initially weighed and placed inside the molds in the middle of the specimen. All samples were
air cured in the laboratory for 28 days and then moved to the moist room. Each specimen was
then checked at 3, 6, and 9 months for extent of corrosion. At the end of each specified period
the rods were taken out of the molds, cleaned, and weighed. No definitive trends could be estab-
lished from this data.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
MATTHYS ET AL. ON EXTENDED LIFE MORTAR 217

I
o
to

00
CN
is
00
I
i^
•ti
^•^
•< ??
•?>
.k L
3
9

(ausojs<£) aSueqo qaSuai

FIG. 28—Corrosion test molds.

FIG. 29—Corrosion test specimens.

'JE P i-^H:
1" 1"
2"

i) Enbedded Condition, 2"x2"xl2" Hold b) Eabedded Condition, 4"x4"xl2" Mold

\
^
P 2"
T
:j -r
I"
2"
c ) Exposed C o n d i t i o n , 2"x2"xl2" Hold d) Exposed C o n d i t i o n , 4"x4"xl2", Hold

FIG. 30—Placement condition of rods for corrosion.

Summaiy—Phase 2
The one extended plastic life mortar in this phase showed equivalent or better performance as
compared to the standard ASTM Type N portland cement lime proportion mortar for all the
tests conducted on specimens constructed from initial mixing and the 6 to 7-h suspension pe-
riod. The EPL mortar exhibited sufficient compressive strength and also good bonding to the
masonry unit. Shrinkage at all suspension times of EPL mortar was less than Type N mortar.
Freeze-thaw resistance of the initially mixed EPL mortar was far superior to that of the con-
struction mortar. The EPL mortar did not appear to promote corrosion.
However, the bond strength and compressive strength of masonry prisms made for the 24-h
suspended mortar were about 50% less than those for Type N mortar. This effect is associated
with retempering of the mortar. In all cases the extended plastic life mortar had good workabil-
ity for a minimum period of 20 h. If retempering is needed it is recommended that the flow of
EPL mortar examined here be kept in the 80 to 100% range.

The Effectiveness of Waterproofing

Coatings, Surface Grouting, and
Tuckpointing on a Specific Project

REFERENCE: Coney, W. B. and Stockbridge, J. G., "The Effectiveness of Waterproofing Coat-

ings, Surface Grouting, and Tuckpointing on a Specific Project," Masonry: Materials, Design,
Construction, and Maintenance. ASTM STP 992, H. A. Harris, Ed., American Society for Test-
ing and Materials, Philadelphia, 1988, pp. 220-224.

ABSTRACT: Water leakage in a building gave the authors an opportunity to try a variety of
widely used procedures to improve the watertightness of masonry walls.
The water permeance of the masonry walls in their original condition ranged from 11 to 60 L/h.
Three types of clear waterproofing coatings, two types of surface grout, and conventional tuck-
pointing were tested. Each procedure was applied to two wall areas. Water permeance tests were
performed on each sample area in accordance with a field-modified version of ASTM Test Method
for Water Permeance of Masonry (E 514). Sample areas were tested prior to repair, after repair,
and in some cases after weathering through one winter.
The tests relate to one type of brick and mortar on one specific building. It would be unwise to
extrapolate our findings to other buildings, but the results are interesting, and we believe they
contribute to a growing wealth of experience.

KEY WORDS: masonry, water permeability, testing

In recent years, tuckpointing, surface grouting, and waterproof coatings have all been used
as methods for decreasing the water permeance of masonry walls, but to our knowledge, their
comparative effectiveness has been studied on few actual buildings in the field.
Water leakage in the Science building at Northeastern Illinois University in Chicago and an
aggressive University architect gave the authors an opportunity to study a variety of widely used
procedures to improve the watertightness of masonry walls. The Science building is four stories
high and was completed in 1972. Figure 1 shows a general view of the building from the north-
west. The bricks conform to ASTM Specification for Facing Brick (Solid Masonry Units Made
From Clay or Shale) (C 216). The mortar was analyzed and found to have Type N proportions in
accordance with ASTM Specification for Mortar for Unit Masonry (C 270). The mortar was
made with masonry cement containing an air entraining agent. Air contents of the mortar sam-
ples tested ranged from 16 to 2 6 % . Petrographic analysis revealed minimal intimacy of contact
between the mortar and brick. The contact area on the samples examined was estimated at only
about 2 5 % .
When tested in accordance with a field-modified version of ASTM Test Method for Water
Permeance of Masonry (E 514), walls were found to have water permeances of 11 to 60 L / h ,
which are some of the highest values we have seen.

'Member of the Evanston Preservation Commission. He is a specialist in rehabilitation and restoration.

^Senior architect and vice president, respectively, Wiss, Janney, Elstner Associates, Inc., Northbrook, IL
60062.

220

FIG. 1—General view of building.

Testing Procedures
As mentioned, the water permeance tests were performed based on a field-modified version of
ASTM E 514. Two test chambers were used. The test chambers each covered a wall area of 12
ft^. Test exposure conditions were equivalent to the standard 5.5 in. of rain per h accompanied
by a 62.5 mph wind. Our field test varied from the standard laboratory test in that the per-
meance was determined by measuring the amount of water entering the face of the wall rather
than measuring the amount of water exiting the back of the wall. Each test was run for 3 h or
until the 15-L capacity of our reservoir was drained. Figure 2 shows one of the test setups.

Test Area
The testing was performed on the masonry walls of the two penthouses on the roof of the
building because of easy access. There was one penthouse near the west end of the roof and one
near the east end. The two penthouses were each 82 ft long by 46 ft wide by 14 ft high.
The walls of each penthouse were divided into six test areas, creating a total of twelve test
areas, as shown in Fig. 3.

Pretreatment Permeance
One E 514 test was run in each of the twelve test areas before any of the masonry treatments
were performed. Test areas were selected to miss any cracks in the walls. At some locations
where cracks could not be avoided, they were tuckpointed before any testing was performed.
Shrinkage separations between brick and mortar were not tuckpointed. Even with visible cracks
eliminated from the test, water permeances from 11 to 60 L/h were measured as mentioned.
The permeance of each of the twelve tests is presented in Table 1.

Wall Treatments
After pretreatment testing, the twelve test areas were subjected to the following repair proce-
dures.

FIG. 2—Water permeance test setup.

FIG. 3—Test areas.

Tuckpointing
Wall Area 3 was tuckpointed in accordance with the recommendation of the Brick Institute
of America (BIA 7F). All existing mortar joints were cut out to a depth of at least V2 in. Walls
were hosed down about 1 h before tuckpointing to remove debris and to wet the brick. A prehy-
drated mortar mix of one part lime, one part cement, and five and one-half parts sand by vol-
ume was used. Portland cement met the requirements of ASTM C Specification for Portland
Cement (C 150). Lime met ASTM Specification for Hydrated Lime for Masonry Purposes (C
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
CONEY AND STOCKBRIDGE ON WATERPROOFING 223

TABLE 1—Summary of permeance testing.

Post
Wall Pretreatment Treatment Percent
Area Permeance° Treatment Permeance" Improvement

1 30 Proprietary grouting of joints 3.2 89

2 11 Acrylic coating 3.7 66
3 30 Tuckpointing with joints tooled 0.5 98
4 45 Alkyl trialkoxy silane coating 15.0 67
5 18 Alkyl trialkoxy silane coating 10.0 44
6 26 Aluminum stearate coating 10.0 62
7 24 Alkyl trialkoxy silane coating 5.5 77
8 22 Aluminum stearate coating 4.5 80
9 60 Proprietary grouting of joints 1.2 98
10 30 Grouting of entire wall surface 0.3 99
11 45 Tuckpointing with joints struck 2.3 95
12 60 Acrylic coating 45.0 25

"Units in liters per hour.

207), Type S. Sand met ASTM Specification for Aggregate for Masonry Mortar (C 144). The
joints were tooled to form a concave joint.
Wall Area 11 was tuckpointed in substantially the same manner and using the same materials
as Wall Area 3, except the joints were simply struck flush rather than being tooled.

Surface Grouting of Entire Wall Surface

Wall Area 10 was surface grouted in accordance with BIA Technical Note 7, July 1961. A
grout coating consisting of three quarters part portland cement, one part sand that passed a No.
30 sieve, and one quarter part hydrated lime was used. Shortly before using, all ingredients were
mixed with water to obtain a fluid consistency. The joints were thoroughly wetted but were
permitted to surface dry before the grout was applied. Stiff fiber brushes were used to work the
grout into the joints. No attempt was made to keep the bricks clean, and after all of the joints
were treated, the entire wall surface was painted with a diluted coating of the same grout. This
method resulted in the joints receiving two coats of grout. The appearance of the wall was
changed appreciably.

Proprietary Grouting of Mortar Joints

A proprietary grout was applied to the mortar joints of Wall Areas 1 and 9 with 6-in. nylon
masonry brushes. One coat was applied. Templates were used to keep the masonry units clean
while applying the grout to the joints. The proprietary grout is a cement-based compound meet-
ing Federal Specifications TT-P-21, Type 11, Class A. Walls were dampened but not saturated
before the application.

Aluminum Stearate Clear Coating

Wall Areas 6 and 8 were sprayed with a waterproofing coating of a product containing an
aluminum stearate-based compound. It was sprayed on the walls in accordance with manufac-
turer's recommendations. One coat was applied. It is advertised as a breathable coating.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
224 MASONRY

Alkyl Trialkoxy Silane Clear Coating

Wall Areas 4, 5, and 7 were sprayed with a waterproofing coating product containing 40%
(wt/wt) alkyl trialkoxy silane and 60% (wt/wt) ethyl alcohol.
The material was sprayed on the walls according to manufacturer's recommendations. One
coat was applied. It is advertised as a breathable coating.

Acrylic Clear Coating

Wall Areas 2 and 12 were sprayed with a waterproofing coating product containing acrylic (a
blend of polymeric resins containing mineral spirits and aromatic hydrocarbons). It was
sprayed on the walls according to the manufacturer's recommendations. One coat was applied.
It is advertised as a breathable coating.

Post Treatment Permeance

After the surface treatments were allowed to cure for at least 15 days, the permeance tests
were rerun. The same equipment was used, and it was positioned in the same positions, reusing
wall bolts installed during pretreatment testing.
All of the treatments reduced the water permeance of the masonry to some extent. The clear
waterproofing coatings were the least effective treatment. Surface grouting the joints in the
walls with the proprietary product was a much more effective treatment. Surface coating the
entire wall with grout and conventional tuckpointing with tooled joints were the most effective
treatments. The surface grouting did, however, change the appearance of the walls. None of the
treatments made the walls completely watertight. We have found clear waterproofing coatings
more effective on other projects than they were on this project. The permeance of each of the
twelve treated wall areas is presented in Table 1.

Permeance After a Winter of Weathering

It was originally planned to retest all of the test areas after they had gone through a winter of
service, but because of the limited success of the clear coatings, only the tuckpointed and the
surface-grouted areas were retested. These areas showed no significant change in ability to re-
sist water penetration between the fall and the spring testing.
Despite its good performance on the retests, surface grouting provides only a very thin layer of
new material over defects in mortar joints and common sense tells us that its longevity will
probably be less than tuckpointing where defects are ground out and a half inch of new, sound
material is installed.

Observations
1. All of the treatments reduced the water permeance of the masonry to some extent.
2. On this specific project, clear waterproof coatings were the least effective treatment. Sur-
face grouting of joints was a much more effective treatment. Surface grouting the entire wall
surface and conventional tuckpointing with tooled joints were the most effective treatments.
3. None of the treatments made the walls completely watertight.
4. Tuckpointing with struck joints was considerably less effective than tuckpointing with
tooled joints.
While it would be unwise to extrapolate our findings on this particular project to other proj-
ects, the results are interesting, and we believe that they contribute to a growing wealth of expe-
rience.

Diagnosis of Terra-Cotta Glaze Spailing

REFERENCE: Thomasen, S. E. and Searls, C. L., "Diagnosis of Terra-Cotta Glaze Spailing,"

Masonry: Materials, Design, Construction, and Maintenance, ASTM STP 992, H. A. Harris,
Ed., American Society for Testing and Materials, Piiiladelphia, 1988, pp. 227-236.

ABSTRACT: One of the most common types of deterioration observed in terra-cotta buildings is
glaze spailing. This separation of the hard, fairly impermeable glaze from the underlying clay
bisque exposes the clay to water infiltration and accelerated deterioration. The mechanisms re-
sponsible for glaze spailing are fabrication defects, incorrect installation, environmental expo-
sure, and improper maintenance. The investigation of glaze spailing consists of a field survey,
field tests, and laboratory analysis. Once a diagnosis of the causes is made, appropriate preventive
measures can be implemented.

KEY WORDS: terra-cotta, spailing, glaze spailing, restoration, masonry, deterioration, testing

Terra-cotta was a popular architectural cladding material from the 1880s through the 1930s,
coinciding with the building of American cities and the rise of the skyscrapers.
The typical terra-cotta block was fabricated by handpressing clay into wood forms. The back
of the block was open with internal stiffeners called webs. The glaze, which was brushed or
sprayed on the surface, was either a slip (clay wash) or an aqueous solution of mineral and metal
salts.
Terra-cotta is relatively durable and permanent because of the excellent weathering proper-
ties and the hard surface of the glaze. But, most terra-cotta buildings are now more than 50
years old and many claddings have started to deteriorate. The historical significance of these
buildings, the unmatched richness of detailing, and the vivid color of the terra-cotta make pres-
ervation of these structures important.
Deterioration of terra-cotta claddings takes many forms, but the one most frequently seen is
glaze spailing. Glaze spailing can first appear as a small blister from which the spailing can
grow to cover an entire terra-cotta block. Once glaze spailing has started, it allows a larger
amount of water to enter the bisque and in many cases further accelerates the breakdown of the
terra-cotta cladding.
It is important to properly diagnose the causes of glaze spailing before initiating a rehabilita-
tion program. This paper discusses glaze spailing and presents techniques for the testing and
diagnosis of the causes of glaze spailing.

Types of Glaze Spailing

Terra-cotta glaze spailing can be characterized by the depth of the failure plane.

'Senior consultant and senior engineer, respectively, Wiss, Janney, Elstner Associates, Inc., Emeryville,
CA 94608.

227

Failure Between Glaze and Bisque

This kind of glaze spalling is cliaracteristic of inherent material failure where the glaze was
not properly fused with the clay. Sometimes this failure takes many years to appear.

Failure in Boundary Bisque Layer

In this common type of glaze spalling, delamination occurs in the boundary layer of the
bisque, as shown in Fig. 1, with a small amount of clay body adhering to the spalled glaze. On
ashlar blocks, this glaze spalling often takes one of these forms:
Edge Spalling Below or Above Horizontal Joint—'i^t spalling originates at the edge of the
mortar joint or in some cases 20 to 80 mm below the joint. After the spalling has started, mois-
ture has access to the exposed bisque, and the spalling progresses rapidly upward or downward
from the joint.
Edge Spalling Adjacent to Vertical Edge—The spalling appears to start at the very edge of
the block. Water enters the bisque, as shown in Fig. 2, through the mortar or from the comers
of the block where the bisque is exposed as the mortar is worn away by the environment. This
type of spalling occurs more frequently on blocks where the glaze covers only the exposed face
and does not wrap around the edges of the block.
Round Spalls in Center of Glaze—A typical spall in the middle of a block is approximately
circular. The spall may have started at a pinhole in the glaze and the diameter increases as the
spall progresses outward from the center. Sometimes multiple small round spalls will combine
into larger spalling.

Causes of Glaze Spalling

The mechanisms generally recognized as responsible for glaze spalling of terra-cotta are fab-
rication defects, incorrect installation, environmental exposures, and improper maintenance.

Fabrication
Inherent Glaze Faults—Inadequate firing of the terra-cotta can result in undervitrification,
which makes the glaze permeable. High permeability can also result from the glaze being too

IIP^

FIG. 1—Glaze spalling in boundary layer of bisque occurs at mortar joint and in center of block.

MORTAR U ^WATER ENTERS BISQUE

JOINT THROUGH DEFECTIVE
MORTAR OR AT
EXPOSED CORNERS

SALT FORM/JION OR
FUNGUS GROWTH
EXERT PRESSURE
BELOW GLAZE

FIG. 2—Glaze spoiling from salt formation or biological growth at mortar joint.

thin or from insufficient flux in certain glaze formulas. Highly permeable glaze allows water
infiltration, which can lead to glaze spalling.
Inherent Bisque Faults—Many terra-cotta blocks were packed by hand into molds. Pressing
the clay layers into the mold sometimes produces microcracks and weak planes parallel to the
surface of the block immediately below the glaze. Spalling has later occurred along these weak
planes.
Crazing—Afiet the firing, the terra-cotta cools and both the glaze and the clay shrink. If the
coefficient of thermal expansion for the glaze is larger than the coefficient for the clay, the
cooling will create tension in the glaze and cause craze cracks in the glaze. The terra-cotta block
has reached its smallest dimension after the firing. As it absorbs moisture from the environ-
ment, the bisque expands and the nonabsorbent glaze remains stable. The expansion of the
bisque creates tension in the glaze and can result in crazing. The craze cracks are typically
microscopic in width and generally do not increase the absorption rate compared to glaze with-
out crazing. The overall effect of crazing on durability is not well understood, but it appears
that crazed glazes are less prone to damage from glaze spalling than uncrazed glazes.

Ittstallation
Terra-cotta was sometimes used as a load-bearing masonry, but more often as a cladding
anchored to the structural framing system. Normally, no provisions were made in the design for
movement, either absolute or differential, in the backup framing or in the cladding. The blocks
were typically supported vertically at each floor level by shelf angles. Ties or Z-shaped steel
straps were framed into slots or holes in each block and anchored the terra-cotta horizontally to
the backup walls of masonry or concrete. Ornamental units generally had multiple anchors.
The terra-cotta was installed with solid cement/lime mortar joints about 5 mm wide. The back
of the blocks were sometimes fully grouted and sometimes left ungrouted, but whether grouted
or ungrouted, the wall cladding had no internal flashings or weepholes.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
230 MASONRY

Strain Buildup—Buildings with concrete frames experience long-term shrinkage of the

frames, and differential movement of building components can also cause high stresses, but
most strain buildup in terra-cotta claddings results from moisture expansion of the terra-cotta
clay body. Wet/dry cycles can cause permanent expansion of the terra-cotta blocks, and this
expansion plus the absence of expansion joints in the facade creates high compressive stresses.
Failure is usually in the form of crushing or buckling in the bisque body or boundary, as shown
in Fig. 3.
Corrosion of Embedded Metals—Most shelf angles and anchors were originally protected
with paint but much of this paint has peeled away. As the mortar ages, it carbonates and loses
its ability to protect the steel from corrosion. Rust scale expansion is often found at shelf angles
and lintels where water is caught at the inside face of blocks. The failure, as shown in Fig. 4,
typically is seen at the toe of the shelf angle or lintel steel and usually results in spalling of the
clay body, not just the glaze.
Incomplete Anchoring—The terra-cotta cladding is typically supported vertically at the floor
levels on shelf angles and anchored horizontally to the backup wall. Incomplete vertical support
sometimes results in glaze spalling, but it has usually caused spalling of the bisque and vertical
cracking from shear stresses between the supported and the unsupported wall sections. Such

FIG. 3—Spalling caused by strain buildup.

FIG. 4—Spoiling at corroding shelf angles.

incomplete supports can be found at outside building corners, as shown in Fig. 5, where the
shelf angles are terminated short of the corner.

