AFRPL-TR-65-iK

O STATE-OF-THE-ART SURVEY
Or METALLIC BELLOWS AND DIAPHRAGMS
• FOR AEROSPACE APPLICATIONS

Dr. L. E. Hulbert, Dr. R. E. Keith, and T. M. Trainer

Battelle Memorial Institute
Columbus, Ohio

TECHNICAL REPORT NO. AFRPLI-TR-6S-M

November 1965

Air Force Rocket Propulsion Laboratory

Research and Technology Division
Air Force Systems Command
Edwards Air Force*Base
Californi&
 NOTICES

Qualified users may obtain copies of this report from the Defense
Documentation Center.

Defense Documentation Center release to the Office of Technical

Services is not authorized.

When U. S. Government drawings, specifications, or other data

are used for %uypurpose other than a definitely related Government
procurement operation, the fact that the Government may have formulated,
furnished, or in any way supplied the said drawings, specifications, or
other data, is not to be regarded by implication or otherwise, or in any
manner licensing the holder or any other person or corporation, or
conveying any rights or permission to manufactare, use, or sell any
patented invention that may in any way be related thereto.

If this copy is not needed, return to AFRPL (RPRPD), Edwards,

California, 93523.
AFRPL -TR-65-I

STATE-OF- THE-ART SURVEY

OF METALLIC BELLOWS AND DIAPHRAGMS
FOR AEROSPACE APPLICATIONS

Dr. L. E. Hulbert, Dr. R. E. Keith, and T. M. Trainer

Battelle Memorial Institute
Columbus, Ohio

I
TECHNICAL REPORT NO. AFRPL-TR-65SI

November 1965

Air Force Rocket Propulsion Laboratory

Research and Technology Division
Air Force Systems Command
Edwards Air Force Base
California
 IBattelle- N"IemnorldI Iflstitite *COLUMBus 'L'A BORAlORE

505 KINCr-AVENUE COLUMBUS.OHIO 43201 -AREA CODE 614. TELEPHONE 299-315i -,CABLE ADDRESS: BATiON

Novb 30,19651

Air Force Rocket Piropulsion Laboratory

Research and Technology Diviiion
Air Force Systems Comm
Edwards Air Force Base
California 93523
Attewtion Lt.*Jobn L. Feld.rni, RPMP
contract No. A? oi4(6ll)_io 5 3 2
AP.W. Prc~ject No. 6753

Gentlemen:
Special Report--"39tste-of-the-Art Survey
of Metallic Bellows and Diaphragm for
Aerospace 4appications"

Enclosed with this letter awe five (5) copies w~ m (1) repro-
ducible meter copy of Technical Report No. AIRPL-UT-65 - Tis special
report has been prepared as an initial part of it 2-1/2-year pr'ograu on the
"Develqpsent of hnsltical Techniques for Bellowsand Diaphragm Design".

Cai March 1, 1965, the Air Force Rocket Propulsion Laboratory,

through contract with Battelle Meorial Institute, undertook a program to
establish analytical design procedures, stress-analysis methods, techniques
for manufacturing control, and other *Actors essential to the successful
deeign and fabrication of metallic bellows and diaphragms. These objectives
were to be accomplished by the following tasks:
1. The survey of knowledge available from. the open literature and
from indstry on the design, analysis, fabrication, and testing
of bellows and diaphragms
2. The Identification of parmfters pertinent to the design of
bellows and diaphragm

3. The develowant of improved stress-amlysis techniques ani

design procedures utilizing mathmtical softls and computer
techniques
ii. The investigation of improved fabrication, asseublyp and testing
* techniques

0ICAT ItD
DED TO T HE ADVANCEM 9NT 0PF CI CNI -
 Lt. John L. Feldmn NDwedmr 30;. 90

5. The verification o,* selected analysis techniques and desilm

procedures through Laboratory evaluation

6. The classification ovf bellows and diaphragms

7. The pig.Oicaton of the program results In handook format

for- distribr-1-11on
to selected fAIIAties.

The enclosed special report has been prepared to iarize the -

res of the literature and i r survey for possible Intirested- facili-
ties. In addition, the report presents recmedations for activities during
the reinder of the ]program whch preclude the duplication of ork forwrly
conducted, anI which require priority to accomplsh tbe objectives of the
Air Force.

We hve appreciated the opportunity to conduct the literature and

industry survey and we look forward to the further achievement of the program
" cbjectives. Any cca'eints on the research- program and on this report will
be appreciated.

Very truly yours,

Aerospace Coonents Division

I o
IA-rTCAT A -

Or !AL1 BU"W AND DUA0M

FOR AVICPACS APPL!CATICES

Dr. L. E. Malbert, Dr. Rt. E. Keith,,

and T. M. Trainer.
 FGRIW~aD

This report sumarizes research conducted under TJWA Contract on.

0o(611)-10532 from March I to July 31, 1965. This contract wVa eutabliuhed
under Air Force Program 8trutitre No. 750G, AM., Project No. 6153, AkW Task
No. 675304. The work was peeormed by Battellet lamriul ii-tW.t: for the
Air Force Rocket Propulsion Laboratory, Research and Techn!0lo Divlsion,
Edwards Air Force bas, with Lt. John L. Feldman and Messrs. A. D'Aicaip-o
and Roy A. Silver serving as contract monitors. The princip l cotribzutr
to the report were:' Dr. L. E. Hulbert, Dr. R. B. Keith, and 2. C. Rodeb,
Senior Research Engineers; Dr. H. J. Grover, Senior Fellow; and T. N. Trabihr,
Program Manager. j
Publiction of this report/does not constitute Air Force approval
of the report' stfindings or conclusions. It is published only for the ex-
change and stimlation of ideas. /I

John L. Feldan /
l/Lt.,. Useh
"~Project Egneer
I
/
/

U
 A 2-1/2-yw pm bas bes wi~etwm to ooda11h awayti
bsra ofteus, stress-oslysls intho"i, techniques fwj iimjimettu
cOmtrcl. &MOthw ftettws smutla to the smoeisfu3 AssSjp-ia t a tics'
of ast&Mc bollws em dqisbawos 2fe 1Wait hns
- of the as
-b~r
1zwbies1 a state-of-tho-art aw*q to issidt j the ft rulmtl ofthe-
bust maws of wampclishing the a refl a op objectives Thu rep
sie the results at the smq aMA eeAts rqctim s flt1Q
ressindw of the pm. AA amotat*A bWblI oelf of- 3T6 iefere~es is
 TANA Or cinus

SUKMAI AID EUCOmIW (IO . . . . . . . . . . . . . . . . . . 2

L1UMR AD IIWC 0****000001 2V!.00

Literature ...... . . . ... . *. . *. * 1

Inustry o•urvY
oo 000 0 0.00 0 000 0...0 5
TPCAL ENZW AID MIA o.o..... • • • • .... 5
Be11ovs Convolutions &a o a o . *. . . .* 0. 0 0 .0 10
Bellows Assemblies o o * o........ 0 0. 0. 0 . . 14
Bellows Anicti a o. o. 00 0 0. o. .0 0 0 0 0 17
Diaphra Configurations o . ,. . 0 0 . o0 o o o .. o o 19
Diapbra Applications o o. . . 0 0 0 0 0 0 o o. . 21
TOMI CAL AALTSIS Cr EAM ICVMA!IK Or MMLOIB AND
DIAPAWS o o o 0 o o ** o0 o o o o * * * 0 0 0 0 0 0 0 22
fte Analytic Appo-h o.. . . . 0. o
o 0 0 0. 0 22
"TbeFiite Difference A hrosh

SDire,-Integation Approach . o. .. . . . .. .. 2
0 PzaR n 0 0 0 a 0 0 0 0 0 0 0 * 0 & 0 0 0 0 0 & 0 0 0 35
Design Procedures for Bellois 0 0 & 0 0 0 0 0 0 0 0 0 0 0 0 35
DesignPvenes for Diapabs 0 00000o 0000* 41
1AIMI CLGY 0* *0 *9 * 0& 0 0 0 0 0a 0a 0 0 0a 0 115
Mterials 000 00000oo&oo oaoo9 o..00
o* a00 15
ProcesmeB * 0
3 Eufacturi 0 0 0 0 0 0 0 0 0 0 0 0 0 0 &1.8
NRMM faTWM13XM o o .e
In-Process Inspoiction o .o .&0. .* .o.
* .o o o oo oo 0o..QC..
o9 o 0 ..0
o 59

EvSluation PzxCO&MOs for BOM a 0 *0*.. * * *0 0 0 6

1valuation Procedtres for Diqwms 00o o o o 0 66
W*0 0 0 0 00
00 0 0 00 00 0 0 0 00 0 0 a0 0 6T

AIOED BnU aP0 000 000 0 00 a 0 00a0 0 00 00 0 1

iv
* PIIle T37=CA RELG AMMMM . .... 15
7 2. AXMA SIMC (XP La'S M=u =WC
W ys - PLMU.
AC213G IAWS *.. .. 28
P 3. N4XA02MG HM~ 8 - FMW MLMI 0 0 a 0.4
rI 4. NNUFXMIU JLI 8 - Ila= LIWB 0 0 0 0 *
17109 5. MACIMU31 PWI 8 - m Wzmor . . . . 57

L= CF TANXS

TAXZ 1. 1KLM AND DIAPR~a FE1 TM AUCBWAC

B 1
331JO S 0 0 *0 0 00 *S 000

TAKZ 2. X&TC3 LMI CCWMa1IW8 AM' NI~~ 21

TAM 3.* TfPIC=~ LGI~X~ M.= o
9.=D-M o . .