Environmental
Terra-cotta is generally durable and permanent because of the excellent weathering proper-
ties and the hard surface of the glaze, but the mechanical, chemical, and biological processes
resulting from environmental exposure will, with time, alter the durability of the facade.
Among the various environmental agents harmful to terra-cotta, moisture is the most impor-
tant. Water is a carrier for airborne chemicals: it transports salts, supports biological growth,
and exerts pressure when it freezes. Water is the single most damaging element to terra-cotta.
Environmental Chemicals—The vitreous glass-like surface of most terra-cotta glazes pro-
vides a durable barrier against most environmental chemicals and generally resists the adher-
ence of dirt and the accumulation of airborne gases and salts from acid rain. The exception is
some slip glazes, especially those with relatively high permeability or those with rough-textured
surfaces. While most glazes are not sensitive to carbon dioxide and sulphur dioxide, the long-
term exposure to the acids usually found in polluted air can cause dulling of the glaze. The
environmental chemicals will, however, more vigorously attack the mortar in the joints and this
will allow water infiltration and accelerated deterioration of the terra-cotta.
Water, Salt Deposits—Water enters the bisque through deteriorated mortar joints, perme-
able glaze, or damaged terra-cotta blocks, and it travels freely through the porous bisque and
masonry backup. Water soluble salts are carried along with the water and these salts are depos-
ited in the boundary layer of the bisque as the water evaporates through permeable glaze. The
buildup of salt deposits can exert considerable internal pressure which ultimately fractures the
boundary layer of the bisque.
Water, Biological Process—Algae, moss, and lichen thrive in a porous environment when
light and moisture are present. Most light-colored glazes permit sufficient sun transmission to
support biological growth. The presence of biological organisms prevents the bisque from dry-
ing out, and the chemical products of their metabolism disintegrate the mortar. The formation
of biological growth below the glaze can exert considerable internal pressure which ultimately
fractures the boundary bisque layer, as shown in Fig. 6. Glaze spalling from biological growth is
commonly found adjacent to mortar joints or adjacent to previously spalled sections where
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
232 MASONRY

FIG. 5—Cracking and spalling caused by incomplete vertical support around building comers.

m^k^-

FIG. 6—Glaze spalling from biological growth.

moisture and air are readily available; but it can also be found in the middle of an undamaged
block, where it can be seen as a glaze blister. As soon as the glaze has spalled, the underlying
bisque will dry and the biological growth will die or will move in under the edge of the adjacent
glaze.
Water, Freeze/Thaw—Frost damage is dependent on the cycles and speed of temperature
variations around the freezing point, the degree of saturation, and the pore distribution and
permeability of the bisque. While certain forms of cracking can be attributed to frost damage, it
appears that glaze spalling is seldom caused by freeze/thaw action in the bisque. This might be
attributed to the large percentage of air voids usually found in the terra-cotta bisque.

Maintenance
Delayed Maintenance—Deteriorated mortar joints, damaged terra-cotta blocks, and spalled
glaze are major sources of water infiltration into the bisque. Repairs and maintenance of the
terra-cotta should be performed at regular intervals to prevent accelerated deterioration. Water
infiltration through cornices, roofs, window sills, and other elements should be reduced by
proper maintenance.
Improper Repairs—More damage can be done from incorrect repair than from no repair.
Most glaze spalling is water related, and often the water infiltrates through deteriorated mortar
joints. While the joints permit water to enter into the bisque, the mortar joints also allow water
to escape in the same way. If the joints are covered with sealant, water that enters the block by
other sources will tend to evaporate through the glaze rather than the joint, and this will result
in glaze spalling.
Improper Cleaning—The hard vitreous surface of the glaze is most often self cleaning, but
dirt sometimes collects in crevices or at rough-textured surfaces. Cleaning of dirty surfaces
should be done with great care. Improper chemical cleaning agents can dull the gloss of the
glaze or attack the mortar, and any abrasive cleaning, especially sandblasting, will damage the
glaze surface and the increased permeability will result in future spalling.
Improper Coatings—Epoxy paints have sometimes been used to coat terra-cotta buildings
with glaze spalling because the glossy epoxy is a good visual match for the glaze. The epoxy
paint traps moisture inside the walls. The water ultimately evaporates through pinholes in the
coating and extensive spalling has resulted.

Techniques for Diagnosis of Glaze Spalling

Field Survey
The extent and the nature of the spalling is first evaluated by a visual survey of the building
facades. Distress in the terra-cotta is recorded on elevation drawings for analysis and detection
of failure trends. An inspection is made of possible sources of water infiltration such as leaking
gutters, damaged cornices, or deteriorated window sills. The glaze is examined closely for pin-
holes, mortar deterioration, and biological growth. In light-colored glazes, biological growth
can be seen as a blue-green color showing through the glaze. In advanced stages, the growth can
be seen as a buckling of the glaze outward from the clay substrate.
The marked-up field survey is analyzed for typical patterns in the spalling. Glaze spalling in
one building consistently started adjacent to the mortar joints and grew in towards the center of
the block. Examination revealed that water was being absorbed through the mortar joints and
through the corner of the terra-cotta block where the glaze did not wrap around the edge. In
another building, spalling occurred at each floor level. It was found that infiltrated moisture
collected on top of the shelf angles and caused spalling from corrosion and from salt deposits as
the water evaporated.

Field Tests
The deterioration and spalling of terra-cotta is closely related to water infiltration, and two
field tests are used to evaluate water tightness. The simplest test is done by attaching an open-
ended plastic tube to the terra-cotta with sealant tape, as shown in Fig. 7. The tube is filled with
a predetermined head of water, and the decrease in height of water in the tube, representing the
amount of water absorbed by the terra cotta, is recorded at intervals of 10 min, 30 min, 1 h, and
hourly thereafter. This test is useful for comparing a variety of substrates on a single building,
such as crazed glaze, glaze with pinholes, intact glaze, treated glaze, and mortar joints.
A portable aluminum test frame can be attached to the facade to perform a modified version
of ASTM Standard Method of Test for Water Permeance of Masonry (E 514). Water and air
pressure are applied simultaneously to the wall. The permeance is determined by measuring the
loss of water in the recirculating system.
Strain relief testing measures the magnitude and direction of the built-up stresses often
caused by expansion of the terra-cotta and absence of expansion joints. Electrical resistance
strain gages are attached to the terra-cotta surface and the gages are read. Then the terra-cotta
block, with the gages attached, is cut loose from the wall and the gages are read again. The
change in gage reading is a measure of the strain in the block. The stress in the block is found by
multiplying the measured strain difference by the modulus of elasticity, as determined by labo-
ratory tests. The glaze in the test area should be firmly attached to the clay body and no glaze
cracks should occur under the gage. Temperature variations during the day can affect the read-
ings and should be recorded when strain readings are recorded. Compressive vertical stresses as
high as 23 MPa have been recorded. This exceeds the maximum compressive strength of some
terra-cotta. From multiple strain measurements, a stress map can be made for the exterior
elevation. The map is used to evaluate if cutting of expansion joints into the facade will relieve
the built-up terra-cotta stresses.

Laboratory Tests
Laboratory tests can be performed on terra-cotta to evaluate the performance of the glaze and
to determine the causes of glaze spalling.
Petrographic Analysis—The consistencies of the glaze and the clay body are evaluated

FIG. 7—Field test for water absorption.

through a stereomicroscopic examination. The density of the glaze surface, the composition of
the material, and the degree of deterioration can be established by an experienced petrogra-
pher. The examination of the boundary layer between glaze and clay body is important, and the
nature, the magnitude, and the depth of cracks often are an indication of the future perfor-
mance of the glaze.
Absorption—The absorption test compares the performance of glazed and unglazed speci-
mens. On one terra-cotta sample, the glaze is ground off while on an identical sample the glaze
is left intact. The sides of both samples are coated so that water absorption occurs only through
the face. The samples are soaked face down in water for 24 h. The weight gain of the glazed
compared to the unglazed sample is a measure of the absorption characteristics of the glaze.
Ideally, a glazed specimen should produce zero absorption if the glaze is intact, sound and
craze-free. New glazes are normally impervious, but tests of 40- to 80-year-old terra-cotta struc-
tures found that the glaze at best reduced water absorption about 70% and at worst only 10%
compared to unglazed terra-cotta. High water absorption through the glaze is not necessarily
indicative of impending failure if the terra-cotta has the ability for water vapor to be transmitted
from the clay out through the glaze. Moisture which is trapped behind the glazed surface is the
more likely cause of spalling failures.
Thermal Coefficients—Tests are performed both on the complete terra-cotta block and on
separate samples of glaze and clay body. Strain gages are mounted on the samples which are
then subjected to temperature ranges representing the normal wall exposure. Strain readings
are taken at the high, the low, and an intermediate temperature.
Moisture Expansion—During the firing, the free water is removed from the terra-cotta. As it
again absorbs moisture, the clay body expands. Some of the expansion is cyclic, but a portion is
nonrecoverable. A general magnitude of the moisture expansion can be determined by a reheat
test. A terra-cotta sample is measured at 21 °C. The sample is then heated to 460°C, again
allowed to cool to 21 °C, and then measured. The measured shrinkage is an indication of some
of the long-term moisture expansion of the terra-cotta. Moisture expansion of the clay body is
the cause of much of the cracking, spalling, and buildup of stresses in the terra-cotta walls.
Glaze Adhesion—Glaze spalling may occur because the bond between the glaze and clay has
deteriorated or because it was not initially well bonded. The glaze adhesion test is performed by
attaching a 25 by 25 by 100 mm test bar to the face of the terra-cotta. The bar is then knocked
off and the fracture surface examined. If the glaze is well adhered to the clay, pieces of the clay
body will come off with the glaze. A clean separation of glaze and clay indicates poor adhesion.

Preventive Measures
The causes of glaze spalling must be addressed and preventive measures implemented to
reduce the ongoing deterioration. The repair of glaze spalling is often part of a larger restora-
tion project. Damaged blocks can be replaced or the spalled section can be patched by applying
a surface coating to the exposed bisque.
Since moisture is a common cause of glaze spalling, the elimination of water infiltration is of
primary importance. Defective mortar joints should be repointed, and horizontal surfaces
should slope to drain or they should be protected against water infiltration. It is generally not
recommended to apply coatings to vertical surfaces. Reducing water infiltration will also cut off
the supply of moisture for biological growth, but treatment with a biocide may be necessary for
some infestations.

Bibiiography
Berryman, N. C , "History of Architectural Terra-Cotta," National Council on Education for the Ceramic
Arts (NCECA) Journal, 1981.

Fidler, ] . , "The Conservation of Architectural Terra-Cotta and Faience," Association for Studies in the
Conservation of Historic Buildings Transactions, Vol. 6, 1981, reprinted in Friends of Terra Cotta
newsletter. Vol. 3, No. 3, Fall/Winter 1984.
Patterson Tiller, de Teel, "The Preservation of Historic Glazed Architectural Terra-Cotta," Preservation
Briefs No. 7, Technical Preservation Services Division, U.S. Department of the Interior, Washington,
DC, June 1979.
Prudon, T. H. M., "Architectural Terra Cotta: Analyzing the Deterioration Problems and Restoration
Approaches," Technology and Conservation, Vol. 3, Autumn 1978.
Robinson, G. S., "The Reversibility of Moisture Expansion," Ceramic Bulletin, Vol. 64, No. 5, 1985.
Stockbridge, J. G., "Evaluation of Terra Cotta on Inservice Structures," Durability of Building Materials
and Components, STP 691, American Society for Testing and Materials, Philadelphia, PA, 1980.
Thomasen, S. E., "Degradation and Rehabilitation of Terra Cotta," Second International Conference on
the Durability of Building Materials and Components, National Bureau of Standards, Gaithersburg,
MD, September 1981.
Thomasen, S. E. and Ewart, C. S., "Techniques for Testing, Analyzing and Rehabilitation of Terra
Cotta," Strengthening of Building Structures—Diagnosis and Therapy, lABSH Symposium, Venezia
1983, final report. International Association for Bridge and Structural Engineering, ETH—Hongger-
berg, CH-8093 Zurich, Switzerland.
Tindall, S. M., "Architectural Terra-Cotta Restoration," National Council on Education for the Ceramic
Arts (NCECA) Journal, 1981.

In-Situ Evaluation of Compressive

Stresses

REFERENCE: Manmohan, D., Schwein, R. L., and Wyllie, L. A., Jr., "In-Sitn Eraluatioii of
Comprefsive Stresses," Masonry: Materials, Design, Construction, and Maintenance, ASTM
STP 992, H. A. Harris, Ed., American Society for Testing and Materials, Philadelphia, 1988,
pp. 237-250.

ABSTRACT: The use of terra cotta as a cladding material emerged in the 1880s and began to
decline in the late 1930s. During this period, high-rise buildings employing terra cotta cladding
included no provisions for differential movement between the cladding and structural framing.
Numerous terra cotta clad buildings are showing signs of distress due to weathering and induced
stresses from frame shortening under load. Determination of residual compressive stresses within
the terra cotta cladding is often necessary for evaluation and repair of the distressed material.
The level of stresses induced in the cladding of a high-rise building in San Francisco was mea-
sured by performing strain relief tests. Uniaxial strain gauges were adhered vertically to the face of
the terra cotta blocks and initial balance readings were obtained. The mortar bed joints were sawn
along the terra cotta units and the strains monitored during and after the cutting. Samples of the
terra cotta block were cut, instrumented, and tested in compression to determine elastic proper-
ties and ultimate compressive strength.
Having obtained the physical properties of the terra cotta, compressive stresses due to frame
shortening in the structure were determined. Strain data obtained indicated stress levels which
were low enough to eliminate the need for stress relief. Optimum saw cutting locations were deter-
mined had stress levels been high enough to require stress relief.

KEY WORDS: terra cotta facade, residual compressive stresses, in-situ evaluation, strain mea-
surements

Glazed terra cotta was a popular facade material for multi-story buildings constructed be-
tween 1880 and 1930. Originally it was employed in the traditional structural load-bearing ap-
plication in masonry walls in buildings of modest height. Subsequently, its use was widened to
high-rise construction where it was used as cladding in purely architectural applications.
Terra cotta, like other building materials, is subject to weathering and deterioration. Most
failures like spalling and crazing are due to moisture [1]. Vertical cracks, however, that run
through several units, stories, or large areas of material are often due to excessive stresses.
These stresses could be due to thermal movements or shortening of the structure relative to the
terra cotta or both. Stress failures are indicative of inadequate design. An understanding of
differential movement and how to compensate for it came after many terra cotta facades were
constructed without flexible joints for stress relief.
Reports of terra cotta restoration on the Woolworth Building in New York [2-4] indicated
very high compressive stresses. Stockbridge [3] reported measuring compressive stresses of 4.8

'Principal, Applied Materials & Engineering, Inc., Alameda, CA 94501.

^Schwein/Christensen Engineering, Lafayette, CA 94549.
^Vice president, H. J. Degenkolb Associates, Engineers, San Francisco, CA 94014.

237

to 23.9 MPa (680 to 3400 psi) in 21 terra cotta clad columns. None of the above reports, how-
ever, presented details of experimental techniques used to measure or relieve these stresses.
This paper discusses experimental techniques employed to determine compressive stresses in
the terra cotta facade of the Pacific Telesis Headquarters Building in San Francisco.
The study was undertaken with the following objectives in mind:
1. To determine the compressive stresses present in the terra cotta units.
2. If excessive, to determine optimum techniques for stress relief that would be effective in
the field.
The Pacific Telesis Headquarters (Fig. 1) is a 27-story building constructed in 1924. The
building has a structural steel frame with reinforced concrete floor slabs. The exterior walls
consist of glazed terra cotta with unreinforced brick masonry backing. The brick and terra cotta
are supported by structural steel ledgers 7.62 by 7.62 cm (3 by 3 in. angles) cantilevered out
from the floor framing. Though the terra cotta is supported on everyfloor,there are no provi-
sions for expansion joints in the cladding.
Figure 2 is a cross-section of a column showing the typical construction. The terra cotta units
are 76.2 cm long by 45.7 cm high by 2.5 cm thick (30 in. long by 18 in. high by 1 in. thick) and

FIG. 1—Photograph of Pacific Telesis Headquarters Building, San Francisco, California.

I
a.
I
o

are attached to the brick by tie wires hooked through attachment holes at the top and bottom of
each unit. The tie wires are embedded in the brick mortar.
Distress to the facade was manifested by spalled glaze and fine cracks in some of the terra
cotta units.

Procedntes
The test procedures to measure stresses due to frame shortening can be summarized as fol-
lows:
1. Instrumenting selected terra cotta units with strain gauges.
2. Measuring and establishing the as-instrumented strain as the base line strain.
3. Saw cutting two horizontal mortar joints and remeasuring strain.
4. Determining the compressive strength and modulus of elasticity of the terra cotta to con-
vert strain to stress. The magnitude of the change in strain, and hence stress, was then assumed
to be the residual stress due to frame shortening.

Selection of Strain Gauge Locations

A site on the south face of the building at the sixth floor elevation was selected for the study.
This particular bay is located just east of the first column line from the southeast comer of the
building. The distress in this area was considered typical of that observed in the terra cotta
cladding on the entire building. Two vertical strips were instrumented, a column area and a
mullion area (Fig. 3). Figure 4 is a schematic which shows the blocks that were instrumented.

FIG. 3—Photograph of elevation instrumented with gauges. The shadows are across the fifth floor win-
dows.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
MANMOHAN ET AL. ON COMPRESSIVE STRESSES 241

S
o
L^ S

O
at
d a 10
d
in
d
z
o
IT a: z
o O o g (E
o O o
3
^^ ^^
IE
I-

As can be seen in Fig. 4, the main concentration of gauges was on terra eotta blocks on either
side of the saw cuts. There was one gauge placed one floor above the saw cuts and three gauges
spaced out two floors below. The four locations on the 4th, 5th, and 7th floors were selected to
determine the extent of stress relief provided on floors adjacent to the one with saw cuts. A total
of 14 gauges were installed on the column units and 12 on the mullion units.

Gauge Installations
Micro-Measurement's CEA-06-500UW-120 electronic strain gauges were selected for mak-
ing strain measurements. These gauges have an active length of 1.2 cm (V2 in.). Based on the
fact that terra cotta is a relatively fine-grained material, it was decided that an active gauge
length of 0.64 to 2.54 cm (V4 to 1 in.) would provide satisfactory results. The other important
selection criterion was that the above gauges are self-temperature compensating [at ambient
temperatures of 10 to 38°C (50 to 100°F)] for a coefficient of thermal expansion of 10.8 X 10~*
cm/cm/°C (6 X 10~* in./in./°F). Prudon et al. [3] measured the thermal coefficient of expan-
sion of the glaze and of the terra cotta to be 16.6 cm/cm/°C (9.2 X 10~* in./in./°F) and 7.92
cm/cm/°C (4.4 X 10~* in./in./°F), respectively, making these gauges ideal for the applica-
tion.
The glazed faces of the terra cotta tile units were uneven and contained small surface asperi-
ties. A cylindrical drum grinder equipped with 100-grit abrasive was used to smooth and pre-
pare the glazed surfaces. Precautions were taken to prevent the glaze from being penetrated.
The surface was finished by hand sanding with 120 and 240-grit emery paper. The gauges were
bonded with a cyanoacrylate adhesive in accordance with the manufacturer's instructions,
wired, and protected with a urethane coating and rubber pad. The wiring was routed along the
face of the building and attached with duct tape. Each of the gauge wires was cut to a common
length and routed inside a nearby office. A strain gauge was mounted on a cut piece of terra
cotta and was removed from the building for use as a temperature-compensating element which
was free of load-induced strains. The compensation gauge was also wired with a length of con-
ductor equal to that of the active gauges.
Each of the 26-gauge circuits was calibrated with shunt resistors which are equivalent to 500
and 1000 microstrain. A Daytronics digital strain instrument was used for measuring strains.
Vishay switch and balance units were incorporated into the circuitry to allow rapid scanning of
the strain gauges.
Strain measurements were made for two days prior to the stress relief operations. These mea-
surements were made to determine the effect of thermally induced strains due to varying sun-
light on the facade.

Strain Measurements
Stress relief operations were carried out in the evenings, several hours after the sun had left
the south face of the building. This was done deliberately to minimize the effect of thermal
strains. While making strain measurements, the temperature-compensating gauge was placed
on the outside face of the building at the sixth floor.
The sequence of saw cutting is reported in Table 1. Strains were measured on the column
gauges after each cut on the column mortar joint. Similar measurements were made on the
mullion. Final measurements were made two days after the saw cutting operations were com-
plete.
Saw cutting was accomplished dry using a regular Skil-saw and a diamond-tipped blade. The
cuts were deep enough to cut through the terra cotta, leaving the brick intact.
Subsequent to saw cutting of the mortar joints, terra cotta Blocks 8 and 9 from the column
and Block 7 from the mullion were removed and taken to the laboratory. Cube samples were

TABLE 1—Schedule of saw cutting and strain measurements.

Time Condition

04/26/86
6:05 p.m. Initial balance of strain gauges.
6:53 p.m. Completed cut on Joint B of column. Measure strain on column
gauges.
7:02 p.m. Completed cut on Joint A of column. Measure strain on column
gauges.
7:22 p.m. Completed cut on Joint A of mullion. Measure strain on mullion
gauges.
7:33 p.m. Completed cut on Joint B of mullion. Measure strain on mullion
gauges.

04/27/86
4:31 p.m. Measure strain on both column and mullion gauges.

04/28/86
4:36 p.m. Measure strain on both column and mullion gauges.

prepared from each of the three blocks and instrumented with strain gauges. Two gauges, iden-
tical to those used in the field, were applied to opposite faces of the cubes. One gauge was placed
on the glazed surface and the other on the fired clay face. The cubes were tested in compression
to obtain stress versus strain data to determine the compressive modulus of elasticity. The cal-
culated modulus for the glaze and fired clay were averaged and the average modulus used for
determining field frame shortening stresses. In addition, the compressive strength of the terra
cotta was also determined.

Revolts
The stress versus strain values for terra cotta samples tested in compression are shown in Fig.
5. The average modulus of elasticity of the two samples was determined to be 10.92 X 10' kPa
(1.56 X 10* psi). This value was used for determining frame shortening stresses in the terra
cotta blocks. The average compressive strength of the two terra cotta blocks was measured to be
23.70 MPa (3300 psi).
The strains measured after each saw cut and at 24 and 48 h after saw cutting are tabulated in
Tables 2 and 3 for the column and mullion gauges, respectively. Corresponding calculated
stress relief has also been reported. Figure 6 is a graphical representation of stresses immedi-
ately after the final cut and 24 h later for the column gauges. Figure 7 is a similar representation
as above for the mullion gauges.
Immediately after the final cut the maximum strain and stress relief measured in the column
were 520 microstrain and 5.6 MPa (800 psi), respectively. The maximum strain relief was mea-
sured at Location 8 between saw cuts. The maximum strain relief in the mullion cladding imme-
diately after saw cutting was 125 microstrain, corresponding to a maximum stress of approxi-
mately 1.40 MPa (200 psi). This location was also in between saw cuts.

Discussion
On the column, the maximum and average stress relief provided by the saw cutting operation
was approximately 5.6 MPa (800 psi) (Location 8) and 3.5 MPa (500 psi) (Locations 7, 8, 9, and

00 <J>
y:
o ^
o_1 8_i
CD_j
m
z z
z
3
1
8
vV ly
I
<0 3
8

z
o
o
o

oin
li^*"^

IS
^

—t o
ID

Sv
CM
^

6 «o
s1 i8= (ISd)
o
»dW SS3U1S

——w w S S S •' : ro <N n rt

« f*^ ^ 1^ O ^
(N (N ^ 00 00 ro vO ; I/) OS f*^ r--^ ^
O O O O <N rn »N
• r-i -"' <N o o

O O >0 U1 O O lO ° . m Q o 1/2 l o
<N ( S -H 1 ^ vO -H Tf . (O 00 Cs) so ^
( N fO (N
. <S — (M

"fc

1^ -^r CO OS o
• t^ 1/2 <N >0
BO ' f*> r^ ro
•a S
c u "^1 O O O O P4 r i <N • <N 'H (N O O
%

2 gI
.a O O O lO O U1 O "
r~ lo o •*
<N ro (N

l| OOOO<N(Nr<i/7f0<N^' a.
o

oi/)ot/)*/j»/2i/)oi/)OLr>t/5oo
Trfnnt^roi/2i/)n»-if*:t^sOi/2Tr
I (Nrs»(Sionfs^^

a So so OS ^ <N
i-i
CO '^ Oi

s-H TTsOsOOsOOcO^OOsO
< in
2 ^ ui^fS^rj^-HsiTT^
3
o o o c> (N <N —' TT r-i -H o o* o o t«)
so s
e u
u 'S3
3 M
E
Ul

«t
rsi rsi »^ f*) ^ ^ "S JO>.
u
13 *-
3
u U

IJ »^<Np^Trt/>soi^cooso^-'rs(fO'^
Kl ^^ ^H ^H f-<
a
55 ^
1

O ^ "^ t^ "-I -O •' o^ 00 o^ f i

(N i/^ r^ ro ^ • f^ O « <N
00 fS CO ; f i 1/1 i-H ^
\o « TT r^ T-^ •" t-s. sO GO O
•-< r^ I/) ^O fN ; f^ m - H ^
O O O O ^ (N • ra f*i "H o

O i/> i/l O O i/> , o o o u^

i-H ro lO I/) O
, fS m -^ 1-1
, <S f*^ »-<
tj.
O ON O 0^ fO
1 S •c

oa ^^ N-- • . >_^ s_^ ^-•

1^ >
_^<_/ .*^
« • h-<riOin 00••'^
O O O O -H fS ts n TH o

1/1 O I/) O O I/) • ". o m o m

ii . <S <S .H rt

I . (N l»i rt

^ ^ M ^ (S

11 O O O O O ^ ^ ^ O ' - ^ O O

1 1 M •* t 'T <N
a.
h
X
O do' 't7\ -^^ ^ ' NO vo' Q (N K ^ ro
- *5 6
^ QLI

II
oooooooooooo
ao e
_e u
'M E
9
•o ^S

II C)
»
3<« j j
o
"3 .^
u o
M M
9

II •H<Nro^u^^f~-oOO^O^^<N e
^

iQ8
lO 1^
1 s.=
in
z P-.
*
S$ o o fS -3
(E
CM
in o
00 lO in

2 ^ in •a
UJ §
CM
>
^^ tf> ;*
-1
m o in •r
K ! t^ in UJ
— (M (E s3
a OL
<M z i
4 o 9

^4
<
1 \i ^i
• J
2 5!

in -
fei zo
• 1 (0 r- (om « •O ( 1

s 4
1-
3
U
<n
z
UJ

1
1
at
1
•^
u
§
•fe

1 z

CVJ
00
I I" O
Kj ID
"
•-
<"
UJ
2 >
CO
_l in
•p: UJ
in Q Q.

i(/> -£i Z
oo §•

1
-*
1
10 *l
_ 2 <> (0
1
K <sn ^
1
lO (1
1 i|^3 oo-
O
o
^i 1D
1 <
3
in
3 U

p
ccto (Ein

4
•"tas
so.
u
•s
UJ ^ CO
Q O »> -a
•e
.ino UJ
K) in oe
< Si
in CO •a
s
UJ Q

> On
•~ B> s
en cc * mom
• N m UJ
• «

ts
No
tlJ 0 .
Z • ^ a. d
o< eor: Z
<
o
z
8 1
g
_)
-J
i
(M ^
:: z
1
1
1. in •
<1 1 'inih
-
• oo-l-
1
->

2
15
"g
I <n
z 1- zUJ <
g U U H-
::5
(- 1
3
H c
o
O t • ^

S
Nm -1'-
•"1^355
u
(C
in
^S; i
fe
**-»
<n in
UJ in
a: O S S!
t-
00
-no JE
lo in u)
15
s
UJ ^
> tQ.
<«
m O UJ 8!
-fw in K n
z _• CM u-
- S
.3
*t;.
o
u P
o
OTP
-oo-i-
: : ' 2 p> z
z g g
o in
(94 WOT
30
<o !

10), respectively. The compressive strength of the terra cotta was determined to be 22.4 MPa
(3200 psi), which is four times the maximum stress measured. In this case, therefore, it was
decided not to proceed with stress-relieving operations on the remainder of the building. The
above results were consistent with the distress noted. There was little evidence of large longitu-
dinal or horizontal cracks traversing several pieces of terra cotta, as would be expected if com-
pressive failure had occurred.
Had the stress relief measured been greater than one half the compressive strength of the
terra cotta, stress relieving may have been required. Prior to making this decision, however, at
least one other area would have been instrumented and relief stresses measured to verify find-
ings of the original data.
As can be seen from Fig. 6, stress relief at Locations 4 and 13 on the column are approxi-
mately 35% of the relief at Locations 7 to 10. These locations are only one block away from the
saw cuts. In the event that it was necessary to stress relieve the entire building, the recom-
mended saw cutting option may have been every other bed joint on the columns. It is likely that
saw cutting every other joint would have relieved the stress uniformly and to acceptable levels.
For reasons similar to those presented above, excessive stresses on the mullion terra cotta would
also have been relieved uniformly and to acceptable levels by saw cutting out every other joint.
In order to determine the optimum spacing of cuts, the following need to be considered:
1. Compressive strength of the terra cotta to establish the maximum acceptable stress level.
2. Decay pattern of stress relief away from cut joint until it reaches an acceptable level.
3. Verification of the stress relief procedures by performing cuts at optimum locations on a
new section of instrumented terra cotta.
4. The configuration of the facade. Window openings, block layout, and future maintenance
of cut and sealed joints.
It should be pointed out that the stress relief pattern shown in Fig. 6 closely follows the ex-
pected behavior of the cladding under residual compressive loading in the structure. The stress
relief due to saw cutting should be highest directly adjacent to the horizontal cuts, diminishing
with distance due to shear drag along the masonry backing-to-terra cotta joint. Within experi-
mental error, negligible stress relief was obtained on adjacent floors.
No attempt was made to monitor strain after removal of the terra cotta blocks. Thus, we did
not measure stress relief due to shear drag from brick backing bonded to the terra cotta through
the mortar joint. Compressive stresses due to shear drag, in our opinion, are relatively low and
of a magnitude of less than 0.70 MPa (100 psi) depending on the quality of the bond; stresses
higher than above would have resulted in shear failure at the mortar-terra cotta interface. Shear
drag stresses would not be higher than 0.70 MPa (100 psi) for the following reasons:
1. Current UBC (Uniform Building Code) requirements [Chapter 30, Section 3004 (b)] for
minimum shear strength of bonded veneer is 0.35 MPa (50 psi).
2. Experimental data obtained in our laboratory of mortar bonded to brick indicate the high-
est shear strength obtainable to be approximately 1.05 MPa (150 psi). This value was measured
using current materials technology and under laboratory conditions. Considering that the sub-
ject project was constructed in 1924, shear strength of bond will definitely be less than 1.05 MPa
(150 psi) and most likely less than 0.70 MPa (100 psi).
The maximum stress relief expected due to shear drag [<0.70 MPa (100 psi)] is, therefore,
substantially less than the maximum stress relief measured on the terra cotta [5.6 MPa (800
psi)].
On the column, strains measured 24 h after saw cutting indicated that there was additional
stress relief of approximately 13% at locations adjacent to the saw cuts. Little or no change was
measured at locations between saw cuts. The changes in strain after 24 h were insignificant.
On the mullion, however, the additional average stress relief at 24 h on blocks adjacent to saw
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
250 MASONRY

cuts (Numbers 5, 6, 9, and 10) was 130% of the average relief immediately after saw cutting.
This is visually apparent from Fig. 7. We are not sure of the reason for the magnitude of delayed
stress relief on the mullion as compared to the column. Nevertheless, it is advisable to wait at
least 24 h to obtain final stress measurements. As on the column, there were insignificant
changes in strain recorded after 24 h.
During this study, it was assumed that the stress distribution across the full thickness of the
terra cotta units and the mechanical property specimens was linear. This is not entirely accurate
as the glaze layer is of higher modulus than the fired clay core and, consequently, exhibits
higher stress values than the core under a given strain. It should be kept in mind, however, that
the fired clay is extremely thick compared to the glaze and will carry most of the load.
Strains measured in the field were converted to stress from modulus values obtained only on
terra cotta. This conversion did not take into consideration the brick backing behind the terra
cotta.
No significant strains were induced by load or thermal changes in the building during the
two-day period when strains were recorded prior to the saw cutting.

ConclnsioD
The procedures described in this paper can provide valuable information on stresses built up
in terra cotta facades due to lack of expansion joints. This information is required during resto-
ration in order to make decisions on the need for and methods of repair.
The relatively low stresses measured resulted in considerable cost savings on the restoration of
the Pacific Telesis Building.

References
[/) Wilson, F., Building Materials Evaluation Handbook, Van Nostrand Reinhold Co., New York, 1984,
pp. %-99.
(2) Stockbridge, J. G., "Evaluation of Terra Cotta on In-Service Structures," Durability of Building Mate-
rials and Components, ASTM STP 691, P. I. Sereda and G. G. Litvan, Eds., American Society for
Testing and Materials, Philadelphia, 1980, pp. 218-230.
[3] Prudon, T. and Stockbridge, J., "Restoration of the Facade of the Woolworth Building," Rehabilita-
tion, Renovation and Preservation of Concrete and Masonry Structures, ACISP-85, G. Sabnis, Ed.,
American Concrete Institute, Detroit, 1985, pp. 209-227.
[4] Gaskie, M. ¥., Architectural Record, Vol. 169, No. 11, Mid-August 1981, pp. 90-95.

Restoration of the John J. Glessner

House

REFERENCE: Sourlis, T., "Restoration of the John J. Glessner House," Masonry: Materials,
Design, Construction, and Maintenance, ASTM STP 992, H. A. Harris, Ed., American Society
for Testing and Materials, Philadelphia, 1988, pp. 251-256.

ABSTRACT: Glessner House in Chicago was designed in 1886 by H. H. Richardson contempora-

neously with the first skyscrapers. The Chicago Architecture Foundation, the present owner of
Glessner House, commissioned tuck-pointing and cleaning of exterior granite facades for the
building's centennial year. Sourlis Masonry Restoration, Inc. was awarded the contract and pro-
ceeded with the project in seven phases: (1) rigging and protection of the building; (2) grinding of
mortar joints; (3) preliminary tuck-pointing of mortar joints to prevent penetration of caustic
cleaning chemicals; (4) cleaning of the granite; (5) repair of the chimneys; (6) pink beading of the
mortar joints; and (7) breakdown of scaffolding and landscaping repairs.

KEY WORDS: granite, masonry, restoration, historic structure, tuck-pointing, mortar joint,
scaffolding, alkaline gel

History of Glessner House

The John J. Glessner House, 1800 S. Prairie, Chicago, IL, was designed in 1886 by Henry
Hobson Richardson for the family of the Chicago industrialist whose name it bears. Construc-
tion had barely begun at the time of Richardson's death in April of 1886. This was a crucial
moment in the history of American architecture for the year 1886 also saw the completion of
Burnham & Root's Rookery Building in Chicago, one of the first structures to use a metal skele-
ton to support exterior stone walls. Although Richardson himself never built the skyscrapers
which would develop from this technique, his expressive use of stone, both in residences such as
Glessner House (Fig. 1) and in commercial buildings such as the Marshall Field Warehouse
(Chicago, 1885-1887), influenced later architects of the modern metropolitan skyline.
The particular stone selected by Richardson for Glessner House was Milford granite, so
named because it was quarried in Milford, MA. The original construction of the building was
carried out by Norcross Brothers, the first general contractor of the nineteenth century. They
maintained their own quarries and lumber mill and were responsible for most of Richardson's
important buildings. Glessner House has exterior masonry-bearing walls of gray granite ar-
ranged in a U-shape around an interior courtyard, thus shutting out the turmoil of the city.
Although the facades are simple with little ornamentation, they are composed of beautifully
proportioned structural elements including particularly fine arches. The plan for Glessner
House was a strikingly modern statement for a period of time when most residences were being
designed in French chateau and Victorian style.

'President, Sourlis Masonry Restoration, Inc., Munster, IN 46321.

251
Copyright by ASTM
Copyright' 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
252 MASONRY

FIG. i—East facade (front).

Historic Preservation
Preserving a historic structure such as Glessner House is difficult and expensive; great care
must be talcen so that irreplaceable elements of the building are not damaged. Also, the restora-
tion must be historically accurate. Both the scholarly interest andthe funding which made pos-
sible this type of restoration came about because of a change in attitude toward our architec-
tural heritage in the 1960s. People began to have a deeper regard for older buildings constructed
of time-honored materials. The movement gained momentum with the National Historic Pres-
ervation Act of 1966. In 1976, the year of America's bicentennial, the Tax Reform Act included
legislation encouraging owners of historic property to preserve it. This law disallowed both the
deduction of demolition expenses and the accelerated depreciation of any new structure built on
the site. As a consequence of these events, restoration of historic buildings such as Glessner
House has become more widespread.
The National Historic Preservation Act of 1966 coincided with the purchase of Glessner
House by the Chicago School of Architecture Foundation (now the Chicago Architecture Foun-
dation). Glessner House was designated a Chicago landmark in October of 1970. After focusing
early restoration work on the interior of the house, the Foundation decided to prepare for the
building's 100th birthday by commissioning complete tuck-pointing and cleaning of the exte-
rior granite facades. This was a major undertaking since the building had weathered two of
America's severe pollution eras: the smoke from early factories and locomotives and the exhaust
fumes from the ever more numerous automobiles. The granite walls of Glessner House had gone
from a sparkling light gray to soot black.

Bidding tlie Project

Sourlis Masonry Restoration, Inc. initially bid the project at one price which included all
tuck-pointing and cleaning of the granite facades and the restoration of five of the seven chim-
neys. For the project layout, we defined the order of work in seven phases: (1) rigging and
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
SOURLIS ON GLESSNER HOUSE 253

protection of the building; (2) grinding of mortar joints; (3) preliminary tuck-pointing of mortar
joints to prevent penetration of caustic cleaning chemicals; (4) cleaning of the granite; (5) repair
of the chimneys; (6) pink beading of the mortar joints; and (7) breakdown of scaffolding and
landscaping repairs.
The original spec called for complete pipe scaffolding of the wall perimeter in order to insure
a minimum amount of contact with the delicate terra-cotta tiles of the roof and with the granite
stones of the facades. We devised a system of rigging the building with pipe scaffold towers at 20
and 12-ft intervals and hanging swing stages in between the towers. Because the swing stage is
the most efficient working platform for a tuck-pointer, this innovation enabled us to bid the
project competitively. As a result, we were awarded the contract.

Rigging and Protection

Eleven towers of scaffolding including one double section were attached to the mortar joints
of the house using a minimum number of masonry bolts. These were red plastic pressure an-
chors installed in 'A-in.-diameter holes drilled 1 in. deep with two anchors used per tower. A
temporary chain link fence was installed in order to keep visiting architecture buffs and the
general public at a safe distance during the project.
Before the actual cleaning could commence, certain components of this historic structure had
to be protected. The window panes of French antique glass (considered irreplaceable) were
coated with a strippable acrylic mask to protect the glass from the caustic alkaline gel which
would be used. Glazing tape was then applied around the entire perimeter of each window
frame. Finally, particle board was nailed over the windows to further protect them from the
chemical spray and from any falling tools or debris. Where the window frames meet the stone
return, we left a V2-in. reveal to allow for complete access in cleaning the granite.
After the building was completely scaffolded and protected, John Vinci, the architect, and
Elaine Harrington, the curator of Glessner House, conducted an inspection to examine some
previously inaccessible areas and to determine the exact extent of the work to be done. During
this survey, several interesting discoveries came to light. It became apparent that pieces of slate
had been used as shims to help stabilize the granite blocks while they were being leveled. This
was a standard nineteenth century masonry technique. Another discovery was that the rosy-red
toned mortar in use throughout the rest of the house had not been used in the arch over the
entry. Gray mortar was used in a row of decorative ball and bead carved granite. Gray was also
used to blend in and unite the separate stones of the inner arch. All these mortar colors would be
carefully matched prior to the tuck-pointing application.