TA=Z 4. TfPXC* WN 1D.MLLM ALM. . . . . . . . . . . . '47

I
ABEV1ATICHB AID 8DUM8

, e coordinates of a point on the midlle surface

distance measured from an arbitrary origin a-Ag the meridisai
in the positive direction of f

R, Re principal radii of curvature of the middle surfe

r R9 sin

B Young' s modulus

V 'oisson's ratio

h thickness of shell
K v2)

D b2 K/12
w, u u components of displacement
i7
angle of rotation of normal

P, PP P components of surface tractions

m, m components of moment

I N, N mbrane stress resultants

%, M%,M moment resultants

- cp ~transverse shear resultants

N, Q effective shear resultants

( )'x partial derivative with respect to x

got 4T, e99 mmbrane strains
KOP K bending strains

'I

.. . . .. .
-I
I bility of providing advanced tetbnoloa for rocket proplion systems thrOtgh
exploratory develosment programs. As a part of this missin, ipo ~t
Ibeperformance and reliability of flui&-system camoneits (valiies ian

regulators) are being pursued vigorously. This is beiing iciPlishjed by -il~

vestigations into individual comondnt parts or modules.- lce*, seals, fittigs,

The ArForce Rocket Propulsion Laboratory, through a contract

attelle Memorial Institute, is establishing analytical design pro-
Dih
cedre.,stress analysis methods.. techniques for maufacturibg control, and
other factors essential to the successful design and fabrication of bellows

I and diaphragms. The finial results of the progrm will be *publihed in hand-
book formt which will be distributed to selected facilities.

f The initial phase of the progras bas involved a literature and

-
industry survey to determine the present state of the art of bellows ad

Irecmsindations
I
diaphragms. This report sumarlizes the results. of the survey aad pr~esents
for activities during the remainder of the program which
require priority to achieve the objectives of the Air Force.
 ieNK AND RECClIOaM WIS

Approximately 50 manufacturers are engaged partially or entirely

in the design and manufacture of metallic bellows and diaphragms for the
aerospace industry. The availability of bellows is significantly different
from that of diaphragms. Bellows are used in sufficiently large quantities
that most manufacturers offer a range of sizes and materials. Although
bellows for aerospace applications are rarely stock items, tooling is avail-
able for "standard" ccnfigurations and delivery can usually be made within a
period of several weeks* Bellows are available as separate items and a;
special assemblies. Diaphragms, on the other hand, are not usually advertised
for sale. While diaphragms can be obtained from certain firms on special
order, most diaphragms are designed and fabricated "in-hodse" as parts of
particular components.

Considerable theory has been developed concerning the behavior of

bellows and diaphragms. However, the theory is complex and the performance
of the actual items is affected by variations in material and fabrication
parameters; conee-quently, bellows and diaphragms are designed almost entirely
on the basis of empirical methods and experimental data developed by each
company. Thus, while it is possible to predict the performance of bellows
similar to those which have been manufactured and operated prefiously, it
is difficult to predict the performance of new designs, or of old designs
operating under new conditions.

Cosidering the wide variety of factors influencing the operation

* of aerospace bellows and diaphragms, the performance of these items is better
than might be expected. However, it is believed that their performance will
be significantly improved through the accomplishment of the major objectives
of the program, i.e.:

l. The developent of impr ved stress-analysis techniques and

design procedures utilizing mathematical models and computer
techniques

2
 2. The investigatiin-fIi1 d-fibidation, assmbly. and
testing techniques
3. The verification of selected analysis technii and design
prcedures thrug lab oratory evaluation
&. he classifticition of berLova and diaphas
5. The publication of the progrm results In eadbook faint. for
distribution'to selected fatilities
A large amount of work has been don on the design, fabrication,
and evaluation of metallic bellows and diaphragms. Mwe may configurations
and perforuince characteristics considered,, va the noW design approaches.
utilized make the application of this InforAatioo difficult. A.wing the
remainder of the program it is recinnded that the eqhsis be directed to
experimentally determining the accuracy and rawg of validity of the manst
mennful theories and synthesizing these theories into an easily used
design procedure. Because of the intensive effort being devoted esewhere- -
to the study of the. large plastic deformation of mentallic- belloa *An Ali%
phrages, it is recomended that the present _prog be concerned Al#-metaiic,
bellows and diaphrages operating larely In tht elastic region. It Is furth~er
recoinnded that the work on diaphragms be limited to corrugated'diqhraWas,
because this type of diaphragm is usually used in aerospace caqpmnnts.

3
 L1TMJE AND DIW~IBY SURVEY

The literature survey included a review of technical papers, text-

books, articles, and government reports on bellows and diaphraM design, per-
formance, and analysis. Selected material in related areas such as manufac-
turing, stress analysis, and testing was also reviewed. The industry survey
included the solicitation of brochures from manufacturers, and trips to repre-
sentative manufacturers and users of bellows and diaphragms to determine the
design, fabrication, and evaluation procedures currently in use.

Literature Survey

An extensive biblioraphy of pertinent published material was pre-

pared. Many of these references had been identified during the conduct of pre-
vious projects at Battelle. Additional references were located by a detailed
search of Applied Mlechanics Reviews Chemical Abstracts, Instrient Abstracts,
Applied Science and Technoloay- Index Nuclear Science Abstracts, and U; S..
___ernen Research and Develosent Reports. Copies of most of the references
were ordered. As the material was received, it was reviewed and abstractedi and
copies of other pertinent articles were ordered. Approximately 500 published
references were reviewed.

To locate additional gover ment reports, machine searches were re-

quested from the Defense Documentation Center ,and from the Scientific and
Technical Information Division of NASA. It was requested that the MC search
include the years 1950 to the present, while the STID search was requested
for the years 1961 to the present. (The file in the latter facility has only
recently .been automated.) A search was also made of the Transducer Information
Center at Battelle, vhich is sponsored by the Air Force, and a bibliograph of
reports was requested from the Interservice Data Excbange Programs. The IDEP
is a group of companies and goverrment agencies that circulates brief reports
on the test resultc of various aerospace components, to eliminate the duplica-
tion of costly testing. As a result of these inquiries, 61 references were
ordered from theDDC, 24 from the STM, 18 from the TIC, and 21 from the IMP.

if
 An ennotateil bibliogirapkiy of the ~most peitlaint- refewenoesrive4
during the literature survey is presented at Up~ and of the risowt..A s
inatio1 of this material will ohom a large-mount--of informaetion-of a tbeicalI
natuve. A significant amount of Informaticai e igian
Oetistxp
mental evaluation, but there is little in-formailon '.n fabricaotio. 'Althug
useful Information may remain in as yet u=located governt~ repm as the
refer ences listed in this report constitute a reasonably comrhensvs--b1blio_
gr&aph of the unclassif-ied literature on metallic bellows and iprs

Initially, letters vere sent to advertised manufc of bellows

arr
and diaphragms requesting design Information and brochures. As the program
progressed, additional firms were identified and trips were xdg to 21 repre-
sentative manufacturers. Table 1 lists the manufacturers identified dur~iag
the survey, and, the types of bellows and diaprasme by these firm.,
Visits mere also mad to four representative users of bellows and
diaphragis: (1) Atordics Internstional, Division of North Anuilian Afilatio,
* (2) CaLae Yknufauturing, (3) Thbe Missile and fpace Division of DWOugls Aifr-
craft, and (if) Parker Aircraft. Atcaics International. has done consierible work
for the AUG on the design and evaluation of bellows for high-temperatuie condi-
tions where fatigue and creep awe primary considerations. Calmec is primarily
a mnfcurer of valves, many of which use electrofoied bellows with low-
spring rates and large deflections. Both the Santa Monica andL Hantington Beact
facilities of DiugWlas are utilizing bellows on the Saturn IV and other aro-
space applications. Parker Aircraft uses bellows for a variety of pressure
measurement wa actuation applications.