Grinding of tlie Mortar Joints

With the scope of the project now defined, we began the actual work. Initially we agreed to
chisel out all the mortar joints and then give the building a preliminary tuck-pointing with
standard Type N mortar and then finish tuck-pointing with a pink beaded joint of Type N
mortar. The mortar joints were to be chiseled out to a 2-in. depth and then tuck-pointed to
within about '/2 in. of the surface. This preliminary tuck-pointing would stabilize the masonry
and insure that none of the chemicals would be water blasted into the interior of the building
during the cleaning process. As the chiseling began, we found that the process was painstak-
ingly slow and not economically feasible. Therefore it was agreed that we would perform a sam-
ple cutting of the joints with electric grinders and, upon acceptance of this procedure, grind
rather than chisel all the mortar joints. This procedure was in fact permitted, although it did
not fall within the strictest interpretation of historical restoration. The grinding proceeded on
schedule.
As the tuck-pointers worked over the building at close range, they carefully saved iron ivy
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
254 MASONRY

nails presumably installed under Mrs. Glessner's direction for training ivy on the facade of the
house. They also collected bits of original building materials found near the foundation below
grade.

Preliminary Tnclt-pointiiig of the Mortar Joints

After the joints were ground out to a 2-in. depth, they were then tuck-pointed to within Vs in.
of the surface with Type N natural color mortar. Sample mortar mix had been made and sample
pointing of the deep mortar had been done for inspection by the architect. Type N mortar was
selected to match the original specification which stated;
The mortar used in the construction of the building above the ground level, unless otherwise specified,
will be composed of the best lime mixed with the best Utica cement, and clean sharp sand in the
following proportions: one barrel cement, one barrel lime, and the due proportion of sand. Below the
ground level the mortar used will be composed of two barrels cement, one barrel lime, and the due
proportion of clean, sharp sand. None but the best quality of lime equal to the best Stearns lime, and
Utica cement, and clean, sharp sand, free from extraneous matter, shall be used in the work.
The joints were struck with a key down the center in order to assure a good bond with the pink
finish mortar which would be applied after cleaning.

Cleaning of the Granite

Upon completion of the preliminary tuck-pointing, the building was completely sealed so that
the cleaning process could begin. Prior to the start of the project, even before the scaffolding
had been erected, we had applied various cleaning chemicals to sample areas of the stone and
had chosen alkaline gel manufactured by Prosoco because it works slowly on granite and poses
no threat to the limestone sidewalk surrounding the house, the terra-cotta roof, or the brick
inner courtyard and walls where they abut the granite.
The sample applications also determined that the method of application would involve two
steps. The alkaline gel would be scrubbed onto the building and allowed to set for 24 h then
pressure washed with 600-lb water pressure. After the gel was rinsed off, an acid-based after-
wash was applied to the building in order to remove any residue of the alkaline prewash, the
actual cleaning agent. The afterwash required no setting time and was immediately rinsed from
the building. This two-step process was done twice to thoroughly clean the granite. The alkaline
gel was removed by pressure washing starting from the bottom of the building and working
upward toward the roof.
Although the dwell time for the alkaline gel was increased from 24 to 48 h on some heavily
soiled areas because of the temperature difference between the August test samples and the
actual application in October, the cleaning process resulted in virtually 100% removal of soot
and grime from the surface of the granite. Split granite samples were used to check the efficacy
of the cleaning process. When the cleaning was completed, what had been a dark stone building
having the appearance of a fortress (Fig. 2) was transformed into a light gray residence with
glittering mica chips and particles of pink stone dispersed throughout the granite facades
(Fig. 3).

Repidr of the Chimneys

The chimneys had deteriorated to the point where plants were growing from some of the
chimney tops and in some sections the mortar was completely missing, allowing our crew mem-
bers to peer through the joints. In one chimney approximately a dozen stones were removed and
replaced; then the entire chimney was tuck-pointed as were the others which we repaired.
During this phase of the operation, we were allowed to walk on the roof as long as we accepted

FIG. 2—Entry door/decorative arch.

FIG. 3—Entry arch after cleaning.

responsibility for any damage and took every precaution to prevent damage. The terra-cotta
tiles are thin, and some have come loose with age. In order to protect these fragile tiles, special
ladders that were padded with foam and supported with wood struts from the topmost struc-
tural tile of the building were designed. The peak tiles were just strong enough to bear this
weight. Chimney repair involved many less man-hours than cleaning but was completed over a
longer time schedule because of the required phases and difficult access to the work area.

Pink Beading of the Mortar Joints

After the cleaning and restoration of the chimneys, we were to begin the bead tuck-pointing
of the joints in the original pink color for the majority of the building and in gray over the front
entry. Two areas of the house had been cleaned but not tuck-pointed: the two servants' porches,
one on the first floor, and the loggia on the second floor. The mortar in these areas had been
protected from the weather and so was left intact for historical purposes. Accurate mortar color
was determined by checking the joints on these porches.
The pink mortar was applied with a 3/i6-in. bead. Tools for this size beading had to be hand-
made by a metalsmith as the standard beading tool of today creates too large a bead. The bead-
ing process was delicate and time-consuming since the joints had to be tuck-pointed and then
struck several times in order to bring the cement content of the mortar to the surface. This
procedure gave a very smooth and uniformly colored joint.

BrealcdoMrn of Scaffolding and Landscape Repairs

All of the window protection and swing scaffolding were removed from the building upon
completion of the final tuck-pointing. The pipe scaffolding was then removed, and the lawns
were resodded. As a final touch, some of the mortar joints required spot tuck-pointing where
the scaffolds had been anchored to the building. This was easily done by removing the plastic
anchors, grouting in the holes, striking the fresh mortar with a beading tool to blend in with the
finished joints. The project which we had begun on August 24 was now completed just before
Thanksgiving.
Glessner House, the last surviving example of H. H. Richardson's four Chicago buildings,
now celebrates its 100th year with granite facades free from the last century's grime and fortified
by this century's restoration techniques.

Acknowledgments
I would like to gratefully acknowledge the editorial assistance of Jacquelyn Scruggs in prepar-
ing this paper.

Bibliography
"John J. Glessner House," pamphlet. Commission on Chicago Historical and Architectural Landmarks,
Chicago, IL, 1984.
Hitchcock, H. R., The Architecture of H. H. Richardson and His Times, Archon Books, Hamden, CT,
1936 revised 1961, pp. 277-278, 290-294, Figs. 105-106.
New Life for Old Buildings, M. F. Schmertz, Ed., McGraw-Hill, New York, 1982, pp. vi-vii.
Van Rensselaer, M. G., Henry Hobson Richardson and His Works, Houghton Mifflin, 1888, reprinted
Dover Publications, New York, 1969, pp. 109-110.

Masonry Cracks: A Review of the

Literature

REFERENCE: Grimm, C. T., "Maioniy Cracks: A Review of the Literature," Masonry: Materi-
als, Design, Construction, and Maintenance, ASTMSTP992, H. A. Harris, Ed., American Soci-
ety for Testing and Materials, Philadelphia, 1988, pp. 257-280.

ABSTRACT: Masonry surface cracks are objectionable because they are the primary source of
water permeance and may be aesthetically displeasing or indicative of structural distress. Cracks
are the most frequent cause for masonry's failure to perform as intended. The types, locations,
patterns, sizes, and causes of cracks are discussed. Methods are described for their prevention and
repair.

KEY WORDS: block (concrete), bricks, corrosion, cracks, expansion, failure, inspection, ma-
sonry, mortar, movement (structural), repair, sealants, shrinkage, strain

Cracking is probably the most frequent cause of masonry performance failure [30] and has
been an engineering concern for at least the last 150 years [J, 7,10,26,36,54,58,70.71,88,91.92,
97,111.112]. More recently, cracked masonry has generated litigation [107-110]. Cracks are
caused by movement (strain), which can not be prevented but can be accommodated. Thus,
cracks can be eliminated or made so small as to be unobjectionable.
A crack is here defined as a break, split, fracture, fissure, separation, cleavage, or elongated
narrow opening visible to the normal human eye and extending from the surface and into a
masonry unit, mortar joint, interface between a masonry unit and adjacent mortar joint, or into
the joint between masonry and an adjacent construction element.

Crack Classification
Masonry cracks may be classified by: (1) structure type; (2) masonry type; (3) location; (4)
pattern; (5) width; and (6) cause. For a given type of structure and material, the location, pat-
tern, and width often provide clues to the cause. Cracks result from strain which induces stress
in excess of strength in compression, tension, or shear. Strain may be induced by the imposition
of loads or by restraint of volume changes in the masonry materials. Volume changes include
those induced by change in temperature, moisture, water or salt crystallization, or corrosion.
Loads may be imposed by movements of foundations, structural frames, shelf angles, roof
slabs, spreading of pitched roofs, wood expansion, or retaining wall deflection. Cracks may also
be caused by vibration, blasting, and fire.
Types of masonry structures include: (1) arches and shells; (2) fireplaces and chimneys; (3)
floors and pavements; (4) revetments and channels; (5) beams and slabs; (6) bearing walls and
columns, and (7) nonbearing walls. Types of cracked masonry units include: (1) brick; (2) con-
crete masonry units (CMU); (3) stone; and (4) terra cotta. Cracks are located: (1) in masonry

'Consulting architectural engineer and senior lecturer in architectural engineering. University of Texas at
Austin, Austin, TX 78712.

257

units; (2) in mortar joints; (3) between units and mortar; and (4) continuously through units and
mortar. Crack patterns are: (1) horizontal, (2) vertical, or (3) diagonal. If vertical, they may be
straight or cogged, and, if diagonal, they may be straight or stepped.
Cracks smaller in width than 0.1 mm are insignificant to water permeance because that is the
least width through which wind-driven rain will enter [i2]. The maximum width of a crack that
will neither impact appearance nor cause alarm is said to be 0.25 to 0.38 mm [14]. Bidwell [11]
classified cracks as fine (up to 1.5 mm), medium (1.5 to 10 mm), and wide (above 10 mm).
Rainer [81] classified crack width as very slight (less than 1 mm), slight (1 to 5 mm), moderate
(5 to 15 mm), and severe (more than 15 mm). Kaminetzky [53] classified crack widths as negli-
gible (less than 0.1 mm), very light (0.4 mm), light (0.8 to 3.2 mm), moderate (3.2 to 12.7 mm),
extensive (12.7 to 25.4 mm), or very extensive (more than 25.4 mm). The term "hairline crack"
often appears in engineering literature to describe vaguely a narrow fracture. The thickness of
human hair varies with race, age, body location, circularity (measurement axis), fiber length,
relative humidity, and tensile stress [111]. Among teen-aged caucasoids for hair from the occip-
ital area of the scalp, the mean diameter is 70 jim with a standard deviation of 20 /xm {X = 0.07
mm, s = 0.02 mm). If the data are normally distributed, probably only 1 of 15 "hairline"
cracks would not admit wind driven rain, that is, diameter <0.1 mm.

Cracking Strain
The theory of fracture mechanics has been applied to masonry by Shrive [94]. He observed
that for a crack to be visible, the surfaces of the newly formed crack must separate, indicating
the previous existence of tensile stresses. Thus, it is tension which causes cracks, whether the
loads are compressive, shear, or tensile.
Compression induces tension transverse to the axial force, which may cause splitting (see Fig.
1). The strain required to cause brick masonry to crack in compression occurs at about half the
ultimate strength [39]. Frequent application and withdrawal of load may cause fatigue and
strength reduction and, therefore, increased cracking probability. As few as 40 cycles of com-
pressive load is said to cause a 30% reduction in static strength [1].
Shear induces diagonal tension. The shear load at first visible crack in concrete masonry
walls was measured at about 64% of failure load with a coefficient of variation of 25% for 15
walls [69,89,90]. Mayes et al. [63] found visible cracks in concrete masonry shear panels at
stresses exceeding 50 psi (345 kPa). Schneider [87] found first crack at an average of 64% of the
ultimate shear strength for 29 concrete masonry piers. Meli [66] found the ratio of shear load at
first crack to ultimate load on various types of masonry walls to be 0.83 with a coefficient of
variation of 20% for 19 specimens.
The first crack in brick masonry in flexure occurs at about 80% of failure strength [46].
Lawrence and Morgan [59] found that the flexural stress in masonry at first crack may be esti-
mated at 30% of the ultimate strength plus 29 psi (200 kPa). The tensile strain at rupture for
concrete masonry walls is about 0.021% [67]. The effect of specimen size and test method on
flexural strength is described in Ref 41.

Cement Shrinkage
Mortar, grout, concrete masonry units, and reinforced concrete shrink upon drying. When
excessive shrinkage is restrained, cracks result. Shrinkage of materials made with portland ce-
ment is caused by water loss and by carbonation. Water loss shrinkage is reversible. Carbon-
ation shrinkage is not. As CMU dry the recession of water in capillaries creates surface tension,
which places the material in compression and thus reduces volume. Because sand and gravel are
stiffer than lightweight aggregates, CMU made with such aggregates have greater shrinkage.
Carbonation is primarily a reaction between calcium hydroxides [Ca(OH)2], released by hydra-
tion of cement, and carbon dioxide (CO2) from the air to produce calcium carbonate [CaCOj]
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GRIMM ON MASONRY CRACKS 259

utile Split
Longitudinal Vertical Tenalla -V
V/

^^^-"-^^-"-^''^TEr^rrrr^' ^-^^-^-^
PLAN
Restraint
Vertical Bow-

Parapet-
-Roof Line

/ Longitudinal Horizontal Tenella Crack

Wall

Grade ' JSAha jdte

ELEVATION

FIG. 1—Restrained parapet: vertical bow or longitudinal split.

-Small Diagonal Tan»iQn CracHt

In Vertical Array

No Shelf Angle -
-Shelf Angia with Eipunsion Joint 'No Expansion Joint

o
u 'j ' ^ S h a l f Angia Diaeontinuous at Cornar -

I \

ELEVATION

FIG. 2—Differential vertical movement at corners due to discontinuity of shelf angle.

and water [ 93]. Carbon dioxide may also react with other cement paste components. Lime also
carbonates and therefore shrinks. The physiochemistry of carbonation is discussed by Kroone
and Blakey [57], Powers [78], Verbeck [103], and Kamimura [51].

Mortar Shrinkage
Unrestrained, 28-day shrinkage of mortar specimens cast in metal molds is said to have a
mean value of about 650 X 10"'% with a standard deviation of 150 X 10"''% [42]. The ulti-
mate (23 year), unrestrained, mortar shrinkage is estimated to have a mean of 1800 X 10~^%
with a standard deviation of 420 X 10"''% [40]. The horizontal shrinkage of mortar in bed
joints in masonry is restrained by shear with the masonry units, which Ritchie [83] found re-
duces mortar shrinkage. From those data it is estimated that the mean effect of restraint is to
reduce mortar shrinkage by about 30% with a standard deviation of 22%, when adjusted for
sample size. Anderegg [5] found that mortar left in contact with brick had shrinkage about
50% less than that of mortar cast in metal molds. Mortar shrinkage is also discussed in Refs 44,
55. and 58.
Mortar shrinkage increases with water-cement ratio, which increases with lime content
[28,33.49,65,73,83,106]. Voss [104] found that some mortars made with dolomitic hydrate
shrink more than some made with high calcium putty. As mortar sand fineness increases, water
demand to provide workability increases and shrinkage increases [13]. Thus, Anderegg [6] and
Conner [ 18] found that mortar joint cracking increased with sand fineness. Masonry sands are
also discussed in Ref 95. Increased air content also increases shrinkage [13].
The use of calcium chloride as a mortar additive can cause cracking by accelerating corrosion
of metals in contact with mortar and by substantially increasing drying shrinkage of mortar
[19]. Mortar cracks may also be due to weathering or sulfate attack [32,84]. McBumey [64]
describes expansion due to delayed hydration of magnesium oxide in mortar as causing severe
cracking of masonry.

Shrinkage of Concrete Masonry Units

Baker and Jessop [9] reviewed the literature on CMU shrinkage. Shrinkage of CMU in-
creases as density is reduced and water absorption of the concrete is increased [50]. Shrinkage
of low pressure cured CMU is about 85% greater than for high pressure cured units [16]. As
with all products, availability of autoclaved CMU should be determined prior to specification.
Table I provides statistical data on CMU shrinkage based on data given in Ref. 16.
When CMU, which had been previously dried to equilibrium with low humidity, were subse-
quently sprayed on one face for 2 h to simulate rain or painted with water cement paint, they
reexpanded approximately one third of their original shrinkage from a saturated condition
[86]. ASTM specifications provide for maximum moisture content of CMU ranging from 25 to

TABLE I—CMU shrinkage, 10'"%.

Low Pressure High Pressure
Probability
of Being All Sand & Light Sand & Light
Statistic Exceeded, % CMU Gravel Weight Pumice Gravel Weight Pumice

Low characteristic 95 130 162 254 400 138 131

Mode 260 269 400 596 192 215 373
Mean 330 270 420 630 190 230 390
High characteristic 5 600 338 610 838 238 323 481

45% depending on CMU shrinkage and average annual relative humidity at the construction
site [8].

Concrete Masoniy Wall Shrinkage

Total prevention of cracks in concrete masonry is said to be technically and economically
unfeasible [20]. Shrinkage of mortar and of concrete masonry units in concrete masonry walls
results in wall shrinkage which is greater than CMU shrinkage, perhaps 30% greater [86].
Total concrete masonry wall shrinkage may range from 0.01 % to 0.1 % [ 14]. More precise data
is given in Table 2.
A rationale for recommending a maximum ratio of wall length to height was provided by
Copeland [21]. Wall shrinkage is restrained at the wall base by bond and friction but may not
be restrained at the top of the wall. For walls of excessive length to height ratio, this phenome-
non results in a vertical crack near the center of the wall. Such cracks are wider at the top and
tapered in width toward the wall base. Depending on the relative tensile strength of CMU and
mortar, the vertical crack may be cogged or relatively straight.
In the absence of excessive wall height, shrinkage cracks are of even width and occur at the
weakest part of the wall. With long walls, vertical cracks occur at the midpoint or at rather even
intervals. Cracks may be vertical or cogged and sometimes stepped, especially near wall ends.
Vertical shrinkage cracks are common at reentrant corners [22,25,96,100].
Rainer [81] reports that wall cracks are most likely to occur at changes in wall dimension; at
comers, openings, pilasters, or other wall stiffener; and in areas of greater exposure as in para-
pets, wing walls, or fences.
The shrinkage cracking of concrete masonry walls is reduced by using two-core rather than
three-core CMU and the use of mortars having higher bond strength [67]. An increase of 10%
in crack resistance is reported. Concrete masonry units should be kept covered and dry during
transportation and job site storage until installed in the wall to minimize in-the-wall shrinkage
[79]. Weaker mortars, because of their greater extensibility (lower modulus of elasticity and
greater creep in tension), accommodate CMU shrinkage to a greater extent than stronger mor-
tars [86]. Expansion joints are usually not required in CMU walls because shrinkage normally
exceeds expansion (see Ref 2).

Sealant Joints
Sealant joints are sometimes called "movement joints." Three types of such joints are used
for crack control in masonry: expansion joints close to accommodate expansion of brick or
stone masonry; control joints open to accommodate shrinkage of concrete masonry; construc-
tion joints seal the crack between masonry and other materials, such as windows and doors. To
avoid considerable confusion, the terms expansion joint, control Joint, and construction joint

TABLE 2—Unrestrained shrinkage of concrete masonry, 10 ''%.

Cure

Aggregate Type High Pressure Low Pressure

Sand and gravel 90 to 130 230

Cinders 120 400 to 450
Expanded slag 340
Expanded shale 310

• *

I
1
t-

>
Ui

I
i

should not be used interchangeably. Some types of control joints arefilledwith mortar, can not
close, and can not be substituted for expansion joints without causing significant cracking. The
design of sealant joints is discussed in Ref 61 and 75.

Crack Control in CMU Walk

Methods for controlling cracks in concrete masonry walls include: (1) limitation on length to
height ratio of walls; (2) limitations on horizontal distance between vertical control joints; (3)
installation of bed joint reinforcement or bond beams; (4) location of control joints at points of
stress concentration; (5) control of moisture content in CMU at time of construction; and (6)
installation of slip joints [29,62].
The maximum spacing of vertical control joints is a function of type of CMU, wall height, and
spacing of bed joint reinforcement. The following recommendations for control joint spacings
[62] apply to walls built of ASTM Specification for Hollow Load-Bearing Concrete Masonry
Units (C 90), Type I (moisture controlled) CMU [8]. Bed joint reinforcement is two No. 9 cold
drawn, steel wires, one in each shell bed. With no bed joint reinforcement, the maximum con-
trol joint spacing is 18 ft (5.5 m) in exterior walls in climates where the average annual relative
humidity is between 50 and 75%. That control joint spacing is increased 33% to 24 ft (7.3 m) for
16-in. (400-mm) spacing of bed joint reinforcement and increased 67% to 30 ft (9.1 m) for 8-in.
(200-mm) spacing of bed joint reinforcement. Those control joint spacings are increased 25%
for interior walls. All of those control joint spacings are increased 6 ft (1.8 m) in dry climates,
where average annual relative humidity is less than 50% and reduced by 6 ft (1.8 m) in humid
climates, where average annual relative humidity exceeds 75%.
In addition, control joints are required at critical points of high stress concentration, that is,
at changes in wall height or thickness, above joints in floors or foundations, and below joints in
slab roofs bearing on the wall, at one or both sides of wall openings, at a distance from wall
intersections or comers not greater than one half the allowable spacing of control joints, and in
composite walls at the same location as expansion joints in the brick masonry. In lieu of a
control joint at each jamb of a wall opening, bed joint reinforcement may be placed in the first
and second joints immediately above and below wall openings, extending at least 2 ft beyond the
opening.
In large wall expanses, bond beams may be used in lieu of bed joint reinforcement. Trough or
U-shaped CMU are used in a continuous horizontal course, which is filled with concrete and in
which one or more reinforcing bars are placed. For two No. 9 bars the vertical spacing of bond
beams is four times the spacing of the replaced bed joint reinforcement. If used, bond beams
should be placed at the base and top of a wall and below windows [62].
Slip joints are horizontal planes of weakness formed by breaking the bond of mortar bed
joints with the CMU. Slip joints are placed at the top exterior corners of walls that support cast-
in-place concrete roofs or floor slabs and at CMU lintel bearings, where a control joint is located
above the jamb at a wall opening.

Cracks in Brick
The plane of cracks in the face of unglazed, extruded brick may be perpendicular to the face
and parallel to the direction of extrusion (for example, vertical for brick layed as a stretcher)
and are usually located at a core in the brick. They may penetrate only slightly or may extend to
the core. Such cracks are typically 2 mm (0.08 or V64 in.) or less in width. They are caused by
inadequate quality control in the brick manufacture process. ASTM Specification for Facing
Brick (Solid Masonry Units Made from Clay or Shale) (C 216) limits facial cracks in face brick
to those which may be seen at a distance of 15 ft (4.6 m) or 20 ft (6.1 m), that is, crack widths of
about 1.5 mm (0.06 in.). Cracks in brick that are parallel to the face and to the direction of
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
264 MASONRY

extrusion are called laminations and are not visible on the surface. All extruded brick are lami-
nated to some extent or other. Although laminations are of some concern to ceramic engineers,
there is no evidence that they affect brick performance [85]. Weathering or excessive compres-
sive stress applied at the face of brick may cause it to spall [40] (see Fig. 4).

Brick-Mortar Compatibility
Cracks in mortar may be due to differential movement between brick and mortar. For exam-
ple, if the coefficient of thermal expansion of brick is greater than that of mortar, vertical cracks
may occur in horizontal bed joints.

-Concrete Slab

Brick

Anchor

FIG. 4—Nonbearing wait.

Hedstrom et al. [45] measured the modulus of elasticity of several mortars at 90% of tensile
strength. The mean value was 2.87 X 10* psi. If the tensile bond strength of mortar to brick is
about 75 psi, the maximum unit strain in the mortar is 75/2.87 X 10* or 26 X 10"''%, which is
about one eighth of that found by Menzel [67]. In any event, if the 28-day shrinkage of re-
strained mortar is 230 X 10"''%, to avoid a cracked head joint the differential strain must be
compensated for by brick expansion. Irreversible moisture expansion of brick at 28 days has a
mean value of 63 X 10"'% [42]. A 7.63-in. (190-mm)-long brick expanding at that rate would
produce a strain in a Vs-in. (10-mm) mortar head joint of 1280 X 10"''%, that is, some 5.6
times greater than that required to compensate for mortar shrinkage.
Since the most likely mortar shrinkage transverse to the plane of a mortar head joint is much
less than the most likely brick moisture expansion, a shrinkage crack in the mortar head joint
most likely will not occur in brick masonry. Palmer and Parsons [ 74] found no evidence that
volume changes in mortar subsequent to hardening destroyed or weakened mortar bond when
extent of bond was good. When water retentivity of mortar was not compatible with the suction
rate of brick, mortar volume change was disruptive. High shrinkage mortar could be combined
with low moisture expansion brick, in which case the bond strength and extensibility of the
mortar becomes important to crack avoidance.
Mortar shrinkage and brick expansion are additive in bed joints. If that differential strain is
290 X 10"''% and the mortar modulus of rigidity is 0.4 X 2.87 X 10' psi, the estimated shear
stress is 332 psi. For ASTM C 270 Type N mortars, Nuss, Noland, and Chinn [ 72] found a mean
28-day shear strength of brick masonry to be 394 psi. When the IRA of brick was reduced by
wetting, shear strengths were much higher. It was also higher when Types M and S mortar were
used. Mortar shrinkage should not cause cracking of well-bonded mortar bed joints in brick
masonry.

Facial Separation Cracl^s

Facial separation cracks are openings in the wall face between brick and mortar, usually
1 mm (0.04 in.) or less in width. They are most frequently caused by inadequate tooling of
mortar joints during construction but may also be caused by thermal contraction of masonry
units and mortar and less frequently by mortar shrinkage. The effect of facial separation cracks
was discussed by Conner [17]. He found that the average cracking in the brick masonry wall of
44 buildings which had no wall leaks was 14.7% (that is, 14.7 ft of crack/100 ft of mortar joint),
which compared with 36.3% in 34 buildings which did leak.
Flexure induces cracks between brick and mortar rather than in mortar, because bond
strength is inevitably lower than the tensile strength of mortar. Horizontal cracks between brick
and mortar bed joints may be induced by shear (Figs. 3 and 5) or flexure (Fig. 4).

Bricli Masonry Expansion

Expansion of masonry may be caused by heat, moisture, or freezing. Such elongation may
cause oversailing of upper portions of a wall over lower portions (see Fig. 5), diagonal shear
cracks (see Figs. 3 and 8), bowing of walls (See Fig. 1, 4, 9, and 13), and flexural cracking at
corners in a vertical straight or cogged pattern (Fig. 11). Restrained longitudinal expansion may
cause delamination and buckling of pavements and bowing of parapets (Figs. 1 and 13).
Cracking due to expansion can be controlled in curtain walls by placing horizontal expansion
joints under shelf angles and vertically at appropriate horizontal intervals. For Vs-in.
(lO-mm)-wide joints and sealants having 50% compressibility, expansion joints in brick ma-
sonry should be spaced at horizontal intervals of about 20 ft (6.1 m) [42]. In addition, expan-
sion joints should be located at the same critical points of high stress concentration as in CMU
walls. Expansion joints in parapets should occur twice as frequently as those in enclosing walls,
unless the parapet is reinforced [30].
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
266 MASONRY

I
I

}
o

I Soil Hea ve

Tapered Flexural Crack

FIG. 6—Deep beam flexure—corner heave and/or midwall settlement or supporting beam deflection

Floor Line Foundation Settlement

Wall Opening

FIG. 7—Foundation settlement.

A wall above grade is more prone to volume change than a foundation wall. If the two are
rigidly anchored together, the restraint often induces cracks (Fig. 5).
Differential movement will also occur between dissimilar materials in a wall, for example,
face brick bonded to CMU in a composite wall [3S\. If the brick expansion is 0.03% and the
concrete masonry contraction 0.03%, the differential is 0.06%. If the modulus of rigidity of the
masonry is 500 000 psi, the first approximation of the shear stress in the masonry is 300 psi.
Differential vertical movement between dissimilar materials can cause parapets to lean in (see
Fig. 12).
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
268 MASONRY

Dcflaction

Wall Exoansion

Di*plac*d W*ll-
Displaced Jamb i i
TTTTr, 1
i^rX^,

^ Wal Opening
Wall
> \

' I ' ' « WWv

FIG. 8—Shear crack.

SECTION SECTION

Uniform
_ Horizontal
Flexural Crack

EXT INT

Uniform
Horizontal
Flexural Crack

FIG. 9—Buckled or bowed wall.

Wall Expansion

Horizontal Flexural
Crack at Piar Head

Horizontal Flexural
Crack at Pier Base

FIG. 10—Flexural crack at pier head and base.

i
1
VSV/////////////////////////K\-\
V i r t i c a l Straight
Flexural Crack Cogged Flexural Crack

ELEVATION

FIG. 11—flexural crack at corner.

Parapet Leans in Due to

Vertical Expansion of
Bricit Face

Cracl(
Roof

Brick Face

Grade

SECTION

Parapet Bows out

Expansion at Corner Due to
Horizontal Expansion

PLAN

FIG. 12—Leaning parapet.

Cracks can occur in bearing wall structures due to differential strain at the intersection of
bearing and nonbearing walls. Horizontal cracks at wall midheight may be due to flexure
caused by excessive deflection or bowing due to unaccommodated, differential, vertical move-
ment (Fig. 4). Such cracks are also caused by excessive deflection of masonry veneer over flexi-
ble steel studs designed for a maximum deflection of L/360 or L/600. A single wythe of brick
masonry cracks at a flexural deflection of about L/2000 [15]. The horizontal expansions of
wood floors has caused bulging and cracks in brick veneer on wood frame buildings.
Masonry having a lower modulus of elasticity has higher strain capacity and is, therefore, less
likely to crack. Therefore, the lowest adequate strength of masonry should be used. The lowest
strength mortar which is strong enough should always be used.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GRIMM ON MASONRY CRACKS 271

Fonndatloiu
Foundation movement may be caused by uneven settlement, moisture movement in plastic
soils, or downhill creep of surface layers. Settlement is caused by soil consolidation, shear fail-
ure, and variable soil types. Clays and silts increase in volume with increased moisture content
and decrease in volume with reduced moisture. Water content of soil changes with season, trees
and shrubs, localized watering and heat, and moisture migration. When ground water reduces
the shear strength of sloped soil, downward slides can result [96], Masonry pavements on
chalky or fine sandy soils may be subject to upheaval due to ice lensing during severe winters
[23], Building on permafrost is a special case with peculiar problems. In coal mining regions
ground subsidence may cause a surface wave 2 or 3 ft (0.61 or 0,91 m) high to pass slowly
through entire communities [27,76]. The influence of trees on house foundations in clay soils is
discussed in a publication of the British Building Research Station 1101]. The settlement of
foundation for masonry walls is discussed by Komomik and Mayurik [56].
Cracks that result from uneven settlement of foundations may take any form, but they are
most often diagonal or vertical and are usually tapered [96] (see Figures 6, 7, 14, 15, 16, and

-Bowed Parapet

7/yy/j ryi
--_:::::w///^:
PLAN
Restraint

Parapet-
S

Vertical Tensile Crack

Wall

jm JML
Grade •

ELEVATION

FIG. 13—Restrained parapet buckled laterally.

Tapered Flexural Crack

Deflection

Soil Heave

FIG. 14—Deep beam flexure—corner settlement and/or midwall heave.

>t ^" ^
Wat. ,
Opening

Deflection ' • • . _ _ ! * I' . __ —

*
Soil Heave Foundation Settlement
f Soil Heave

FIG. 15—Deep beam flexure with wall openings—comer heave and/or midwall settlement or supporting
beam deflection.

Deflection''

FIG. 16—Deep beam flexure with wall opening—comer settlement or differential column contraction.

Deflection

FIG. M—Deep beam flexure with wall opening—comer settlement or differential column contraction.

17). Vertical cracks wider at the top than at the base indicate flexure, sometimes due to founda-
tion movement (see Fig. 14).

Frame Movement
Wind loads [4] cause structural frames to sway. Frame drift may induce racking of masonry
walls supported on frames (Fig. 18). Concrete columns are subject to thermal, creep, shrinkage,
and elastic deformations, which usually are different for each column in the same building. The
average contractions of concrete columns in highrise buildings is reported by the Portland Ce-
ment Association to be about 0.032% for shrinkage, 0.04% for creep, and 0.042% for elastic
deformation, but extreme values may be 65% greater [14]. Unaccommodated differential
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
274 MASONRY

FIG. 18—Shear in nonbearing wall due to frame side sway (drift).

movement between adjacent columns will crack masonry walls attached to the frame (see Fig.
16). Thompson and Johnson [102] report differential vertical movements of Vz in. in 15 ft of
wall length as insufficient to cause cracking (0.0028L), but others have suggested limiting such
movement to 0.0014L. Such movement, when excessive, often causes diagonal cracks extending
to and from the comers of wall openings [88].
Horizontal cracks near floor level may indicate excessive floor, beam, or slab deflection.
Walls supported on beams having excessive deflection may crack horizontally anywhere, verti-
cally near midspan, and diagonally near span third points [14,96]. Masonry supported on a
steel spandrel beam may crack due to torsional rotation of the spandrel beam [52].
The net expansion of brick masonry due to freezing, moisture, and heat, less restraint and
mortar shrinkage, is estimated to have a mean value of about 0.03% but one chance in 20 of
being as much as about 0.07% [42]. When brick masonry is anchored to a concrete frame, the
differential vertical movement may average 0.14% and could be as much as 0.26%, that is, 3.12
in. (79 mm) in 100 ft. (30.5 m). If that magnitude of differential vertical movement is not ac-
commodated by horizontal expansion joints between the masonry and the frame, cracking will
result. A restrained movement of only 0.01 in. (0.25 mm) in 20 ft (6.1 m) can produce a bulge of
1 in. (25 mm) [22].

Cracks at Shelf Angles

In the absence of a horizontal expansion joint under the shelf angle, which supports masonry
on a structural frame, differential vertical movement between the masonry and the frame pro-
duces a load on the shelf angle, which may be sufficient to lift the angle from its wedge insert
anchor, to yield the angle [43], or to cause spalling of brick [ 77,98,99]. Mortar in the joint at
the toe of a shelf angle can cause spalling of brick at the shelf angle [ 77,98,99] (see Fig. 4). The
deflection of shelf angles should be carefully controlled to insure that the expansion joint under
the shelf angle does not close excessively [43]. Total shelf angle deflection should not exceed Vi6
in. [30].
Vertical cracks frequently occur at comers of walls supported on shelf angles, when the shelf

angle is not continuous around the corner, leaving the masonry at the comer continuous verti-
cally past the shelf angle (see Fig. 2).

Cracks at Roofs
Horizontal cracks at corners near concrete roof slabs may be due to slab curl caused by differ-
ential shrinkage between the top and bottom of the slab [24,48] (see Fig. 3). Roof movement
can cause diagonal cracks in masonry walls parallel to the roof movement (see Fig. 5) and hori-
zontal cracks in masonry perpendicular to the roof movement [ 96]. Roof movement may be due
to concrete shrinkage or thermal movement in steel roofs. Horizontal cracks near eaves may
indicate lateral movement of pitched roofs, vaults, or shells. Dimensional change of wood plates
rigidly anchored to masonry walls may cause masonry cracks. The holes in such plates through
which the anchor bolts pass should be larger than the bolts, and the anchor nut should be
tightened only by hand.

Vibration-Induced Craclcs
Most building vibrations generated internally are caused by machines (cranes, elevators,
fans, pumps, and punching presses) or by people (walking, jumping, running, dancing). Exter-
nally generated vibrations are commonly caused by road or rail traffic, subways, sonic booms,
strong wind, earthquake, blasting, excavation, soil compaction, or pile driving [80]. Relatively
small vibration may add to built-up stress concentrations and lead to unexpected masonry
cracks even when vibration levels are within recommended limits [80]. In tall buildings wind-
induced vibrations can lead to cladding cracks. Dowding and Corser [31] describe cracks
caused by blasting due to: (1) vibration of the structure or its foundation; (2) impact of flying
rock; (3) permanent ground distortion; and (4) air blast [31].

Othet Craclc Causes

Cracks in chimneys may be caused by sudden and wide temperature changes or by the freez-
ing of condensate from the combination of natural gas. Severe fire causes cracking and bulging
of masonry as well as surface spalling or possibly vitrification of clay brick. Although severe
damage to masonry may be caused by earthquakes, well designed and built masonry may be
crack free after imposition of significant seismic loads.
When steel corrodes, the ferric oxide occupies more than twice the volume of steel from which
it was formed [43]. Corrosion of imbedded reinforcing steel may cause a crack at the wall sur-
face along the length of the steel. In walls, horizontal cracks at regularly spaced vertical inter-
vals may be due to corrosion of bed joint reinforcement or wall ties.

Cracli Inspection
Although no absolute determination as to the cause of masonry cracking can be made solely
on the basis of visual observation, cause clues are readily obtained. What to notice about cracks
[32]: (1) direction (pattern); (2) extent (where it begins and ends); (3) width (uniform or tapered
and if so how); (4) depth (through the paint, the plaster, and the wall); (5) alignment (in plane
or laterally offset); (6) edge sharpness (rough, rounded, or broken edges may be indicative of
compression failure); (7) cleanliness (new cracks have clean sides, not coated with paint, dirt, or
algae); and (8) crack dynamics (static or changing in size, shape, or direction).
Information on the date of crack occurrence is suspect because a crack is very seldom noticed
at first unless its formation is accompanied by a loud noise. Hearing a noise and then finding a

crack is not uncommon, but a cause-effect relationship is seldom justified [32]. Crack width
may be gauged by use of the Avongard Calibrated Crack Monitor (2836 Osage, Waukegan, II).
Under about 15 foot candles of illumination visual acuity is about 0.5 min of arc, that is,
under that illumination, a crack can be seen at a distance up to about 6900 times its width [47],
for example, a crack width of 0.1 mm can be seen at a distance of about 2 ft-3 in. (960 mm).
The frequency with which masonry should be inspected for cracks varies from one to five
years [105]. The legal liability assumed by architects and engineers who inspect building fa-
cades has caused considerable concern [60].

Repair
Tests made at the Building Research Station in England have shown that the capacity of 9-in.
(229-mm)-thick brick walls to carry vertical loads is reduced no more than 30% by a stepped or
slanted crack up to 1 in. (25 mm) wide, provided that the damage is not accompanied by consid-
erable transverse movement [82]. If a wall is out of plumb not more than 1 in. (25 mm) or bulges
no more than V2 in. (12 mm) in a normal story height, no repair would usually be needed on
structural grounds alone [82].
Crack repair methods may be classified as those which do not significantly change wall ap-
pearance and those which do. Fine cracks [less than V16 in. (1.5 mm)] are not very conspicuous
and in brick masonry would often be made more unsightly by repointing [82]. Such cracks can
be filled by surface grouting, which will prevent objectionable water permeance and not greatly
change wall appearance, if the masonry surface texture is relatively smooth. Clear coatings for
masonry typically do not bridge cracks and, therefore, do not prevent water permeance. Crack
repair methods for masonry are discussed in Ref. 34, 37, and 38.