TPMEIKm- AND. ZWAULA

bellows anid dia3iLraws are thin elmuats whose Geflection charcter-

Istics are utilized to provide moemt In the structure of fluid systaie.
Usually thene Itaks are used when the need for motion Is coined with the weed
 MI~tr~ilx x 4 N 14 N

_-uoou - - -

UUITTIW N* x IN x '
I
Ij

$4 Ha L-LT M x xW x X x X

TT~jagC MN. x x x

xtoqdeoti _ - -- - - - - -- _

E-44 Sai NV W

sa9qDUT 01 4,

w Tesd ga.ms0

Tod ga.ms
*j XW(
d'-qns
&A - -

0~ sja x ___

S;Bo 1 -t 59 -c
fff xW (oi

021 u '.4A

0 i4. 0 0,4 4

r. . 0

8 .3
V * * k

r46
 - -

- - -- O~ - - - -- - K-

I- -V - --

xteaU H HH HH
TOM - m - -
H H H H -

zoqdgoti - -

*~~~~~ H HH H H
x
a-o gm 0--
HH H H ~~~~~~U A.-I !5
z -B - - - -- -

H H H H H H H

I' H- HH
Todgam

STod 'sams
43, j------------------ 8- - - - - -

M -goa -

&r

r4 saIogT P1

4 I Gg U1

-* V -BJ

saqo -- -

sart -O CO P-C4 -%

a iin

dO 54 A
vl 0 0 V

~r

-0I
50
V4
. 41 0d

Ax 7 4 0
 ~-XX 4 ~~I- I'- v P

mN N

09CouI N~ N N M NV

IO~I~N N N N N N x
WUZ~~U~~ty L

aS Oji X NjN N NN
H.I1.-ltN x x - N

Nfsj.NN N N
9STUs -am -o -

m0 N

Td -com
a -said q~__U'U

Saqoujl c - -

-
g~a - -- -

sam 0. %D

wo ~ 0

Iit,

t0
 UTQ3 x x Ix

Te -0=8

"CuID X X

Tsd Io1m XX

sa=

Xi X XI

-sojX XW
jqoj-- --

E4 9d Io.ans

fCID X1K

.4E- totjXV

-nows~oU
 for a hermetic seal. Bellows and diaphrass are made in %W- ftseskand shapes
from a variety of materials. This section of.the repbrt mwized-the prInc-
pal types of metallic bellows and diapbrajas that have been manufaztured, and
the oajor types of applications in which they ,havebeen used.
Ey autual agreement with the Air Force, the pr0 S is con.eerned
only with bellaws and liapbragas which operate either totally vithin, or
largely within the elastic state. Thus, the work which has been dode to
develop bellows and diaphragms as expulsion devices for fluid systeis is not
described in this report, although references -on the subject are included in
the bibliography.

Bellows .Convolutions

Most metallic bellows are cylindrical elements which conUin annu-

lar circmaferential corrugations. In flexible hose, the corrugations may be
formed in a continuous helix. Some bellows are not cylindrical but have elkp-
tical or other nowircla cross sections. However, these shapes are suffi-
ciently unusual that they are not included in the progra. "
Bellows are classified according to one of the four primary methods
Of manufacture, i.e., (1) fomd, (2) welded, (3) deposited, or (4) Macined.
Within these classifications, the bellows are usually categorized aci:dIg
to the appearance of the convolution cross section, as shown in Table 2.

Formed Bellows

Formed bellows constitute .approx~mtely75 percent of the bellows

which are manufactured. They can be produced in a variety of materials and
sizes at relativey low cost. Diameters uto four feet are readily avail-
,A able and one manufacturer advertises diameters up to 50 feet. Althchgb Table
d 2 shows on3y single-ply confJiwations, most formed bellows can bt made with
multiple plies. Thrae- and four-ply bellows are c-on. ftltiple plies are
used to provide a lover sping rate.than would be obtained with a single ply
equal in thickness to the total thickness of the multiple plies.

Sant oroidal. Semitoroidl bellows are attractive for materials

with relatively low ductility. The form also offers good pressure-capability

10

, - -- ---- . ---- - - --- --- .-

° .
 !Aa~2 VCRS MUM. OW a A cn ime

Sn'taridal my high very pow very good

U-apd,
Rmhgh
xt fair- vay good

U-w~e~pIa.RI igh fair vry good

Moei ]itRl falr vewry good

feW
sh&Sjt ehig poor mey ood

18.1l, lat. ?resmwvemy hih POWr *XCelleut

lo goo good

stod R31pe VWr 1OW 0xcelnt "oo

*e4(amr be varied) low good faIr 1

high fair mnr

and stability. 51C convolutions may be -truly semicircular, el1iptical, or -' '
some combination of curvas. A lw deflection capability per convolution and
a high spring rate are major limitations of this configuration.

n . When flat sections are placed between the semitorOidal

sections, a U-shaped, or flat-plate-bellows configuration is formed. Over 50"
percent of all the bellows are of this type. The shape is amenable to any of-
the methods for manufacturirg formed bellows, a variety of performance charac- -

teristics can be achiew.d by varying the radii and depth of convolutioni, and,
supporting devices are easily installed externally or internally.

S-Shaed. The S-shaped bellows is similar to the U-shaped belivfas. -

By sianting the straight sections between the semitoroidal sections, or by

connecting the semitoroidal -sections with curved sections, it is possible to
form more convolutions and thus achieve more deflection per unit length. Zie-
S-shaped bellows is not as easy to manufacture, and is not as amenable to the
use of cupporting devices as the U-shaped bellows.

Toroidal. Toroidal bellows have been developed to reduce the pressure-

induced stresses in the bellws. By using a shape which is essentially cir-
cular, the effects of pressure are more evenly distributed along the convolu-
tion. In addition, the stresses in the convolution are less affected by an
increase in bellows diameter than is the case with the other convolution
ehapes. The Marquette Coppersmithing Company claims that their "OMEGA" shape
distributes the stresses more evenly than a true toroidal bellows. Zallea
Brothers advertise a "HyPTor", or modified toroidal shape which is satisfac-
tory for intermediate pressures and is more flexi.lz thin n'true toroidal
shape. Althcugh the toroidal bellows permit hig;2 operating pressures, it is
more difficult to manufacture than the other tormed bellows and has a high
spring rate.

12
 Welded Bellows

Approximately 20 percent of the meaifictured beilow*s have 'vild

convolutions. Welded bellows are made up of shapedi &daphr#'4-sifi' dr
l~
nittely welded together at the ininer and outer dimisteirs- ilthoghb tihey are
more expensive to manufacture than forned bellows, welded bellows offir three
significant advantages over fozd bellows: (1) a wider choice of mteri'alsB-,
(2) more deflection per unit length, resulting in shorter assemblies or
loiager strokes, and (3) a wider choice of performance characteristici-s-bediise
of a greater variety of convolute diumnsions and shpes. Although Sealol Is
offering a two-ply welded bellows, most welded bellows have aL single Ply.

Because the diameter of the welded bellows is deteruined by the

diameter of the stamped diaphragm, welded-bellows sizes are limited by the.
available stamping equipment,? In goneral, welded bellows are available in
sizes from 1/2 inch to T inches outside diameter, but bellows in excess of~ 12.
Inches in diameter have been produced.
*Table 2 summrizex thes otypos of welded convolutions,, and their
primary characteristics. Most welded bellows are of the nested-ripple configu-
ration because this design. makes maximum use of the advantages of low spring
rate and ccmaqutness. However, the other configurations have attrictive
charabteristics for certain applications.

Deposited Bellows

Two kinds of deposited bellows are comssrcial.y available: chemically

deposited, and electrodeposited. Both methods can be used to produce NW~
shape that can be deposited on a mcined? ?adrel. In each method an alminia
Mandrel is machined for ea.-h bellowe, and after. the bellows mterial is
deposited, the mandrel is dissolved. The prly adVantages of the procosses
are the ability to produce: (1) very thin-wiled bellovs,. (2) nonvelded
* bellows, (3) very small bellows, and (14) scm special shaped bellovs.
Chemically deposited bellows are produced by Ncwtals Corporation
* from a material called "lieeimetal". Although the material is acknow10deed to

j 13
 I-A

be about 95 percent nickel, its exact cca sition and the method-1q.ch-t
is deposited are proprietary. Chemically deposited bellows can be--: ie with
wall thicknesses ftcu 0-0003 inch to 0.005 Inch,,. and vith. dtametegs, from 0.060
inch to 7.00 inthes. IlectrodepQsited or electroplaed bellows are .usu&.ly
produced in nickel or nickel-cobalt alloy. Sizes are. avilable from .Q.063
inch to 1.250 inches in diameter, with wall thicknesses varying from. 0*0003.
inch to 0.006 inch.

Machined Bellows

Machined bellows are turned or ground from bar stock, tuting, 9r.
forged rings of most materials used in other types of metallic bellows, as
well as of materials not found In sheet stock. High-strength, high-endureae,
heat-treatabl.e tool steels, in addition to high-strength, lov-modiflus titaium
alloys can be used. The design of machined bellow is customized, with most
wheined bellows having high spr~tng rates. Machinied bellows -have been made
from 1/4i inch to 60 inches in diameter, for ]*essures as high as 12,000 psi.

Bellows Assemblies

Although bellows can be used in an extremely wide variety of ways,,

certain types of assemblies have become relatively cboon. In general, these
assemblies, some of which are shown in Figure 1, prevent certain types of
motion and li.mit other types of motion.

Bellows with a, length-to-diameter ratio greater than 1 my be dis-

torted by internal pressure Iii a =ner referred to as "squirm%. When squirat
occurs, the -pressure my not be.-sufficient to distort the individial convelu-
*tions, but the bellows may deflE~db as a columnj causing damage to one -or note.
convolutions. Metal braidinig is commonly used to provide extermal support
for such bellows and to protect the bellows from external damage. This con-
* figuration is most camon .u flexible metal hoses., Braiding may cause
abrasion of the bellows and it may accelerate corrosion o~? the bellowts.

114
I~eAsml ~ trSYLne Ass*I

Gimbalmfl Assmby GdM Auy -

Universal Asesobly Xitm l is uu1er Assi~21&

(Permduuion bas been grergd to use the aro repoductions by the Awoqu*
Corpation/Jkruen Division, Los hAogs, CWIitamla)

8* FI~~ 1. !YPICAL IEL

A5.05

15
 Sleeve

A bellows may be provided with an internal. ilarnito reduce fluid.-

scrubbing contact with the bellows. Suih contact can as fo losses,
abrasion,, noise, and flow-induced vibration which can lead to early failure.
An internal sleeve can also be used to limit the aaount of bellowis compression.