References
[1] Abrams, D. P., Noland, J. L., and Atkinson, R. H., "Response of Clay-Unit Masonry to Repeated
Compressive Forces," Proceedings of the 7th International Bricks Masonry Conference, Brick Devel-
opment Research Institute, University of Melbourne, Melbourne, Australia, February 1985, p. 565.
[2] Allan, W. D. M., "Shrinkage Measurements of Concrete Masonry," Journal of the American Con-
crete Institute. Detroit, MI, Vol. 26, No. 6, 1930, pp. 699-713.
[3] Ameny, P. and Jessop, E. L,, "Masonry Cladding: A Report On Causes and Effects of Failures,"
Proceedings of the Seventh International Brick Masonry Conference, University of Melbourne, Mel-
bourne, Australia, February 1985, p. 261.
[4] American National Standard Minimum Design Loads for Buildings and Other Structures, ANSI
A58.1—1982, American National Standards Institute, Inc., New York, NY, 1982.
[5] Anderegg, F. O., "Some Properties of Mortars in Masonry," Proceedings, American Society for
Testing and Materials, Philadelphia, Vol. 40, 1940, p. 1134.
[6] Anderegg, F. O. and Anderegg, J. A., "Some Volume Changes in Mortar and Concrete," ASTM
Bulletin, American Society for Testing and Materials, Philadelphia, December 1955, pp. 60-63.
[ 7] Anderson, G. W., "The Design of Brickwork for Differential Movement," Techniques. No. 6, Brick
Development Research Institute, Parkville, Victoria, Australia, January 1979.
[8] 1985 Annual Book of Standards, American Society for Testing and Materials, Philadelphia, Vol.
4.05, 1985.
[9] Baker, L. R. and Jessop, E. L., "Moisture Movement in Concrete Masonry," InternationalJoumal
of Masonry Construction. London, England, Vol. 2, No. 2, 1982, pp.-75-80.
[ 10] Baker, M. C , "Introduction to the Problem of Cracks, Movement, and Joints in Buildings," Cracks,
Movement, and Joints in Buildings, Division of Building Research, National Research Council of
Canada, Ottawa, Ontario, August 1972.
[11] Bidwell, T. G., The Conservation of Brick Buildings, The Repair, Alteration, and Restoration of Old
Brickwork, Brick Development Association, London, England, August 1977, p. 6.
[12] Birkeland, O. and Sevendsen, S. D., "Norwegian Test Methods for Rain Penetration through Ma-
sonry Walls," Symposium on Masonry Testing, STP No. 320, American Society for Testing and
Materials, Philadelphia, February 1963, pp. 3-15.
[13] Bloem, D., "Effects of Aggregate Grading on Properties of Masonry Mortar," Symposium on Ma-
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
GRIMM ON MASONRY CRACKS 277

sonry Testing, ASTM STP 320, American Society for Testing and Materials, Philadelphia, June
1962, pp. 67-91.
[14] Building Movement Joints, Portland Cement Association, Skokie, IL, 1982, p. 30.
[15] Compressive, Transverse, and Racking Strength Tests of Four-inch Brick Walls, Structural Clay
Products Research Foundations Research Report No. 9, Brick Institute of America, Reston, VA,
August 1965, p. 17.
[16] Concrete Masonry Shrinkage, National Concrete Masonry Association, Herndon, VA, 1961.
[17] Conner, C. C , "Factors in the Resistance of Brick Masonry Walls to Moisture Penetration," Pro-
ceedings of the American Society for Testing and Materials, ASTM, Philadelphia, Vol. 48, 1948, pp.
1-35.
[18] Conner, C. C , "Some Effects of the Grading of Sand on Masonry Mortar," Proceedings of the
American Society for Testing and Materials, ASTM, Philadelphia, Vol. 53, 1953, pp. 933-945.
[19] Control by Cracking in Concrete Structures, ACI 224 R-80, American Concrete Institute, Detroit,
MI. July 1985, p. 13.
[20] Copeland, R. E., "Procedures for Controlling Cracking in Concrete Masonry," Concrete Products,
Chicago, IL, Vol. 67, No. 9, September 1964, pp. 48-52.
[21] Copeland, R. E., "Shrinkage and Temperature Stresses in Musonry," Journal of the American Con-
crete Institute. ACI, Detroit, MI, Vol. 53, 1957, pp. 769-780.
[22] "Cracking of Concrete Masonry—Causes and Suggested Remedies," Concrete, Vol. 56, No. 4, Chi-
cago, IL, April 1948, p. 6ff.
[23] "Cracking in Buildings," Building Research Digest, No. 75, Building Research Station, Garston,
Waterford, England, October 1966.
[24] "Cracking Tendencies in Brick or Stone Masonry Walls at the Structural Slab," Journal of The
American Concrete Institute, Detroit, MI, January 1947, Vol. 18, No. 5, pp. 606-608.
[25] Cracks Control in Concrete Masonry Unit Construction, Federal Construction Council Technical
Report No. 48, National Academy of Sciences, Washington, DC, 1964.
[26] Cracks, Movements, and Joints in Buildings, Division of Building Research, National Research
Council of Canada, Ottawa, September 1976.
[27] Crawford, C. B., "Deformation Due to Foundation Movements," Cracks, Movement, and Joints in
Buildings, Division of Building Research Council of Canada, Ottawa, Ontario, August 1972.
[28] Davis, R. E. and Troxell, G. E., "Volumetric Changes in Portland Cement Mortars and Concrete,"
Proceedings of the American Society of Civil Engineers, Philadelphia, Vol. 25, 1929, pp. 210-260.
[29] "Design of Concrete Masonry for Crack Control," NCMA-TEK No. 53, National Concrete Masonry
Association, Herndon, VA, 1973.
[30] "Differential Movement," Technical Notes on Brick Construction, No. 18, Bricks Institute of Amer-
ica, Reston, VA, April 1963.
[31] Dowding, C. H. and Corser, P. G., "Cracking and Construction B\sisting," Journal of The Construc-
tion Division, American Society of Civil Engineers, New York, NY, March 1981, pp. 89ff.
[32] Eldridge, H. J., Common Defects in Buildings, Her Majesty's Stationery Office, London, England,
1974, p. 85ff.
[33] Evans, D. N. et al., "Properties of Some Masonry Cement," Journal of Research of the National
Bureau of Standards, Research Paper 2427, Washington, DC, Vol. 51, No. 1, July 1953, pp. 11-16.
[34] Filler, J. D. and Kriegh, K. D., "A Guide to Pressure Grouting Cracked Concrete and Masonry
Structures with Epoxy Resins, National Technical Information Service, Springfield, VA, February
1973, AD-755-926.
[35] Fowler, D. W. and Grimm, C. T., "Differential Movement in Composite Load Bearing Masonry
Walls," Journal of the Structural Division, Proceedings of the American Society of Civil Engineers,
Vol. 105, No. ST 7, New York, NY, July 1979, pp. 1277-1288.
[36] Fricki, K. E. et al., "Problems in Masonry Walls—A Case Study," Proceedings of the First North
American Masonry Conference, University of Colorado, Boulder, CO, August 1978, p. 113-1.
[3T] Grimm, C. T., "Masonry Maintenance and Restoration—A Guide to the Literature," Structural
Renovation and Rehabilitation of Buildings, Boston Society of Civil Engineers, Section/ASCE, Bos-
ton, MA, Nov. 1979, pp. 71-90.
[38] Grimm, C.T., "WaterPermeanceof Masonry Walls—A Review of the Literature, "Afaionry.Ma^e-
rials. Properties, and Performance. ASTM STP 778, American Society for Testing and Materials,
Philadelphia, 1982, pp. 178-199.
[39] Grimm, C. T. and Fok, C.-P., "Brick Masonry Compressive Strength at First Crack," Masonry
International, University of Edinburgh, Edinburgh, Scotland, Vol. 1, No. 2, July 1984, pp. 18-23.
[40] Grimm, C. T., "Durability of Brick Masonry—A Review of the Literature," Masonry: Research,
Application, and Problems, ASTM STP 871. American Society for Testing and Materials, Philadel-
phia, PA, 1985, pp. 202-234.
[41] Grimm, C. T., "Flexural Strength of Masonry Prisms vs. Wall Panels," Journal of the Structural
Division. American Society of Civil Engineers, New York, NY, September 1985, pp. 2021-2032.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
278 MASONRY

[42] Grimm, C. T., "Probabilistic Design of Expansion Joints in Brick Masonry," Proceedings of the 4th
Canadian Masonry Symposium. University of New Brunswick, Fredericton, NB, June 1986.
[43] Grimm, C. T. and Yura, I. A., "Shelf \ng\ei lor Masomy Weneer," Journal of the Structural Divi-
sion, American Society of Civil Engineers, New York, NY, in review.
[44] Hansen, T. C , "Effect of Wind on Creep and Drying Shrinkage of Hardened Cement Mortar and
Concrete," Materials Research and Standards, American Society for Testing and Materials, Phila-
delphia, January 1966, pp. 16-19.
[45] Hedstrom, R. O., Litvin, A., and Hanson, J. A., "Influence of Mortar and Block Properties on
Shrinkage Cracking of Masonry Walls," Journal of the PCA Research and Development Laborato-
ries, Portland Cement Association, Skokie, XL, January 1968.
[46] Hendry, A. W. and Kheir, A. M. A., "The Lateral Strength of Certain Brickwork Panels," Proceed-
ing of the Fourth International Brick Masonry Conference, Groupement National de I'lndustrie da la
Terre Cuite, Brussels, Belgium, April 1976, p. 4. a. 3.1-4.
[47] lES Lighting Handbook, Illuminating Engineering Society, New York, NY, 1959, p. 2-9.
[48] "Job Problems and Practice," Journal of The American Concrete Institute, Detroit, MI, Vol. 43,
January 1947, pp. 606-608.
[49] Johnson, H. V., "Cement-Lime Mortars," Technologic Papers of the Bureau of Standards, Superin-
tendent of Documents, U.S. Government Printing Office, Washington, DC, Jan. 29, 1926, Vol. 20,
No. 308, pp. 256.
[50] Kalouseb, G. L., Relation of Shrinkage to Moisture Content in Concrete Masonry Units, Housing
Research Paper No. 25, Housing and Home Finance Agency, Superintendent of Documents, U.S.
Government Printing Office, Washington, DC, March 1953, p. 4.
[5/] Kamimura, K., et al., "Changes in Weight and Dimensions in the Drying and Carbonization of
Portland Cement Mortars,"Afagazj'neo/CoBcrefe/iesearcA, Vol. 17, No. 10, March 1965, pp. S-14.
(521 Kaminetzky, D., "Preventing Cracks in Masonry Vf a.lls," Architectural Record, Vol. 136, November
1%4, pp. 210-214.
[53] Kaminetzky, D., "Verification of Structural Meqancy," Rehabilitation, Renovation, and Preserva-
tion of Concrete and Masonry Structures, American Concrete Institute, Detroit, MI, 1985, p. 141.
[54] Keller, H. and Suter, G. T., "Concrete Masonry Veneer Distress A Case Study," Proceedings of the
Third North American Masonry Conference, University of Texas at Arlington, Arlington, TX, June
1985, p. 3.
[551 Kessler, D. W. and Anderson, R. E., Studies in Stone Setting Mortar, BMS Report 139, National
Bureau of Standards, Superintendent of Documents, U.S. Government Printing Office, Washing-
ton, DC, Nov. 23, 1953, p. 23.
[56] Komornik, A. and Mayurik, A., "Restrained Settlement of Masonry Buildings," Proceedings of the
International Conference of Soil Mechanics and Foundation Engineering, Japanese Society of Soil
Mechanics and Foundation Engineers, Tokyo, 1977 and 1978, VI, pp. 613-618.
[571 Kroone, B. and Blakey, F. A., "Reaction Between Carbon Dioxide Gas and Mortar," ACIProceed-
ings, American Concrete Institute, Detroit, MI, Vol. 56, 1960, pp. 497-510.
[58] "Lack of Design Data, New Building Techniques Cause Facade Failures," Engineering News
Record, 2 Feb. 1978, p. 9.
[591 Lawrence, S. S. and Morgan, J. W., "Investigations of the Properties of Small Brickworks Panels in
Lateral Bending, Experimental Building Station, North Ryde, N.S.W., Australia, TR 52/75/418,
January 1975.
[601 Le Patner, B. B., "Caveat Architectus: Facade Inspections and the Design Professional," Architec-
tural Record, July 1981, p. 57.
[6/] Mansfield, G. A., Sirrine, C. A., and Wilk, B., "Control Joints Regulate Effects of Volume Change
in Concrete Masonry," Journal of the American Concrete Institute, Detroit, MI, July 1957, Vol. 54,
pp. 59-70.
[62] Masonry Structural Design for Buildings. TM 5-809-3/NAVFAC DM—2.9/AFM 88-3, Chap. 3,
Department of the Army, Navy, and the Air Force, Washington, DC, 16 Jan. 1985.
[63] Mayes, R.L.,YutaroOmote, andClough,R.W., Cyc/icS/ieBr Testso/Masonry fters, Vol. I.. Test
Results, National Technical Information Service, PB-264-424, Springfield, VA, May 1976, p. 77.
[64] McBumey, J. W., "Cracking in Masonry Caused by Expansion of Mortar," Proceedings of the
American Society for Testing and Materials, ASTM, Philadelphia, Vol. 52, 1952, pp. 1-20.
[651 Mears, A. R. and Hobbs, D. W., "The Effect of Mix Proportions Upon the Ultimate Air-Drying
Shrinkage of Mortars," Magazine of Concrete Research, Slough, England, Vol. 24, No. 79, June
1972, pp. 77-84.
[66] Meli, R., "Behavior of Masonry Walls Under Lateral Loads," Proceedings of the Fifth World Con-
ference on Earthquake Engineering, Rome, Earthquake Engineering Research Institute, El Cerrito,
CA, 1972.
[671 Menzel, C. A., "General Considerations of Cracking in Concrete Masonry Walls and Means for

Minimizing It," Development Department Bulletin D 20. Portland Cement Research and Develop-
ment Laboratories, Skokie, IL, September 1958, pp. 1-15.
\68] Monk, C. B., "Testing High-Bond Clay Masonry Assemblages," Symposium on Masonry Testing.
ASTM STP 320, American Society for Testing and Materials, Philadelphia, February 1963, pp.
31-66.
[69] Moss, P. J. and Scrivener, J. C , "Concrete Masonry Wall Panel Tests—The Effect of Cavity Filling
on Shear Behavior," New Zealand Concrete Construction, Concrete Publications, Private Bag, Pori-
rua Bay, NZ, April 1968.
[ 70] Myren, B. J., "Cracks in Housing Project VfMs," Engineering News Record, 17 Mar. 1938, p. 390.
[ 71] Niermann, T. H., "Cracks in Brick Walls," Engineering News Record, 5 May 1938, pp. 640-641.
[72] Nuss, Larry K., Noland, J. L., and Chinn, J., "The Parameters Influencing Shear Strength Between
Clay Masonry Units and Mortar," Proceedings of the (First) North American Masonry Conference,
University of Colorado, Boulder, CO, August 1978, p. 13.
[ 73] Palmer, L. A., "Volume Changes in Brick Masonry Materials," Research Paper No. 321, Journal of
Research, National Bureau of Standards, Superintendent of Documents, U.S. Government Printing
Office, Washington, DC, Vol. 6, June 1931, pp. 1003-1026.
[74] Palmer, L. A. and Parsons, D. A., "A Study of the Properties of Mortars and Bricks and Their
Relation to Bond," R.P. 683, Journal of Research, National Bureau of Standards, May 1934, Vol.
12, pp. 609-644.
175] Panek, J. R. and Cook, J. P., Construction Sealants and Adhesives, Wiley-Interscience, New York,
NY, 1984.
[76] Parate, N. S., "Some Observations on Masonry Structure Behavior Due to Ground Movement,"
Proceedings of the Second North American Masonry Conference, University of Maryland, College
Park, MD, August 1982, p. 31.
[ 77] Plewes, W. G., "Failure of Brick Facing on High-Rise Buildings," Canadian Building Digest, No.
185, April 1977, p. 3.
[78] Powers, T. C , "A Hypothesis on Carbonization Shrinkage," Journal of Portland Cement Research
and Development, Portland Cement Association, Skokie, IL, May 1962, pp. 41-49.
[ 79] "Prevention of Cracking," Rock Products, March 1942, Vol. 45, No. 3, March 1942, p. 79.
[80] Rainger, J. H., "Vibrations in Buildings," Canadian Building Digest, No. 232, National Research
Council of Canada, Ottawa, Ontario, Canada, May 1984.
[81] Rainer, P., Movement Control in the Fabric of Buildings, Nichols Publishing Co., New York, NY,
1983.
[82] "Repairing Brickwork," Building Research Station Digest, No. 4, Garston, Waterford, Hurts, En-
gland, 1960.
153) Ritchie, T., "Effect of Restraint on the Shrinkage of Masonry Mortars," Materials Research and
Standards, American Society for Testing and Materials, Philadelphia, Vol. 6, No. 1, January 1966,
pp. 13-16.
[84] Ritchie, T., "Influence of Lime in Mortar on the Expansion of Brick Masonry," Reaction Parame-
ters of Lime, ASTM STP 472, American Society for Testing and Materials, Philadelphia, 1969, pp.
67-81.
[85] Ritchie, T., "Measurement of Laminations in Brick," Bulletin, American Ceramic Society, Colum-
bus, OH, No. 9, Vol. 54, 1975, pp. 725-726.
[86] Russell, W. A., Shrinkage Characteristics of Concrete Masonry Walls, Housing and Home Finance
Agency, Housing Research Paper 34, Superintendent of Documents, U.S. Government Printing Of-
fice, April 1954.
[ 87] Schneider, R. R., Shear in Concrete Masonry Piers, Masonry Research of Los Angeles, Los Angeles,
CA, undated.
[88] Schubert, P. and Glitza, H., "Resistance to Cracking of Masonry Subjected to Vertical Deforma-
tion," Proceedings of the Fourth International Brick Masonry Conference, Bruge, Belgium, April
1976, p. 4.6.9.
[89] Scrivener, J. C , "Concrete Masonry Wall Panel Tests—Static Racking Tests with Predominant
Flexural Effect," New Zealand Concrete Construction, Concrete Publications, Private Bag, Porirua
Bay, Porirua Bay, New Zealand, July 1966.
[ 90] Scrivener, P. J., "Static Racking Tests on Concrete Masonry Walls," Proceeding of the First Interna-
tional Masonry Conference, Gulf Publishing Co., Houston, TX, May 1969, p. 185.
[91] Severud, F. N., "Avoiding Cracks in Building Walls," Engineering News Record, 6 July 1939, pp.
45-46.
[92] Shakir, A., "Failure of Masonry Structures," Proceedings of the Third North American Masonry
Conference, University of Texas at Arlington, Arlington, TX, June 1985, p. 21.
[93] Shideler, J. J., "Carbonation Shrinkage of Concrete Masonry Units," Journal of the Portland Ce-
ment Association Research and Development Laboratories, Skokie, IL, Vol. 5, No. 3, September
1963, pp. 36-51.
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
280 MASONRY

194] Shrive, N. G., "A Fundamental Approach to the Fracture of Masonry," Proceedings of the Third
Canadian Masonry Symposium, University of Alberta, Edmonton, Alberta, Canada, June 1983, pp.
4-1.
[95] Sinha, B. P. and Currie, D. W., "Survey of Scottish Sands and Their Characteristics Which Effect
the MoTtait Strength," IntemationalJoumal of Masonry Construction, United Trade Press, London,
England, Vol. 2, No. 1, 1981, pp. 2-12.
[96] Sorensen, C. P. and Tasker, H. E., Cracking in Brick and Block Masonry, Experimental Building
Station, Technical Study No. 43, Australian Government Publishing Service, 1976.
[97] Stevenson, J. H., "Cracking Walls," Engineering News Record, 31 Aug. 1939, p. 38.
[98] Stockbridge, I. G., "Cladding Failures—Lack of Professional Intetface," Journal of The Technical
Councils ofASCE, American Society of Civil Engineers, New York, NY, December 1979.
[ 991 Suter, G. T. and Hall, J. S., "How Safe Are Our Cladding Connections," Proceedings of The First
Canadian Masonry Symposium, University of Calgary, Calgary, Alberta, 1976, pp. 95-109.
[ 100] "The Avoidance of Cracking in Masonry Construction of Concrete and Sand-Lime Brick," Building
Research Digest, No. 6, Building Research Station, Gaston, Waterford, Herts, England, 1949, re-
vised 1957.
[101] "The Influence of Trees on House Foundations in Clay Soils," Digest, No. 298, Building Research
Station, Garston, Waterford, England, June 1985.
[102] Thompson, J. N. and Johnson, F. B., "Design for Crack Prevention", Insulated Masonry Cavity
Walb, National Academy of Sciences, Publication No. 793, Washington, DC, 1960.
[103] Verbeck, G., "Carbonization of Hydrated Portland Cement," Cement and Concrete, ASTM STP
205, American Society for Testing and Materials, Philadelphia, 1958, pp. 17-36.
[ 104] Voss, W. C , "Lime Characteristics and Their Effect on Construction," Symposium on Lime, Ameri-
can Society for Testing and Materials, Philadelphia, 1939, p. 14.
[705] "Wall Cladding Defects and Their Diagnosis," Building Research Establishment Digest, No. 217,
Building Research Station, Garston, Waterford, England, September 1978, p. 2.
[106] Watstein, D. and Seese, N. A., "Properties of Masonry Mortars of Several Compositions," ASTM
Bulletin, No. 147, American Society for Testing and Materials, Philadelphia, Aug. 1974, pp. 77-81.
(107] Society of Mt. Carmel v Fox, 31 IlAp3d 1060, 335 NEZd 588 (1975).
[108] Schreiner v Miller, 67 la 91, 24 NW 738 (1885).
(109] President and Directors of Georgetown College v Madden, 660 FZd 91 (4th Cir 1981).
[110] Society of Mt. Carmel v Fox & Fox, 90 IlAp3d 537, 413 NEZd 480 (1980).
[///] Robbins, C. R., Chemical and Physical Behavior of Human Hair, Van Nostrand ReinholdCo., New
York, NY, 1979, p. 179.
[112] Bartlett, W. H. C , "Experiments on the Expansion and Contraction of Building Stones by Varia-
tions of Temperature," American Journal of Science, first series. Vol. 22, 1830, pp. 136-140.

Author Index

Anderson, E. D., 18 Krogstad, N. V., 96, 118

Arnold, R. C , 118
M
B
Manmohan, D., 237
Bretnall, W. C , 57 Matthys, J. H., 193
Brown, R. H., 7 Monk, C. B., 96
Monk, C. B., Jr., 118

Chen, H.-L., 145

Chin, I. R., 96
Coney, W. B., 220 Raths, D. C , 41
Ribar, J. W., 33
Robinson, G. C , 7

Dehghani, A., 193

Dondanville, L. J., 118
Dubovoy, V. S., 33 Schwein, R. L., 237
Searls, C. L., 227
Shah, S. P., 145
Sourlis, T., 251
Gensert, R. M., 57 Stockbridge, J. G., 220
Grimm, C. T., 169, 257

H
Houston, J. T., 193 Thomasen, S. E., 227

W
Johal, L. S. P., 18 Wyllie, L. A.,Jr., 237

281
Copyright by ASTM
Copyright 1988 byInt'l
A S T(all rights reserved); Mon May
M International 19 10:24:50 EDT 2008
www.astm.org
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
STP922-EB/JUI. 1988

Subject Index
C 230-80, 199(ftn, table)
C 246, 206
Absorption {See also Initial rate of absorp- C 270, 8,42,44, 49,50,52(table), 193-194,
tion)
199, 220
brick C 270-82 32
boiling water, 186{fig) C 270-84 (also, UBC 24-20), 7-17, 21,
cold-water 18S(figs)
initial rate, 11, 186-187(figs) 22(ftn, tables), 148(table)
statistical methods for testing, 182-192 C 476-83, 21
C 1072, 43, 45, 49, 51, 52(tables)
tests on terra-cotta, 235 C 1072-86,8
ACI (See American Concrete Institute) E 447-84, 23, 148(table)
Acrylic clear coating, 224 E 514, 43, 44(fig), 45, 46, 51-52, 53(fig),
Adhesion, terra cotta, 235
Air content 54-55, 220-221, 234
Avongard calibrated crack monitor, 276
extended plastic life mortar, 196
Air content of mortar
influence on bond strength, 9-ll(fig),
22(tables), 32, 33 B
Air-cured cube strengths BIA {See Brick Institute of America)
cement only extended plastic life mortars, Binary operations, 179, 182
200{table) Biological growth as a cause of glaze spalling,
Air content 231-232(fig)
extended plastic life mortar, 196 Bisque—failure between glaze, 228
Algae, 231 Bisque faults, 229, 231
Alkaline gel, 251 Block (concrete), literature review of masonry
Alkyl trialkoxy silane clear coating, 224 cracks, 257
Aluminum stearate clear coating, 223 Block pier specimens—materials, 20, 22(ta-
American Concrete Institute bles)
ACI 318, Section 11.6—flexural shear and BOCA (See Building Official Conference of
torsional shear, 120-121 America)
ASTM Special technical publications Boiling water absorption—brick, 186(fig)
STP 589, 1 Bond, masonry, 33, 41-56
STP 778, 1 Bond strength
STP 871, 1 brick and mortar, 7, 8-9(figs), 14-15(figs),
ASTM Standards 33-34, 46-47, 52
C 67, 43, 46(table), 51, 52(table), 201 Bond wrench testing techniques, 33, 45(fig),
C 67-83, 169 46
C 67-85, 147 Brick
C 90, 263 masonry, 41, 42-43(figs), 169
C109, 8 properties
C109-86, 147 applied statistical methods, 185-192
C 110, 199(ftn, table) (figs)
C 140-75, 21 Brick and mortar properties
C 144, 45, 223 bond strength, 7, 8(fig). 33-34
C 144-84, 149(table) boiling water absorption, 186(fig)
C 150, 222 cold-water absorption, 185(figs)
C 185-80, 199(ftn, table) compressive strength, 151(table), 187(fig)
C 207, 222-223 expansion, 189-190
C 216, 42, 46, 220, 263 initial rate of absorption (IRA), 11, 12(fig),
C 216-85a, 147 186-187(figs)

283
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
284 MASONRY

Brick and mortar properties (cont.) Compression

pier specimens, materials, 20-21, 22(ta- brick masonry prisms, 186-187(figs)
bles) cracks, 63-67(figs), 158(fig), 186(figs)
prism, 45(fig), 147, 151(table) cube, extended plastic life mortar, 197-
tests, 147 198(figs)
Brick Institute of America (BIA) literature review on masonry cracks, 258
shear reinforcement and torsional shearing prism testing, 205(figs), 206, 207-210(figs)
stresses, 120-121 Compressive load as cause of tensile splitting,
Standard: Recommended practice for engi- 62(fig)
neered brick masonry, 120 Compressive strength
Technical note 288, 96-97 brick masonry, 100, 169, 186-187
Brick National Testing Program, 97-98 concrete masonry block, 21
Brick prisms mortar, 7-8(fig), 12-13, 33,100,151(table)
bond wrench test, 33, 45(fig), 46 terra-cotta, 237-243, 249
compressive strengths, lOO(fig) Compressive stresses of terra-cotta in-situ
Brick spandrel panel, prefabricated, 118-145 evaluation, 237-246
Brick suction, log-normal fit, 184(table) Compressive tests, 145-147
Brick surface terture, 34-35(table, figs) Computer analysis, 103, 107, 116
Brick testing and evaluation, 42-44, 47- Concrete masonry, 21, 41, 58
55(figs, tables), 97-98 Concrete masonry shrinkage
Brick veneer construction, 41, 96, 103- literature review on cracks, 260
109(figs), 116 Construction materials
Bricks specifications and standards, 167
literature review on masonry cracks, 263- Construction procedures
267 preconstruction testing program, 41
Brittle behavior, 28 Construction Technology Laboratories(CTL)
BV/MS construction (See Brick veneer) research on masonry mortars, 33
Building Official Conference of America Cope, 46, 50-54
(BOCA code), 120 Corrosion
Bumham and Root's Rookery Building, Chi- extended plastic life mortar, 214, 218-
cago, 251 219(figs)
literature review on masonry cracks, 257
masonry buildings, 57-95
Corrosion of embedded metals
terra-cotta glaze spalling, 230
Carbonation shrinkage, Crack classification, literature review on ma-
literature review on masonry cracks, 258 sonry cracks, 257
Cavity wall, 41 Crack repair methods, literature review on
Cement content in extended plastic life mor- masonry cracks, 276-280
tars, 199, 200(table) Cracking
Cement shrinkage, literature review on ma- and spalling, 232(fig)
sonry cracks, 258 corrosion induced, 91-95 (discussion)
Chemicals, environmental masonry buildings, 28, S7-95(figs)
terra-cotta glaze spalling, 231 Cracks, masonry
Chimneys compression, 63-67(figs), 69, 74-75(figs),
historic preservation and restoration repair, 158, 169, 186
254, 256 literature review, 258-280
literature review on masonry cracks, 275 repair methods, 276-280
Cladding Crazing, 229, 237
terra-cotta glaze spalling, 227, 230, 237 CTL (See Construction Technology Laborato-
Clay brick masonry units ries)
extended plastic life mortar testing, 201 Curtain walls, 60-61(figs), 63-90(figs), dis-
Clay-masonry prisms, 34 cussion, 91-95
Coatings—waterproofing, 220 Cyclic loading, 23-24, 145-147
Cold water absorption of brick, 185(figs) masonry piers, 23-24
Combined wind and gravity force, 118-144 slow-rate, 152
Compound lever loading system, 118 tests, 145-147(fig), 152, 156, 157(figs)
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
SUBJECT INDEX 285

Deflections, 59(table), 70(fig), 140-143 Fabricated masonry assemblies for precon-

Density functions, probability struction testing, 41
statistical methods applied to masonry, Facade replacement program, 118
184-192(figs) Facing brick
Design (See Seismic design) performance standards, 42-43
Deterioration, 92, 169, 227 Failure, 28, 153
Differential movement, 59-60 Failure
Displacements cracks, shear, 67, 75(figs) literature review on masonry cracks, 257
Distress, masonry buildings, 57-95 FEM {See Finite method analysis)
Distribution, statistical methods applied to Field survey of glaze spalling, 233-234(fig)
masonry, 169-192 (equations, figs, Field tests for water infiltration of glaze spall-
tables) ing, 234(fig)
Dynamic analysis, tests of model masonry sin- Finite method analysis(FEM), 122
gle-pier, 145-147 Flexural bond stress
Dynamic response, 158-163(figs, tables) masonry walls, 97-98(fig), 107
Dynamic shaking test, 152, 155(figs), 161 ordinary vs computer analysis, 109, 112
(fig). 162 (fig), 116
Flexural bond test
extended plastic life mortar, 202, 203(fig)
Flexural shear, 120
E Flexural strength of brick masonry, 169, 187
Flexural tensile strength of masonry bonds,
Earthquake resistance, 145-147 233-234(figs)
Efflorescence, 211, 215(figs) Flexural tension of brick, 52, 98(fig)
Elasticity modulus, 189(fig), 206, 243 Flow test, extended plastic life mortars,
brick veneer testing program, 99 195(fig)
End-boundary condition, 96 Foundation wall, literature review on masonry
Envelope curves, 147 cracks, 267-271(figs)
Envelope of hysteresis, 147, 156 Fracture mechanics, literature review on ma-
Environmental chemicals sonry cracks, 258
terra-cotta glaze barrier, 231 Freeze/thaw
EPL (See Extended plastic life) damage to terra-cotta, 233
Epoxy coating damage to terra-cotta, 233 deterioration, 92, 169
Expansion, brick resistance, 211, 214-215(figs, table)
elasticity and compressive strength, 189(fig) tests on brick and mortar, 46
freezing, 189(fig) Freezing expansion of brick, 189(fig)
literature review on masonry cracks, 265 Frequency distribution, 170(table)
moisture, 169, 190(fig) Full-scale testing, 118-144
thermal, 190(fig)
Expansion joints on curtain walls, 60-61(figs)
Extended plastic life (EPL) mortars
G
air content, 196
air-cured cube strengths, 200(table) Glaze spalling, terra-cotta
cement content, 199, 200(table) adhesion, 235
clay brick masonry units, 201 biological growth, 232(fig)
compression cube, 197-198(figs) epoxy coatings, 233
flexural bond test, 202, 203(fig) failure between bisque, 228-229(figs)
flow test, 195(fig) preventive measures, 235
masonry units, 202(table) tests, 235
performance criteria, 193-194 Glessner House, John J.
physical properties, 202 history, 251
strength retrogression, 198 restoration, 251-256
suction effect, 197(flg) Granite restoration, 253-254
testing, 195-198(figs), 199-200 Gravity and wind force, combined, 118-144
type M, N, S mortars, 199 Gravity loading, 120, 122
water retention, 196(fig) Grout conditions, 120
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
286 MASONRY

Grout mixture for pier testing, ASTM C476- cement-based mortar,18, 19(table), 34
83,21 cracks, literature review, 257-280
Grouting, 223 distress, 57-90—discussion, 91-95
literature review on cracks, 257-280
mortars, 33, 148-149{tables)
H performance failure, literature review on
Histagram—sampling, 170 cracks, 257
Historic structure performance standards, 37, 41
preservation and restoration, 251-256 piers, 18, 23-24, 145, 147
Hystereses loops, 147, 156 prisms, 21-23 (tables)
properties
coefficients of variation, 171(table)
vs reinforced concrete masonry, 58(table)
Inertia force, 153 vs structural steel, 58(table)
Initial rate of absorption—brick IRA, 11,12- restoration, 251
13(figs), 33, 186-187 shrinkage, 206, 211-217
In-situ evaluation of compressive stress on standards, inadequacy of ASTM C 270, 7-
terra-cotta, 237-243 17
Inspection, literature review on masonry seismic design, 147
cracks, 257 terra-cotta, 227
Interaction, 57 testing and evaluation program, 21, 41
IRA {See Initial rate of absorption) Masonry assemblage shrinkage, 206
Masonry buildings, distress, 60-90{figs) 94-
95
Masonry units, extended plastic life mortars
Leakage, water (See Water leakage) physical properties, 202(table)
Lichen, 231 Masonry walls
Literature review on masonry cracks, 257-280 flexural bond stress of brick veneer/metal
Load stress distribution on brick spandrel stud, 97
panels, 126-131(figs) flexural strength of brick panels, 187(fig)
Load factors (See also Dynamic loading. Cy- performance, 41, 52(table), 97-99(fig, ta-
clic loading) ble)
deflection, 140-143{figs) preconstruction testing, brick veneer, 41,
failure, 153 52
rotation, 141(fig), 143-144(figs) Material properties, 97-98, 147, 148(table)
testing, 152-153 Material properties of brick, 149
Load in tie, 108(fig) Materials, strength, 97
Loading bents, elevation, 134-139(figs) Materials testing, 20, 22, 169-192
Loading, gravity, 120,122 Maximum and minimum properties
Loading mechanisms and procedures, test statistical methods for brick masonry sam-
plan, 132, 133(fig) pling and testing, 178-179
Loading test, 152(table), 156, 157(figs) Mean and variance, 170-175(equations, ta-
Log-normal distribution, 183-185(figs, tables)
bles) Mechanical properties of materials, 147
Metal Lath/Steel Framing Association (ML/
SFA), 96-97
M Metal studs, 103, 110-116(figs, tables)
Maintenance of terra-cotta, 233 Metal tie flexibility, 103
Malaysia Pavilion buildings, 96 ML/SFA (See Metal Lathe Steel Framing As-
Marshall Field Warehouse, Chicago, 251 sociation)
Masonry Model masonry piers
bond strength, 33-34 structural behavior, 145-147, 148(table)
brick and mortar, 220-224 test results, 156, 157(figs), 158-165
buildings Modulus of elasticity in compression, 99,189,
corrosion, 57-95 206, 243
curtain walls, 57-60 Model prisms
discussion, 91-95 stress-strain curves, 149(fig)
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
SUBJECT INDEX 287

Moisture expansion, 169 Performance standards of extended plastic

Moisture expansion, terra-cotta tests, 235 life mortars, testing, 193-199(figs)
Mortar Performance standard, mortar
batch properties, 199(table) inadequacy of ASTM C 270, 7-17
bed joint—evaluation testing, 42 Permeability of water, 44-46
bond strength, 7-9, 46-47, 52 Permeability of water in terra-cotta glaze, 231
brick prisms, 34, 147 Permeance testing, 222(fig), 223 (table), 224
compressive strength, 34, lOO(fig), 151 Physical properties of extended plastic life
(table) mortars, 202 {See also Mortars, Ex-
cyclic loading, 145, 149(table) tended plastic life mortars)
joints, preservation of Glessner house, 251- Pier specimens, 20-21(figs)
256 Pier testing, 153-155(figs)
literature review on masonry cracks, 258 Piers, masonry, 18-32, 145, 147
model masonry and model brick properties, Plastic life mortars, extended, 189-194
148-149(tables) Portland cement-based mortar, 18, 19(table)
performance standards, 37, 41 extended plastic life, 194
testing, 21, 41-56, 194 Preconstruction testing of masonry materials,
prisms, 45, 147, 151 41. 54-55
shaking, 145, 149(table) Prefabricated brick panels, 118
shrinkage, literature review, 260 Prefabricated panels, 62, 68-70
terra-cotta, effect of water infiltration, 235 Preservation, historic, 252
Mortar standards Prisms
inadequacy of ASTM C 270, 7-17 brick masonry compressive strength, 188-
Mortars 189(figs), 202, 205(figs), 219
cement-based, 18-19, 34 clay-masonry, 34
extended plastic life, 193-194 compressive strength, 100, lOl(figs), 219
masonry, 33, 148-149 masonry, 21-23(tables), 145-147
types M and S, 19(table), 21, 22(tables), 32, mortar-brick, 34, 45-46, 147, 151
97 Probability density functions, 171-185 (equa-
type N, 147—compared to extended plastic tions, figs, tables)
life mortar, 199 Profilometer, surface, 34
Moss, 231 Property specifications
Movement joints, 57-95 inadequacy of ASTM C 270, 7-17
Movement (structural)
in masonry buildings, 57-60, 94-95
literature review on masonry cracks, 257-
258, 259(figs), 273-274 Quasistatic cyclic loading, model masonry
pier, 145-165
N
National Historic Preservation Act of 1966,
252 Ready mix, 193
Reinforced concrete masonry properties com-
pared to masonry and structural steel,
58(table)
Reinforced structural masonry, 118-144
Ordinary conventional analysis Reliability, 182
strength of materials, 97 Repair, literature review on masonry cracks,
Ordinary vs computer analysis, flexural bond 257
stress, 109
Residual compressive stresses, 237
Resistance, 193, 211, 214-215
Restoration of terra-cotta glaze, 235, 250
Restoration—Glessner House, 251-256
Penetrometer test, extended plastic life mor- Retarded strength, 193
tar, 196(fig) Richardson, Henry Hobson—design archi-
Performance failure of masonry tect, 251
literature review on cracks, 257 Rigid connections, 57
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
288 MASONRY