Slide

A slide aiseumgy is similar-to a sleeve assebly ecept that tele-

-

scoping sleeves provide axial guiding and prevent squirm. Slide assemblies
are attractive for system subjected to high surge ppessdres and temperatures.

Universal

A universal assembly contains two bellows joined by a Joismon -on-

nector for the purpome of absorbing ary combination of the three basic move-
ments, i.e., axial deflection,. lateral deflection,, apd.-ngular deflection.
Limit rods are often used to distribute the movement between the two bellows
and to stabilize thbe- comon connector. This configuration ciii tolerate mare
-lateral deflection or offset than one bellows equal in length to the two
bellows.

Intenally Linked

*nternall -linked assemblies utilize struts inside-the bellows in

the flow strean to limt moemt In all directions. &auch assemlies are
*sil~le, coapct, and lightweight, but they Introduce significant' flow losses.

A hinged bellows is designed to permit angular dflection in one

plane only by the use of a pair of pins through hinge plates attached to the
ends. The hinges and hinge pins must be designed to resist external forces
and thrust due to internal pressure.

16
 7

Gimbaled,

*GWmbad, bellcxri assemblies provide for angular deflection in sa

direction from the tubing axis. Internl,* or external gimbiii can bet WIiueC.
They are made with two sets of hinge plates attached to a~-glial riijg and:
function in a manr similar to a universal joint.

Pressure Balaned

A pressure-balanced bellosa -assembly is designed to absorb axial move-

munt and/or lateral deflection while absorbing the end thrust -by means of ti
do-vices interconmecting the flow bellows with an opposed, bellows also subjected
to line pressure. This type of assembly is noxuall7y used where a change of
direction1 occurs in a run of piping.

Bellows Aplicatioins -

Bellows are u~sed primarily in two ways; (1) as mtion camvenstors,

and (2) as calibrated transducers. Typical applications are discussed biiefly.

jexnsion Zoints

iNyW bellows are used as expansion joints to compeanst for me-

snt occurring in pipe lines as a result of teaperature changes &aor as a
result of external loading. With the wide variety of piping systems, move-
amt can be transmitted to a bellows as compression, extension, offset, rotaL-
tion, or combinations of these mtions. Most of the "standard" bellows as-
semlies have been devoloed to compensate for therml cbeas in industrial
tubing and piping systems.

flexible Connectors,

Bellows are often used to comensate for structural deflections,

*misalinmnt, and tolerance accumulation. These functions ma be fuilfilled,
by a short,. plain bellows, or by relatively long, flexible metal hoses with
* metal braiding or other types of exterior covering.

17
 Pressure adTmrtr esr

Che of the most comon uses -fr bellois is the'actuation bf some

device as a function of a change in t~he pressure- or tvaqeraturz 'or a fluid-
system. The fluid-saey be external or internal-.to the bellows. For- temperatUre
sensing, the bellows system often incorporates a liquid whose vapor pressure
reflects the temperature being sensed. The vapor pressure is then used to
actuate the bellows. When the bellows moves in response to pressure.chne,
the movement is utilized to actuate a device such as a vave or swItch. Because
the bellows responds very quickly to changes, the wovement-*can be used as a
* proportioning signal as well as an on-off signal. Lijkpbrsaas and Bo~hrdon
tubes are also used extensively for these functions-.

Shaft and End-Pace Seals

Although most reciprocating shafts &re sealed with some type of

packing, increasing requirements for reliable, hertic sealing have resulted
in the -wider use, of bellows for. this purpose. ,The ability- of a bellows to
provide sealing and to act as a spring and motion caenator.makes it Ideal-
as a means of pressing race seals together on rotating shafts. Because of
their compact construction and low spring rate, welded bellows are 'cavonly..7-
used in face-seal assemblies.

%jdraulic Motors and Actuators

For. inll power requirements, part icularly for remote operation,. bqllo-Is
an be used for jeonoting mechanical work to hydraulic work or for converting
bydraulic work Ito mechanical work. For exile a liquid-filled system my
consist of two bellows connected by tuidng. The movement of one bellows causes
movement of the other bellows.

Vibration Dameners

just as springs with-frictional elemnts for enrgy adsorption ore:

used as vibration dampeners. iv~ mechanical systems, bellows are used,"

2 18
 vibration dameners for fluid-containing structures. Because the beilowi
itself responds -to certain vibration frequincias, care waet' be taken to &"asure
that the bellows is not excited by the frequencies transmitted from the system
structure.

- jAccmiulatorm and &iock Absorbers

In a bydraulic system, a bellows can function as-& flexible-con-

taimer to mintain oil flow or to absorb srge pressures. Tbe primary -cq12ire-
muts are long stroke, resistance to high differential pressure, and 4quick
response*

Volum CgMensators

For liquid systems of mn3.l volme, bellows furnish an attractive

mans of compensating for fluid expansion or contraction. sruvetically sealed
* floated instrmnts such as aros and acceleromters are typical applications.

Flexible Couplings

Altho~ag bellows are not usually designed to transmit torsional

loads, bellows have been found to be attractiv'e for transmitting Simll torques,
particularly for Inatrumentation-type equiWnt. Mie bellows can tolerate
sow mislimnt and the torsional stiffness of the 'bellows Insures acurate
rotationa traunission. Bellovs-type flexible couplingg are available for
shafts up to 1/2 Inch dimeter.

DI*wan Configrations

A diapbragm is a thin diak-like element which deflects In a direc-

tion si*bstantislly perpendicular to its flexible surface. 11tallic diaPbragmS
* are classified as flat or nearly flat, and corrugated. They awe used primarily
as actuators to transform pressure into linear notion and force. Corrugated
diapragas are preferred for aerospace components such as valves and regulators
because their average sensitivity ovnr a large range Of pressure Is greater than
flat diaphragm of the sme size, their zero-position under no-load in more

19
stable, such larger deflections can be cbtained without permanent defomtion,
and a variety of pressure-deflection characteristics ma be obtained-for a-ivn
size disphragm by usirg different depths or shapes of cotrugations. This TWf
is concerned only with corrugated diaphragms.

In. contrast with the variety of informtion on bellows sizes and

shapes contained in the open literature and listed io catalogs, little inform-
tion has been obtained' on comsrcially avpailable corrugated diapbragas. The
most comlete.dicussion of possible diapbragn configurations and their per-,.
forance characteristics is contained in a pV&bliction by Newell (l* which has
been approved by the Diaphragn Research Suconittee of the Amrican Society
of Mechanical kagineers. I
In general, the work reported by Neell~l and by Wildhack, et al.()
shows that the effect of increasing the wnber of corrugations is to increase
the Initial flexibility as well as the average flexibility over the usable
range, although diaphragas with moe corrugations ma be moe nonlinear. Dia-
hbragas with shallower corrugations exhibit better Initial flexibilities within
lmits but at the expense of decreasing the linear rage According to Wildhack,,
et a. dsbaamaewith triangular andtrpziashesrodelnr
force-deflection characteristics for deflections up to at least 2.percent of'

angular shape being slightly more flexible than the corresponting diapbragM
with the sam nober of circular corrugations. Although eapirical relation-
ships haeve been developed, the design of corrugated diaphragms is reported to
be largely a process of trial and error.
According to a survey by Giannini Controls, ()corrugated disphragas
are manufactured in sizes from 0."T to 6.0 inches In diameter, and are
usuall.y Joined togerther In pairs to form a capsule. They can be used to
sense pressures from 0.5 to 4i00 psi; ho-wever, the swjority of units are used
to seum pressures from 5 to 100 psi, and most units are less than 2.5 inches
in diameter. Most diaphragms will not be aore than 1 percent nonlinear if
the displacemnt is kept below 2 percent of the diameter. Xf nonlinearitY
of aore than 1.percent is acceptable, displacements up to 5 or 6 percent can
be obtained.

U151minces are given on pop 6T.

20
 The puablication by *Wl()lists 15 inJw aWA seVeral,1minor AMU-
cation classifications for corrugated dispbraas. !Ia most iqpartant qpl cS-
tions frcu the standpoint of this progrext we: (1) in~se and t1eatr
ses ps (2) linear sbaft seals, and (3) rolum coarensetwe. Sinde these
applications have been discussmed previously for bellows,, they will aft-be
'.1repeated here. It should be noted that the primary advsntaps,,of dimpbrsgas
scompared with bellows are greater sensitivity ian awe copit sapso-for
som components. Low deflection and low pressure capability w~e the p1"inary
- limitations of dimpbraas, althoug the pressure c4paIlity is often altere by
the use of springs.

21
 (ff L~I AIM DIAPUMD

The theoretical. investigation of the performance of bellows and

diaphragms har bad a long and sometime colorful history. The' nuin-theoreti-
cal develomnts have taken place in the last half century .since the jeri*a-,
tion of the ff. Reissner-Heissner equations for the linear elairtic deftemtioss
of the shell of revolution. Prior to the development 'of comuters, the inves-
tigations were limited to attempts to derive appoxlaate forkulas for :tbe
stresses and deflections In axisymietric shells of special shapes. Hcmevsrj.
the development of digital computers and the accompanying development ofL
numerical methods has -made- it possible to obtain direct nmerical £iolutions.-
for the deflections ad stresses in %xisymtric shells of arbitrary men-d-
ional shape.
This report section gives a brief description of approachen3 thp-t
have been used for developing solutions to the elastic defomtion of ballows
and diaphragms. The approaches may be divided, into three major cate~otbes:
(1) analytic, (2) finite-difference, and (3) direct-Integration. h cat-
integeation approach will be discussed in sow detail, since it is believed
to be the most useful.

The Analytic Approachi

The analytic solution for a general shel! of rev,-Autioax having an

arbitrary neridional, profile has not yet been fond. 5Hidever, approxigate
solutions have been derived for certain shells having constant meidional
curvature. These include conical, toroidal, spherical, and Cylindrical shells.
Most bellovs, and diaphragms can be considered to be nud~e up of a =nuber of
segments of shells of this type, and it is possible to synthesize solutions
to these bellows and diaphragas by *ppropriately comining tbm solutions for
their segmnts. An attempt to calculate the strespes In bellowa &Ad di=phras
in this way will be called the analytic approach.
An noted above, the theoretical analysis of shells of constant
curvature has been studied intensively for the last half century. The bilio1-
grapby lists some of the papers that have beec written in this area. The
nuawe of papws bearing on the theory of azisymmetric shells is so great

22
 that only aselected list could be incorporated. ldevor, except for the
toroidal bellows.- only amal rnmer of bellows or dispaa "lca.s bave
beer. salved by combning the various analytic solutiont.
An early a"Iication of this technique to the calculation of -the
stresses in a diaphragm was reported, In a pMpe by Grover aibd aUiY'k) nis
solution required the inversion of a la0 x la0 mtrix an a desk calculAtor!
Of comuse, present-day applications of this technique use a computer to solve
the mtrix equaions. A well.-knaun application, of this technique to the
.bellovs, problem is given in the paper by Laups, and Wel()Also, in a recent
(tTY809)
study~~IuY Atomics internationO Division of North Awrican Aviation,
Inc. developed a compfter prora that used tbb analytic sipretach for analyzUK
single-ply toroidal and convoluted bellows with and without reinforcing rings
and single-sveep welded bellows.
The chief disadvantaM of this approach is that it Is limited to
bellows cr diaphragm conflguraticns. that can boae up1of sections that can
be soled analytially. lech section In addition to having constant sarid-
ioms] cuvature, must also have constant thicnss and waft be Isotropic and
hooezkeotW. These permters can be allowed to vary fras sgmnt to, segmnt
to account for the nomniifors thickness cc varistion In the elastic properties
of the bellows., Howevar, unless the bellow is broken into a fairly large
imir, of segents, this mW not be a satisfactory way to approxiinte shells
with va71116 thickness or elastic properties.
FIrUaL]y, all of the attemts to vbtain relatively simple apcoxite
formulas for bellows problms by mking further a!! 4ifin "Gssiions wre
amalytical apprachbes. the paper by Turnr and Ford gives one example of
such an qpracimnt solution. J)hW others are Included In the bibhiogapbr.

The Finite-Difference ARwch

The finite-difference approach Involves ch.oosing a set of grid Points

*aloing, the bellows SMd roxisating the differential equation by fini1te-
differewe equations defined at the grid ponts. this sp~robch has been used
I .as the basis for sL number of comuter solutions for prclas-invblving abells,
Or revolution. .,2ipk

23
 This technique is quite general and may be applied to the solution of arbitrarily
shaped thin shells of revolution with varying thickness and elaitic pir er ..
Considerable progress has been made in solving the large matrices of Mtie °'"
that are encountered in this technique. These matrices are quasdia96Win-, "
can be solved by successive elimination and back substitution technique. As
a result, bellows problems can be solved by finite-difference codes such. as he
AVCO code(12) or the CEGB code in a matter of one cr two minutes.on os.-
puter machines of the IBM 7090 class. (15)

Finite difference programs have been written with the capability of

analyzing a variety of bellows configurations and loading conditions. For
example, the AVCO code permits the analysis of bellows with both axisammeric
and nonsymmetric loading in the linear elastic range. It also permits the
anelysis of multi-layer shells in which the layers are completely bonded together
along the entire length of the shell. The code developed at Bell Aerosystems
1 rmits the analysis of the linear and nonlinear axisymmetric elas-
Company
1 7)
tic deformation of bellows and the couputer code recently developed at IT(
j i permits the analysis of axisyinetric plastic deformations of shells as well as

deformation.
,tkastic
The chief disadvantage of the finite-difference techniques is that
the accuracy of the solution depends on the choice of the mesh point spacing.
Since this dependence is not known beforehand for a given problem, the user
will either have to choose a mesh spacing finer than necessary or risk hallin&
to run the problem over again if his original choice of mesh spacing is not
fine enough. This dilemma is even more serious when an iterative solution m-st
be obtained for nonlinear problems. The direct-integration approach provides
a way of overcoming this difficulty.

The Direct-Integration Approach

A number of direct nerical integration schemes have been developed

for ordinary differnntial equations. Since most problems involving axisynmetric
thells can be reduced to one-dimensional problems involvii4 ordinary differen-
tiai equations, numerical integration techniques can be used to solve them.

* The capability of taking varying thickness and elastic parameters into

account includes accounting for discontinuous changes in these quantities.
This permits analysis of end fittings, weld beads, etc.
 This fact has been long recognized. The eu-enied:,,,-h , have
been solved by finite-difference techniques a als6-be solved-1y 'diz
ct4 -
integration techniques. One of the first ccmuter solutions to - c'ivolUted-
diaphbragm problem employed the Runge-Kut~a integration approach.
quent teo-tis work a number of computer solutions ve obtaij d with he-Rjflii;-
Kutta technique.(195 However, it soon became apparent that the dtkedt-xttegiti6n
techniques became inaccurate if applied to some shell problems. This phenomenOn
is discussed in the paper by Septovski, et al. (14) Briefly, the difficulty -
stems from the fact that self-equilibrating boundary loads on shells give rise
to stresses only in a narrow "edge-effect" zone near the boundary. fn solving
a shell problem, it is necessary to take into account the boundary loads at
both ends of the shell. However, the direct-integration approach integrates
from one end of the shell to the other. As the effects of the boundary loads
begin to decay, the numerical integration procedure begins to encounter tru-4ca-
tion errors. If the shell is long enough, the truncation errors completely
obliterate the desired answers. This effect might not be serious if the inte-
gration is carried out over only a half-convolution of most bellows. However,
an integration over a full convolution would probably be Inaceurate.

As a result of this deficiency in the direct-integration technique,

many investigators turned to the finite-difference technique discussed above.
An indication of this was that more of the general computer programs dis-
covered in the literature search were based on the finite-difference than the
direct-integration approach.

An ingenious way to extend the range of applicability of the direct-

integration technique to include long shells has been developed by Dr. A.
Kalnins. A similar technique apparently was derived independently by
o er dogdanoff, by Mirabal and Dight and by Cohen.

The essence of this technique is that the shl is first broken up

into relatively short segments. The solutions for these segments are obtained
by the direct-integration teclique.- Then these solutions are combined in
such a way that the continuity equations between sesnts and the boundary con-
* ditions at the ends of the shill are satisfied siuultaneously. This approach
is similar to some of the analytic approaches that have been used. The princi-
pal difference is that the solution for each segment is determined by numeri-
cal integration rather than by analytical formulas.

25
 It was mentioned earlier that direct integration couldbe us-edt
solve the same bellows problems that have been considered using the finite-
difference technique. For example, the Yale code(20) permits the analysis
of the symmetric and nonsynetric deforution of Isotropic or orthotropic
single- or multi-layer shells of revolution. (Sere, as in the AVCO code, a,
multi-layer shell is considered to have layers coeletel bonded topether
along the entire length 9f the shell.) Arbitrary variations in thickness and
elastic properties, including discontinuous variation, are easily accounted for.
Thus., the effects of end fittings and weldments may be considered. Cohe
C "
has also developed a ccmputer program tMat permits the linear elastic analysis
of orthotropic shells subject to either symetric or nonsyuetric loads.
Recently, the Yale code was extended to permit.the analysis of the nonlinear
axisymmetric elastic deformation of bellows. (2 ) As yet, there does not appear
to be any program that utilizes the direct-integration approach to solve for
the plastic deformation of bellows. However, this problam appears to be amenable
to the direct-integration approach.

Vrus, it appears that the multisegment direct-integration approach

is fully as general as the finite-difference approach for shells of revolution.
The big advantage of the direct-integration approach over the finite-difference
approach is that the integration step size is automatically selected by the
ccputer to give a uniform desired accuracy at all parts of the bellows. This
means that the designer need not worry about choosing the mesh size as in the
finite-difference approach. Further, with the direct-integration approach,
he has control over the accuracy of the answers in contrast to the analytic
approaches where the accuracy depends on the accuracy of the approximations
made in deriving the analytic solution.

The multisement variation of the direct-integration approach seems

to have eliminated the major disadvantage of this approach. It appears to be
applicable to all probleas involving thin shells of revolution and should be-
come the standard way of approaching these problems.