Roofs, literature review on masonry cracks, Strength

275 compressive, 169
Rotation, 141, 143-144 flexural, 169
Strength of materials analysis, 99-102(figs),
116, 121
Strength retrogression, extended plastic life
Sampling, brick masonry, 169-177 (tables, mortar, 198
figs) Stress distribution, brick spandrel panels,
Sampling and testing of concrete masonry 124-131(figs)
units, 21 Stress due to stiff masonry, 57
Sarabond mortar additive, 91-95 Stress failures of terra-cotta, 237-240
Scaffolding for historic restoration, 256 Stress in mortar, 108(fig)
Scale factor, 145 STRESS program—computer model, 103
SCPRF (See Structural Clay Products Re- Stress relief, 243, 245-248(figs, tables), 249-
search Foundation) 250
Sealant joints, literature review on masonry Stress-strain curves of model prisms, 149(fig),
cracks, 261, 263 205-210(figs)
Seismic design testing, 147 Stress-strain data for terra-cotta, 244(fig)
Seismic zones, use of masonry cement, 18 Structural behavior, 147
Shaking, 147, 153-155(figs), 158-165 Structural Clay Products Research Founda-
Shear tion(SCPRF), 97
behavior of masonry, 18, 25-31(figs), 67, Structural steel properties
147 compared to masonry and reinforced con-
displacement cracks, 75(figs) crete masonry, 58(table)
literature review on masonry cracks, 258 Studs, computer analysis vs ordinary analysis,
reinforcement(BIA), 120 96-116
Sheer strength (See Shaar behavior of ma- Studs, metal, 103, 110-116
sonry) Suction effect, extended plastic life mortar,
Shear stress, 156, 158(fig) (See also Shear be- 197(fig)
havior of masonry) Surface profilometer, 34
Shelf angles, literature review on masonry Surface texture parameters of brick, 34
cracks, 274-275
Shrinkage, 206, 211, 212-217(figs)
Shrinkage, literature review on masonry
cracks, 258-261 Temperature changes, effects on masonry
Single piers, 147 buildings, 78(fig)
Slow-rate cyclic loading, 152 Tensile splitting due to compressive load, 62
Soft joint, 96 Tension due to temperature changes, 78(fig)
Spalling of terra-cotta, 227-228, 233, 237 Terra-cotta
Span-deflection ratios, 103, 111 cladding material, 237
Spandrel beam masonry piers, testing 23- compressive strength, 237-243, 249
25(figs) facade, 237
Spandrel panel, prefabricated brick analysis freeze/thaw damage, 233
and testing, 118, 119{fig) glaze spalling, 227-228
Specifications installation, 229-230
performance standards of masonry maintenance, 233
inadequacy of ASTM C 270, 7-17 moisture expansion, 235
Spring constant, 96 restoration, 227
Square panel, 145 spalling, 227-231
Standard deviation sampling, 171 strain relief tests, 237-243
Static behavior, 165 stress failure, 237-240
Statistical methods for brick masonry thermal coefficient tests, 235
sampling and testing, 169-192 thermal expansion, 229
Steel grout void conditions, 120, 121 water infiltration, 235
Stiffness degradation, 96, 147 Tension tests, 147
Strain in terra-cotta, 230, 237-243 Test results of mortar performance standards,
Strain measurements, 242, 243(table) 7
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
SUBJECT INDEX 289

Testing Vertical displacement, 156

brick masonry, statistical methods for sam- Vertical load tests, 120
pling and testing, 169-192 Vibration, literature review on masonry
concrete masonry units, ASTM (C 140-75), cracks, 275
21,41
extended plastic life mortars, 195-200
W
load factors, 152-153
model masonry, 147, 152(table) Wall height, 104-106(figs)
water permeability, 44-46, 53, 220-224 Wall panels, 187
Thermal coefficient tests on terra-cotta, 235Wall stress, 104-106(figs)
Thermal expansion, 190(fig), 229 Walls {See also Curtain walls)
Tie flexibility, 103, 116 brick veneer, 103-107(figs)
Torsional behavior of brick spandrel panel, concrete masonry, literature review on ma-
122(fig) sonry cracks, 261-274(figs)
Torsional sensitive, C-shaped cross section, exterior performance, 41
118 masonry buildings, 57, 108-109(figs), 116
Torsional shearing stresses, 120 Water and salt deposits, effect on terra-cotta,
Tuckpointing, 222-224, 252-254, 256 231, 233
Type M mortar, 21, 32, 99 Water infiltration of terra-cotta glaze, 235
Type N mortar, 147 Water leakage in mortar-brick combinations,
Type S mortar, 21, 32, 97, 99 I I , 12(figs), 13-15(figs), 16(table), 46-
Tjrpe M, N, S—extended plastic life mortars, 47, 50-53(figs, table)
199 Water leakage resistance, influence of air con-
tent and water flow, 16(table)
Water loss shrinkage, literature review on ma-
U
sonry cracks, 258
UBC (See Uniform Building Code) Water permeability test for brick, 44(fig), 46,
Uniform Building Code (UBC) 53(fig)
allowable shear stresses, 18, 32 Water permeance, 220-224, 258
chapter 30. sect. 304(b), 249 Water retention of extended plastic life mor-
Unique compound lever loading system, 118- tars, 196(fig)
144 Water retention of mortar, relation to bond
strength, 8-9(fig), 33
Water-to-cement ratio for strength develop-
ment, 9
Variance Waterproof coatings, 223-224
statistical method for brick masonry sam- Wind and gravity force, 118
pling and testing, 170-175 Wind frames, 70-72(figs)
Variation coefficient Wind load, 104-106(figs), 120, 122, 123(fig)
masonry properties, 171(table) Wind load stress distribution, 128-131(figs)
Veneer, brick, 41, 103 Window panel, 73(fig)

Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize
Copyright by ASTM Int'l (all rights reserved); Mon May 19 10:24:50 EDT 2008
Downloaded/printed by
Linda White (Barry Levin AFFILIATION Associates, INc) pursuant to License Agreement. No further reproductions authorize

Structural Masonry
91% (11)
Structural Masonry
309 pages
Everyday Details
100% (1)
Everyday Details
116 pages
Masonry Repair
100% (1)
Masonry Repair
19 pages
Masonry Design Guide
100% (10)
Masonry Design Guide
62 pages
A Practical Guide To Single Storey House Extensions
No ratings yet
A Practical Guide To Single Storey House Extensions
81 pages
Encyclopedia of Trusses
100% (1)
Encyclopedia of Trusses
52 pages
Strip Foundation Construction Guide
No ratings yet
Strip Foundation Construction Guide
8 pages
Builders Foundation Handbook
100% (12)
Builders Foundation Handbook
124 pages
BuildersGuideResidentialSteelFloors PDF
100% (2)
BuildersGuideResidentialSteelFloors PDF
104 pages
Masonry Construction Details Guide
100% (3)
Masonry Construction Details Guide
31 pages
Masonry Design Guide PDF
No ratings yet
Masonry Design Guide PDF
29 pages
Informguide Brick PDF
100% (1)
Informguide Brick PDF
8 pages
Residential Structural Design Guide: Second Edition
100% (12)
Residential Structural Design Guide: Second Edition
403 pages
Residential Building Construction PDF
No ratings yet
Residential Building Construction PDF
434 pages
Dincel Wall - Construction Manual - 2023
No ratings yet
Dincel Wall - Construction Manual - 2023
176 pages
Diagnosing Common Building Defects
100% (1)
Diagnosing Common Building Defects
51 pages
FLorida Code Flood Resistant Construction
No ratings yet
FLorida Code Flood Resistant Construction
9 pages
Residential Bearing Wall Removal: What Is A Bearing Wall and What Does It Do?
No ratings yet
Residential Bearing Wall Removal: What Is A Bearing Wall and What Does It Do?
7 pages
Masonry
100% (1)
Masonry
178 pages
Introduction To Masonry
100% (7)
Introduction To Masonry
4 pages
E 447 PDF
100% (1)
E 447 PDF
13 pages
ASCE - SE Exam Review - Masonry Design Slides
No ratings yet
ASCE - SE Exam Review - Masonry Design Slides
46 pages
Ce479: Design of Building Components & Systems FALL 2012 - J. LIU
No ratings yet
Ce479: Design of Building Components & Systems FALL 2012 - J. LIU
70 pages
Masonry Construction Materials Guide
No ratings yet
Masonry Construction Materials Guide
8 pages
Masonry Repair: Key Techniques and Issues
No ratings yet
Masonry Repair: Key Techniques and Issues
39 pages
A Good Place To Start: Tip 2: For ASD, Consider Using The 2011 MSJC Provisions
No ratings yet
A Good Place To Start: Tip 2: For ASD, Consider Using The 2011 MSJC Provisions
12 pages
Field-Testing-Guide of Mortar, CHB
No ratings yet
Field-Testing-Guide of Mortar, CHB
48 pages
Dutch Unreinforced Masonry Properties Analysis
No ratings yet
Dutch Unreinforced Masonry Properties Analysis
8 pages
Thin Layered Mortar Joint Masonry Study
No ratings yet
Thin Layered Mortar Joint Masonry Study
10 pages
Essential Terms in Brick Masonry
No ratings yet
Essential Terms in Brick Masonry
78 pages
Masonry Structures Behaviour and Design PDF
No ratings yet
Masonry Structures Behaviour and Design PDF
809 pages
Brick Testing Methods and Standards
No ratings yet
Brick Testing Methods and Standards
6 pages
Lecture2 PDF
No ratings yet
Lecture2 PDF
21 pages
Masonry: Essential Questions
No ratings yet
Masonry: Essential Questions
7 pages
Comprehensive Guide to Masonry Units
No ratings yet
Comprehensive Guide to Masonry Units
4 pages
Masonry Materials and Testing Overview
No ratings yet
Masonry Materials and Testing Overview
61 pages
2015 DORMS - Design of Reinforced Masonry Structures
No ratings yet
2015 DORMS - Design of Reinforced Masonry Structures
410 pages
Masonry Techniques and Materials Guide
No ratings yet
Masonry Techniques and Materials Guide
16 pages
Overview of Masonry Construction Techniques
No ratings yet
Overview of Masonry Construction Techniques
7 pages
Clay Brick Masonry Strength Analysis
No ratings yet
Clay Brick Masonry Strength Analysis
4 pages
Masonary Design Working Stress
100% (1)
Masonary Design Working Stress
21 pages
103 Makmak Destroyer
No ratings yet
103 Makmak Destroyer
240 pages
Theoretical Framework
No ratings yet
Theoretical Framework
7 pages
5 Rammed Earth
100% (1)
5 Rammed Earth
138 pages
Comprehensive Guide to Masonry Units
No ratings yet
Comprehensive Guide to Masonry Units
42 pages
ASTM C270 91a Mortar
No ratings yet
ASTM C270 91a Mortar
9 pages
Strength Behavior of Prism Masonry
No ratings yet
Strength Behavior of Prism Masonry
12 pages
Masonry Techniques and Tools Overview
No ratings yet
Masonry Techniques and Tools Overview
4 pages
TL 362 Scope Effective Aug 21 2023
No ratings yet
TL 362 Scope Effective Aug 21 2023
4 pages
Masonry Construction Manual
96% (47)
Masonry Construction Manual
390 pages
Acceptability of Masonry Gary Porter 1
No ratings yet
Acceptability of Masonry Gary Porter 1
2 pages
MASONRY
No ratings yet
MASONRY
4 pages
Appropriate Selection of Interior and Exterior Materials, Finishes, Products, Components
No ratings yet
Appropriate Selection of Interior and Exterior Materials, Finishes, Products, Components
77 pages
RILEM TC 127-MS: Masonry Testing Standards
No ratings yet
RILEM TC 127-MS: Masonry Testing Standards
5 pages
Standard Test Methods For Measurement of Masonary Flexural Bond Strength
No ratings yet
Standard Test Methods For Measurement of Masonary Flexural Bond Strength
19 pages
Ce 434-534 Masonry Class Notes
100% (1)
Ce 434-534 Masonry Class Notes
135 pages
Tek 14-02 (1997)
No ratings yet
Tek 14-02 (1997)
6 pages
CHAPTER 2.4 - Masonry and Finishings Handouts
No ratings yet
CHAPTER 2.4 - Masonry and Finishings Handouts
12 pages
Comprehensive Guide to Masonry Techniques
No ratings yet
Comprehensive Guide to Masonry Techniques
10 pages
FN Masonry
No ratings yet
FN Masonry
38 pages
The Essential Guide To Framing
100% (13)
The Essential Guide To Framing
121 pages
Residential Interior Design - A Guide To Planning Spaces
100% (38)
Residential Interior Design - A Guide To Planning Spaces
289 pages
Modern Construction Handbook
89% (55)
Modern Construction Handbook
498 pages
The Complete Guide To Decks
100% (40)
The Complete Guide To Decks
322 pages
Black & Decker The Complete Guide To Garages
97% (29)
Black & Decker The Complete Guide To Garages
241 pages
Detail For Wood Frame House Construction - Illustrated Handbook
94% (16)
Detail For Wood Frame House Construction - Illustrated Handbook
133 pages
Handbook of Building Construction
93% (41)
Handbook of Building Construction
507 pages
Black & Decker The Complete Guide To Bathrooms
92% (26)
Black & Decker The Complete Guide To Bathrooms
289 pages
Black & Decker The Complete Guide To Attics & Basements PDF
100% (6)
Black & Decker The Complete Guide To Attics & Basements PDF
242 pages
Construction Cost Estimating (2021, Routledge)
100% (25)
Construction Cost Estimating (2021, Routledge)
526 pages
Framing Basics PDF
91% (11)
Framing Basics PDF
130 pages
101 Tiny House Design Concepts
94% (16)
101 Tiny House Design Concepts
425 pages
Plumbing Skills
100% (11)
Plumbing Skills
131 pages
Black & Decker Heres How Painting
100% (24)
Black & Decker Heres How Painting
98 pages
The Complete Guide To Bui
98% (42)
The Complete Guide To Bui
257 pages
Interior Graphic and Design Standards
100% (13)
Interior Graphic and Design Standards
634 pages
Black & Decker The Complete Outdoor Builder
100% (11)
Black & Decker The Complete Outdoor Builder
529 pages
Black & Decker The Complete Guide To Landscape Projects
97% (34)
Black & Decker The Complete Guide To Landscape Projects
241 pages
The Complete Guide To Custom Shelves & Built-Ins
100% (32)
The Complete Guide To Custom Shelves & Built-Ins
242 pages
Patios & Walkways PDF
100% (11)
Patios & Walkways PDF
257 pages
Tiny House Design & Construction Guide
95% (22)
Tiny House Design & Construction Guide
145 pages
Black and Decker - The Book of Home Improvement (2017)
93% (14)
Black and Decker - The Book of Home Improvement (2017)
563 pages
Black & Decker The Complete Guide To Contemporary Sheds
100% (35)
Black & Decker The Complete Guide To Contemporary Sheds
242 pages
Black & Decker The Complete Photo Guide To Home Repair
100% (7)
Black & Decker The Complete Photo Guide To Home Repair
560 pages
House Plans
89% (27)
House Plans
64 pages
Residential Building Construction
100% (21)
Residential Building Construction
434 pages
Black & Decker Complete Home Improvement With 300 Projects and 2,000 Photos PDF
100% (8)
Black & Decker Complete Home Improvement With 300 Projects and 2,000 Photos PDF
602 pages
6037 - Black & Decker The Complete Guide To Finishing Walls & Ceilings
100% (28)
6037 - Black & Decker The Complete Guide To Finishing Walls & Ceilings
244 pages
Construction Standards
100% (6)
Construction Standards
379 pages
2020 Samsung SMART Hospitality Displays: About Samsung Electronics Co., LTD
No ratings yet
2020 Samsung SMART Hospitality Displays: About Samsung Electronics Co., LTD
11 pages
Lesson 3 - Art Appreciation, Art Creativity, Imagination and Expression
No ratings yet
Lesson 3 - Art Appreciation, Art Creativity, Imagination and Expression
13 pages
2025 All-State Band Results
No ratings yet
2025 All-State Band Results
4 pages
To Purchase A Twilight Bay Catalog or Locate A Retailer in Your Area
No ratings yet
To Purchase A Twilight Bay Catalog or Locate A Retailer in Your Area
57 pages
Fiber Rope Damage Assessment and Acceptance Criteria - New Guideline
No ratings yet
Fiber Rope Damage Assessment and Acceptance Criteria - New Guideline
7 pages
Superconductivity: Key Concepts and Applications
No ratings yet
Superconductivity: Key Concepts and Applications
8 pages
Guideline and Procedure in Accounting and Reporting of CCSDAF Firearms and Ammonition in The Custody of PROs and Other LEAs (Signed DCSG
No ratings yet
Guideline and Procedure in Accounting and Reporting of CCSDAF Firearms and Ammonition in The Custody of PROs and Other LEAs (Signed DCSG
8 pages
Technical Specifications for Air Handling Units
No ratings yet
Technical Specifications for Air Handling Units
3 pages
2024 11 29 Revised Result May June 2024
No ratings yet
2024 11 29 Revised Result May June 2024
3 pages
Spremanje Studenata Etf-A, Spr-A, Fit-A, Pmf-A, Udg-A Tel. Tel. 067 555 439, Web: WWW - Spremanje.info
No ratings yet
Spremanje Studenata Etf-A, Spr-A, Fit-A, Pmf-A, Udg-A Tel. Tel. 067 555 439, Web: WWW - Spremanje.info
16 pages
Copycat Starbucks Vanilla Cold Foam Recipe
No ratings yet
Copycat Starbucks Vanilla Cold Foam Recipe
1 page
HVAC Market Entry Strategy in Brazil
No ratings yet
HVAC Market Entry Strategy in Brazil
12 pages
Reading Assessment
No ratings yet
Reading Assessment
3 pages
Four Year B.E. Degree Course Mechanical Engineering A.U. College of Enginering Scheme of Instruction and Examination
No ratings yet
Four Year B.E. Degree Course Mechanical Engineering A.U. College of Enginering Scheme of Instruction and Examination
55 pages
Netflix's Shark Whisperer Wants Us To Think Sexy Conservation' Is The Way To Save Sharks - Does It Have A Point
No ratings yet
Netflix's Shark Whisperer Wants Us To Think Sexy Conservation' Is The Way To Save Sharks - Does It Have A Point
1 page
Reading A1 A2 Reading For Summarizing - Strategy 9
No ratings yet
Reading A1 A2 Reading For Summarizing - Strategy 9
9 pages
Understanding Disease Stages and Movements
No ratings yet
Understanding Disease Stages and Movements
16 pages
Grey Kangaroo 2025 Solutions
No ratings yet
Grey Kangaroo 2025 Solutions
13 pages
HCAI Awareness and Control Survey
No ratings yet
HCAI Awareness and Control Survey
4 pages
Faith and Authority in Action
No ratings yet
Faith and Authority in Action
10 pages
Using Zend Framework 2 Sample
No ratings yet
Using Zend Framework 2 Sample
44 pages
Super Gene 801-900
No ratings yet
Super Gene 801-900
489 pages
Sdarticle 007 173782
No ratings yet
Sdarticle 007 173782
5 pages
Improving Style
No ratings yet
Improving Style
1 page
University Application Summary for Kodjoman Sanogo
No ratings yet
University Application Summary for Kodjoman Sanogo
1 page
Financial Markets and Institutions Overview
No ratings yet
Financial Markets and Institutions Overview
4 pages
Understanding Galinsky's Parenthood Stages
No ratings yet
Understanding Galinsky's Parenthood Stages
5 pages
KZN Maths Grade 10 June 2024 P1 and Memo
No ratings yet
KZN Maths Grade 10 June 2024 P1 and Memo
10 pages
Iannis Xenakis A Stochastic Approach of PDF
No ratings yet
Iannis Xenakis A Stochastic Approach of PDF
13 pages
As Economics 23-25june - Merged
No ratings yet
As Economics 23-25june - Merged
80 pages