Derivation of Linear Thin-Shell Equations

The remainder of this section will be devoted to outlining the

derivation of the linear thin-shell equations in a form useful for applica-
tion of the numerical-integration techniques. The multieseent approach

26
 ai some features of the integration techniques will then be discus'*l 76r
simplicity of presentation, only the case of axisysmtric deformatibn of shel
* of revolution will be discussed. The application of the numerical rnst)bods is
practically the same for the case of axisymtric or nonaxisyamtrile loads,
and it is believed that the procedure will be better~am4.ratood, 4fhe equa-
tions art -kept simple. The more general eqgions for nnym i od
are derived in Kanins' paper. Cohen's paper gives equations for oithotropic
shells.
The underlying shei-ti~ory ai-iptions tbit 4Mi -be -used for deriving
the basic equations (in addlIon to the relative thinness of the shell) are:
(1) that normils to the shell's middle surface-defora, into normals to the
deformed middle surface without extension (this. iples that transverse shear
and normal strains are negligible), (2) that the transverie:bral 'stress is
negligible, and (3) that the components of the displacement u, and.w ani the
rotation 0 are sufficiently smau to neglect second-order terms while'the
middle surface strains, e and 9 are negligible. These assmptions formt the
(P
basis for the classical linear thin-shell equations of R. Reissner-Neissner
and Novozhilov*.
* However, In order to use the dIrect-integration approaches, it is
more convenient to express the shell-theory equations In terms of linear first-
odrequations in the quantities that appear in the rntural--boupdary condi-

tions,. these parameters are (see Figure 2) the nridional nerane and bending
resultant N. and N the transverse stress resultant
stress thQerd
ional displacement u the transverse displacement vw, and the rotationP

Tne derivation of the shell equations begins with the utatftknt of

the equilibrim equations, stress-strain relations, and strain-displacement
equations. These equations have becm quite standardized, at least for smll
axisymetric deformations. The equations will be stated In the form given
by IKminins(P0 ad may be obtained from his 3quations (6) tbr~ough (13) by
eliminating derivatives with respect to 9 and eliminating terms that include N
jI ~ p0 , Q0 , and xe(These quantities are zero for auisymA-tria dsformatiens-3"

I .. *An elegant derivation shoving the relationship between theme equations Was
presented recentiy by R. A. Clark(25).

27

7I
I -k

I r

FIUE2 XA ETO OF S -S MIDESUFC

INDIATIN ESULANTSACTIG
STRSS I
THE- rz-PLAN

28w
 The equations of equilibrium for axis~metric deformtionsre:

(1) N , + (N - r + o -.
r ]10) cos +
-P m

The lengths r an R are- shown in lge 2. mbcipt y after the

~ ~
(3)~~~~~ ~ 1- MM +aMVi.)cow
-r%+ra-0

coczs denotes the derivative vith respect to ,: N, = o $ d

the c rcumferential membrane and bending stress resultants; p and pip are. com-
ponents of the mechanical surface loads in the normol and meridional direc-
tions, and a is the component of the moment of the surface loads.

The stress-strain relations are given by:

:e
N =K(C e %(P

(5) N -K ( + ,e,)
(6) M9- D (xq + vx

There K = Eh/ ( - wj), D =h K/1.2, 1 is Young's modulus, v is Poisson's

2

ratio hi- shell thickness, eP and € are the neridional and circumferential
R-.-
strains, and andx. ore changes in principal curvatures in the meridional
and cirecu ferential directi$s because of bending.

Me stral-1s$plfaement relations are written as:

(8) 6,3 o
(V c(~ +V siny()

. (n) K, = /R,
(12) o*"~ ~os /r .
• * . P :

(1)(-
29
 In order to apply the nmerical integration approach, these 12
equations must be reduced to six equations in X~,N 1 ,~ u. 1 v, -and ~
Further, it is more convenient to integrate along the arc-length, a, of '.be shell.
This requires that the independent variable be taken as a rather than f. Is
is a simple transformtion, since:

(13) dR d

As a first step in eliminating the unwanted variables, the quanti-

ties N0 and M9 are expressed in terms of the fun' antal variables. Substituting
(5) and (9) in (4)gives:
(14) No = ,N + (1-_,v) Ke.
= vf + (1 _ v2) 1r (u Cos f
ucpsn i P)
Substituting (7) and (12) in (6) gives:

% (1 _ ) % os(P

Substituting (II4) and (15) in Equations (1) through (3) gives the
following equations:
(16) E- , +Cos ,P{(1 - V) -F(1 -v2) _.(u os y + ,, sin ,p)}
(16
+rp--0.

(1 ) rsin (u cos ( sin )= .

r-C

(1)rM
'12)+Ro (i v) (i- 91 C~bw~o rq1+ m 1P 0.

30
 Note that only the first tern in esch of thes eqW*i ".cd-
derivatives with respect to p. Using'the relation (13) aMd doing scm
ninr
rearranong gives:

(19) 1, -( -" ) + ( -1 "

)Kcs9 (, o-i ,)

(2o ) )Q
fua r R r
qiu Co C4+Vsi P

(21) N :-2-2
r !(1- v)KX 0.(1~ V2)p Cos ()+ a

j These are three of the eqiations that awe-n giled4 Te other eqft.
tions of a simlaer type are me Med, for the derivatives- of the displacests.
Solvinguation (1) forv

V
(P. =us"- R.
"ror (P sP r+
,C )
3, ,C (

Subtituting Equations (8) and (9) 1U(5), WA sayving for u -1 yields:

+C, - 1 - (P CO
u4Po + V sin CO)

or from (13):

(23) U, 3 r 10osp
CP r
The expression for P is derived by svibstitutig Iuio~ns (11)
and (12) in qation,(T) and solving forB to give:

X V
v~Cost
-- ~so that:
r

31
 Equations (19) through (24) are the desired first-order4ati-s '
for the fundawentinl variables. Once these quantities are found, the other
variables such as N9and N. ma be easily -found. (!be- requisite fow;alas-for
N. and M, are Equations (i14) and (15)).
It my be noted that no approximations were ma in-deriving the
Equations (19) through (21&) from Equations (1) through (12)." 0? course, the
original equations incorporate the "tin-shell" approximations noted at the
beginning of the derivation.
Equations (19) through (24) are solvable by nmia -Integration
techniques. Such techniques are applicable to the solution of initial value
problems In -which all of the- information about the aolution Is knownat an
Initial point and the values of the solution at the remaining points of the
interval are determined by Integrating the differential equations. -(Actually
the shell problem requires a modification In the technique sice, oaly tlbee
conditions are given at eaich edp of the shell.) It shll be. shown presently
how this is handled. The tecimiques can be illustrated by considering a first-
order differential equation:

(25) Y' -f (XoY) a :9x fb, vhere y =#/dx.

Let the initial value of y at x -a&be calledyo. It Is assmmad that YO and
f (a*. y0 are known and that it Is desired to calculate y(x),. a, s x !c b. The
nirical techniques consist of procedures to integrate Equation (25) step-
vise in =all increments of x. The first step consists of calculating
ff(x., y ) wbere x, 3z a + x, mad 6x Is som mall interval. The integratiLOU
is complicated by the fact that Y, is not known. It ust be determined
siariltaneously with f (X1 y 1 )
Two major types of integration schemas used for one-dimensional,
*initial-vaue problems are the RL-ge-Matta tjpe Proceurs and the predictor-
corrector type procedures. The Runge-MttA procedures are equiient to
expanding f(x, y) In Taylor's series about the Initial point (a, yo) and
evauating y{and y1 approximitely by a truncated Taylor's series (usually
of fourth order). The derivatives of f (x, y) are evaluated at (a, yo). A
number of variants of the Eunge-Kutta, Process exist dependin on the form--
lation of the approximate Taylor's series.
32
 The predictor-corrector schemas differ, from- theRup_ t
In tbat after the approximate vaus of Y, almd are .OR;d e4Uqhal
at xl't new valuesaOf YJand e found by "correin" :the .m~n-
ar"
the prediction step. In most of the Preditr-crrector crato
'1 be applied iteratim!ely (each tim with the uMmast vl
~can
asthmta~
fy)~tlb
exact value of-y 1 is deteramd. The, chief deficiency . _ _

corrector schemas is that as xww as four successive- values Of. .0014tiN~ r

needed a input to the difference formulas. In order to start t~a 1tesalatlu
by this scheme, it is usuaLly necessary to calculate theme severalvaisb
som other method such as the Runge-Lkitta method.
The accuracy of any of the numerical-Integration. schemes Aepends;
on the spacing, 6z, at each step of the iteration., This accuracy my be
determined with error. formulas In so cases or by integraing at each point
over two values of 6x and camparing the results. Using one or the. 9ther
estimate of the accuracy of each Integration step, it is then poassile tg
automatically adjust the step length to sctaln the desiredk acclarac In the.
integration. A umer of Integration codes have been written with this- fac~ility.
Detailed discussions of these integration techniques are contained in-a. DP*e
* of books on numerical analysis such as Fox. MW)~pof- these books. discusf.
the integration of system of equations of the type (19) -. (24).
It was mentioned earlier that the shel" preblem requires sieCift1.
cat ion of boundary conditions on both ends of the shbell.* The direct intepa-"
tion techniques asm that all. of the Infoat ion is ariven at tbe begiaing
of the shell. This requires a slight modification of- the tchbnique. Iti
effect, the modification consists-of starting out with initial values of
unity for the six variables N~ m~v . . . . y~ one by one. The shell equa-
tioms are then Integrated across the shell to get six equations relating
the values of the variables atb. the final end to unit values at the initial
end. (If there is pressure on the shell, one awre integration Is necessary
to obtain the particular solution.) 'Since three vaues of the variables are
known at each end, the remaining six unknown values are obtained by solving
the six equtions.

* The Mltieaw Approach

When the shel]. is too long the matrix of the six equations becomes
very nearly singular and they cannot be solved for the six~ unknown parinters.

33
 (20)A
a and others have overcm thiLs deficiency by first beigthe
shel into sborter segment*. 2he shell eqiations are intejiated- dMr each
shel ogmez.
injus th - sa bv
'- to get aii equations biti~
the stresses and 'diiplacbets at each -eh of each aeeu0t. Then~ the equaions
for all the pegments are solved sailtiaud3y vith the -_q ion of cciin4Uiy
for stress resultants and displacements b~etveen the stpebta and -the bouiliri
conditions at the ends*of the shell to obtafn tbi stlon~ of the etre Mhell.
A detailed description of this phase of the solution is given by both Cohaen(3
and Kalnins. (2)Use of this wultiaegment approach elminates the loss o?
accuracy of the direct-itegration appr~oach and permits the analysis of long
shells.
It is noted again that the ninerical-Integration tecbniq.ae can be
applied to Equations (19) through (9i) without azW further simplification.
Therefore the onl~y restrictions In thetypes of problems that vy be solved
are the restrictions Imposed in deriving the equations tbftales (ioe.,
emal] deflections, thinness of the khefl, etc.). This mans, for instance,
that the thickness and elastic properties can be alaie. to var in an arbi-
trary mnner over the length of the shell. It vas pointed out oari~r that
the awe technique *is used for solving the more- c~ca ted problems of non-
syatric loading except-that there are 8 equations as shown by Maain.
Again, for orthotropic shells, the sam technique is applicable although the
equations are comlicated still further an shown by Cohen. It appears, there-
fore, that the direct-integration technique is the most accurate and uiiwr-
sally applicable technique available for aolving problems Involving shells Of
revolution, such as bellows and diap8r080s

341

-L
 Iasi=_________

Little is ]mown about thi datiidle of the procedures tsed by asnu-

facturers to design bellows and diaph~ragm Because it Is difficult to -predict
the performance of new configurations, the ability to develop satisfactory
designs directly affects the competitive position of a copaW and the methods
used to establish design details are considered proprietary. However, manu-
fazturers have been willing to discuss certain aspects of their do-sign pro-
cedures and sone information has been gained from the literature. This section
outlines the general design procedures which appear to be in use by manufiacturers.

Design Procedures for Bellows

Manufacturers must predict the following bellows performance charac-

teristics: (1) stresses and deflections, (2) spring rate, (3) effective
area, (4i) pressure drop, (5) instabilit:j, (6) resonance, (T)corrosion, and
-~~ (8 operational life. All manufacturers request the opportunity to design o
select the proper bellows for each application because oftthe r4y factors
hat must be considered. However, approximate performance cherazteristics for
st dard bellows are available to potential users in company-brochures, and
j\ des nographa are soumetimes included. One ccmmny offers a design slide
rule, v1'1Sle another ccup-an provides extensive de-sign charts. The procedures,
used by th ufacturrs to produce these design lids and to design specific
bellows atre d cussed briefly.

Stresses and Defi Ons

The general met"0\1used bpy manufacturers in developing dtesifff pro-

cedures for predicting streafts and deflections is to select form the litelra-
ture a theoretical approach which most closely approximates the typas cf bellows
to be manufactured. Available appj' baches (acme of which are discusael in
the previous report section) are base&a\on plate,, beam, or shefl1 theory. Be-
cause the approaches are complex and do tiqt usually describe Lecurately the
bellows to be manufactured, most ccinpsnies ~bave simplified s.nd modi-fied4 the-
IV
theoretical analyses of particular configurations.

35
 When a theoretical apprxlmtion.has been selected, the comga
produces bellows of the type to be manufabtured and the performance of the
actual bellows is compared against the theoretical prediction. ftreso-coating
analysis and strain-gage readings on bellows are comon means of knalyzin
actual stresses. Empirical correction faztors are then formulated to make.
the theoretical approximation agree as closely as p6ssible with the stresses
and deflections of the actual bellows. The empirically abdified desIgn pio-
cedure is then used to design similar bellows.

The most detailed nonproprietary design procedure available has been

devaloped by Atomics International 6 ) The basis for the procedure is discussed
at length, and design calculations are presc-nted for three configurations,
ie., (1) U-shaped, (2) U-shaped ring-supported, and (3) toroidal. Comparison
of the results of the calculations with bellows produced by any given company
would probably indicate the need for empirical correction factors for the
bellows produced by that company.

Many companies have accumulated experience with a large n=ber Pf

bel" --
s sizes and types, and have developed extensive design relationships
whi. can be used to predict the stresses and deflections of,.belcrs similar
to those that have been produced. However, if any company undertakes the pro-
duction of a substantially new bellcws configuration, it is at present necessary
to conduct extensive laboratory tests to establish new empirical correction
factors for the theoretical approximation.

Some companies have begun to use computers to assist in bellows

design. In general the computer programs have been limited to the automation
of existing design procedures. A few companies have utilized computers to
incorporate some of the basic theory, and it is expected that further develop-
ment of the theory will result in a wider use of computers for designing
bellows.

Spring Rate

Although the theoretical approaches used to predict stresses can also

be used to predict spring rates, the spring-rate predictions may be even less
accurate than the stress predictions, because spring rate is affected by

36
 inaccuracies in all of the convolutions, wil'e maximum stress is a restlt of
effects peculiar to each convolution. In most cases, manuNacturers predict
the spring rate of a bellows by means of equations derived from the measured
spring rate of a similar bellows. If an accurate spring-rate vtlue is needed,
it must be measured for each bellows. The effect of hysteresis is almost-alvays
determined by actual spring-rate measurements, rather than by prediction.

For approximating purposes, most manufacturers list spring rates fior

standard bellows sizes and materials. In the literature, Blair 2 le a
28)
generalized spring-rate formula with typical exponential values. Nathemy(
presents spring-rate formulas for seven typical bellows convolutlqns, and
descibe
by ndeson(6)
the design procedure described by Anderson includes both axial and rotational
spring-rate calculations for three bellows convolutions. These relationships
are useful for approximating the spring rates of certain configurations.

Effective Area

Sr If there is a pressure differential across the bellos, the bellows

acts like a piston, and the area over which the pressure apparently acts is
known as the effectoe area. The effective area is approximately equal to
an area calculated from -he mean diameter of the bellows. Manufacturers'
brochures no=lly give the effective area for standard bellows. However, when
this quantity mu-st be known accurately (for example, for face seals) tie
effective area is measured for each bellows. For some welded bellows, the
effective area changes with extension and compression.

Pressure ]o2

Most manufacturers estimate pressure losses in bellows and flexible

metal tubing from test data or from calculations based on flow through a
pipe with an assumed equivalent length or an assumed roughness factor.
Daniels 29 ) describes the calculation of pressure losses based on an assumed
roughness factor for two-inch and three-inch annular and helical flexitle
metal hose.

37
 Hawthorne(0 presents a relatively new method of calculation-based
on the assunption that the convolutions behave as uniformly spaced ,orifices.-
and that pressure drop is caused by a succession of individual flow expansions.
This approach has been investigated 1,n some detail at Plexofics..
For iternally-linked bellows, Daniels(31) presents experieit i _-
determined pressure-drop factors for two types of assemblies: the cbiAiink
type, and the gimbal-ring type. This article exemplifies the emprial
approach which is often used by -,Lnufacturers to determine bellowo perfomnce
characteristics, in place of a more complicated and. perhaps less accur'ate..
theoretical approach.

Instability

A bellows my become unstable as a result of internal presspre .if.

the ratio of bellows length to diameter is greater than one, or if the bellovs
is offset or angularly deflected. As far as can be determined, these .condl-
tions are not predicted theoretically by most manufacturers. If there appears
to be a possibility of instability, tests are conducted with the bellows of
interest mounted for the test as it would be in operation.
In 1952 F.rfrgx'~ developed a mthod fort cal~.ilating the Pressune
at whi!:h a bellows may bedoue unstable. Be concluded that the pressure buckling
of a bellows is closely analogous to the buckling of an axially loaded colwrI.
Although this approach has been referenced by subsequent discussions'of initabil-
ity and is reccmended by Winborne'(8) , there is little evidence that the
mthod is widely used by the manufacturers. Newland ha(vryreenl
extededHaringx' theory to the more complicated case of the universal bellows

approach is not eetivetgtosb 8iom elde(3)aa nesn6

concerning the instability of bellows under internal pressure 'when angularly
deflected have produced some interesting results. Hdvevao, it is not known
whether this work is used by manufacturers.

Resonance

A bellows may be placed in resoniance by vibration from the sup-

porting etructure or by movement of fluid throuagh the bellows. Flow-*indUCed
 vibration, which usually occurs in gaseous systeos, is predicted from test
data. Approximations for resonance due to forced vibrations can bVe mde by
calculations based on the bellows functioning as a spring. Vail - 3 5 dis-
S-cusses the calculation and laboratory evaluation of the effect of extension,
compression,-and fluid damping on the resonance of a welded-bellows. An-ex-
perimental and theoretical investigation of the forced vibration of bellows
was included in a study by the Bell Aerosysteme cOMa ( 6 ). Both longitudinal
and beam mo-des of vibration were considered. In &=itIon, consideration was
given to a vibration mode in liquid-filled bellows caued by the pumping action'
of the opening and closing of the individual bellows convolutions as they
deflect. In general, however, the vibration characteristics of bellows and
structures are sufficiently difficult to predict that laboratory tests of each
installation are normally required to demonstrate the susceptibility of the
configurat in to resonance. Rekate and Schwartz ( r ) describe such an
investigation.

Corrosion

Bellaws material is so thin that corrosion rates such &s those

associated with the chemical industry cannot be tolerated, and highly corrosion-
resistant materials must be utilized. However, highly corrosion-resistant
met-.s may suffer from one or more of the following: (1) crevice corrosion,
(2) galvanic (dissimilar metals) corrosion, (3) corrosion fatigue, and (4)
stress-corrosion cracking.

In bellows design, care is exercised to eliminate crevices and

other areas where moisture and other environments might be trapped. Attention
is given to the elimination of couples, particularly those having widely dif-
- ferent galvanic properties. Sx h metals, although highly resistant when used
individually, may be incompatible when used together. Corrosion can also
reduce the fatigue limit of ma metals, and this factor must be considered.
-. Any aerospace materials, especially the high-strength alloys, are susceptible
to stress-corrosion cracking in environments such as humid air and chloride-
1containing atmospheres. Configurations are designed to keep residual and
applied stresses below those known to promote cracking in specific environments.

39

4
')erational Life

In the absence of corrosion (and neglecting creep in high-tmperate

applications) the operational lifetime of a bellows is determined by its -
ability to resist failure by fatigue uader repeated stressing. The fatie
lifetime of a bellows is difficult to predict from fatigue data usually:
available for the bellows material. -The available material data are often ,in
the form of S-N curves from constant-amplitude uniaxial stressing of spocfr.ins ,
in a well-defined metallurgical condition and with good surface finish.
Operational behavior of the bellows mW.involve:
.1. Complex loading: cycles of varying amplitudes of pressure or
of deflection, or of bath

2. Complex stress distributions: varying degrees of biaxiality

and of stress gradient at various locations in the material;
details are not always easily comwutible b-om the applied
loads

3. Complex effects of fabrication: varying surface conditions,

residual stresses and avmount3 of cold-work.

There is no satisfactory theory for including all of these complexities.

Accordingly, semi-empirical rules are used for preliminary design.
Tnese vary in degree of sophistication among the various manufacturers. The
rules often in.lude such simplifications as:
1. Neglect of, or very crude allowance (by some sort of strength-
reduction factor) for fabrication variables
2. Use of such approximations, in fatigue-strength analysis, as a
maximum stress (or a maximum shear-stress) criterion of failure,
a linear Goodan-diagram, a Miner-Palmgren cum Alative damage
rule.

For design to resist low-cycle fatigue, there is increasing use of the rules
being developed for pressure vessels (see, for example, Ref. 36). -There are
divergencies among manufacturers and users in details of alro.-ar ations and
even in nomenclature.
Similar rules, with one or more coefficients to be evaluated from
load-cycling tests on typical bellows are used to provide more realistic
operational-life-prediction charts. Such charts may be developed in terms
of percent of maximum allowable deflection and of percent of maximum allowable

------
pressure. If there is a reasonable &mount of experimental data, the charts
may be quite dependable for interpolation; however, the empirical nature of
the guiding rules makes unlependable extrapolation beyond the conditions of
test. Moreover, although charts from various companies may be similar in
appearance, predictions are often applicable only to bellows for which a 2hurL
was developea.

Desi%!Proedares for DiaphraMs

Performance characteristics for diaphram~is are not generally avail-

able to potential users. Not only are the diaphram design procedures of
manufacturers considered proprietary, but most diaphragms are fabricated nn:d
used in zompon-ents by bazh manufacturer, and ther.e_ is generally no requir'-
merit for a component user to be concernd with th-c diaphragm. Howev-1r, din-
phragm manufacturers welcome the opportunity to d.-sign special diapharagn
con1igurat ions.

Tine design of diaph=agus, as with bellows, presently involves n

bination of analytical and experimental approache-s. In the analytical approach,

equations are derived relating the load-deflection c!harateristics to the
geometry of the diaphragm and the elastic constants of the mwterial. The
experimental approach is to make diaphragms of various sizes, shapes, and
materials and to load them in various ways and measure their deflection, drift,
and hysteresis. From julicious selection of sizes and shapes for experiment,
it is possible to deduce design criteria and generally applicable formulas
which may be used to design other diaphragms of similar shape and materials.
Informtion available in the literature on each of these approaches is sunn-
marized briefly below.

Analytical Approach

Althoug:h mathematical description of the performance of corl-119ated

diaphragms of arbitrary shapeb is difficult, considerable anta3sls has been
carried out for diapbragms with concentric circular corrugations subject to
"ujiform loading. Two approaches have been taken toward the difficult appli-
cation of shell theory to such configurations. 0c involves the use of shell

141
 theory directly. Tae other involves si~lifying pbys~cul asmaitions such
that approximite calculations can be*md.-

Application of Shell Theory. The firs., attempit of this type was ade
by Grover and Bell~ by piecing together known approximate solutions to the
shell equations for elementary seridional'shapes having constant curvaiture'.
Their results gave information for stresses and m ents, but not for dis-
placements. Wildhack, Dressler; a~i Lloyd(2) ,ssquently developedl an anal-
ysis procedure involving the use of a digital computer and compqted stresses
an! deflect ions for the configuration investigated by Grover and Dell. This
procedure, whiich is based on eat elastic-shell equatio~ns, can be used to
analyze stresses and displacwu-ts for any corrugated diaphragm haiing arbitrary
meridional shape. Dressler(9 then described the investigation of the Grover-
Bell configuration for thin, medium, and thick material by mans of the com-
puterized analysis method. This paper shows how thickness influences the
coupled bending and stretching quantities in the diaphraga.

Use of -Simpified Phzjsical Assuptions. In this apooch apprcci-

m~tions are made such that algebraic solutions are feasible. Mrh- 9 bO, Chro(0
described two diaphrasa types which wre determined by the corrizgation shape.
He theorized that In both types., the elastic stress 'which correpods to a
definite deflection of the diaphragmt increases in approximate proportion to
t'he thickness of the metal. In 19!47: Pfeiffer described a hypothetical
Rodel for the distribution of stress in diapbrafts. It is constructed of
f members in series, -iheahconsisting
of one flexural and one tension spring
acting in parallel. This bypothetical model leads to a correlation between.
pmre'sure on the diaphreaa corresponding diaphragm deflection, effectivre dia-*
meter, thickness, andi "plate modulus", a measure of elasticity.
An ingenious approach by RaringKc is based on the analysis of a fic-
titious flat plate which is "equivalent" to the corrugated diaphragm. In
tresucessive pape.rs (4s24)he calculated the rigid~ity, the stresses, and,
for large deformations, the degree'of nonlinearity of The relation between
deflection and load. Subsequently, Haringxc diescribedI furtber bow the

introduction of a fewr simplifying restrictionis laads to 14he concept of aA

142 -
 design chart which can be used to give the dimensions of a diapbaim to meet
." specified requirements. A design chart is given, and the solution of a design
problem is described. This zhart is very useful Tot-n diaphra.
- configurations. In 1956, Flindt ( 5)
extended Ikringx' theory to include the
calculation of the thrust which can be exerted by a restrained diaphragm
under uniform pressure conditions.
In 1958, Akasaka and Takagishi (46) analyzed the natural frequencies
of the corrugated diaphragm as an equivalent flat plate, and derived approxi-
mate formulas for the lower-order frequencies of symetric and awmnetric modes
including the effect of the concentrated mass at the center of the diaphragm.
Eperimental results agreed well with the theory.

ExrtiCental Approach

The literature contains the results of two primary sources of expe-ri-

- mental work: .(1) the investigations at the National Bureau of Standards, and
(2) work at the Taylor Instret Company.

The National bureau of Standards has maintained an interest in

diaphragm design for many years and has cooperated with other government
agencies and industry in several investigations. The results of the major ex-
perimental work are sumarizedin the paper by Wildhack, Dressler, and Llaid(2)•
The pressure deflection characteristics of corrugated diaphragms are correlated
by methods of dimensional analysis. Experimental results indicate that the
performance for diaphragms of any given shape may be computed from a dimension-
less formula derived from experimental data on other diaphragms of that shape.
The dimensionless formula is given, and its application to various diaphragms
is discussed. Although this work has been relatively extensive., it is not
known whether the results hive been of value to diaphragm manfi4cturers.

The most complete discussion of diaphragm design contained in the

literature is presented by Nevell(l ) . Wch of the information is based on
experimental work conducted at Taylor Instrument. An effort was made to
show how diaphragm characteristics are affected by changes in uateritl, dim-
ensions, and treatment. The first part deflines a diaphragm and its performance

43
characteristic si methods of inmeawng and representing thechrtitcs.
anid ways of using then. The second part deals with vinys In iitich. diaphwar ,
can be constructed and how various design details and treatments affect per-.
formance characteristics. This maal. is probably as clear-an indication as c-'n
be obtained of the design procedures used by diaphragm mufacturers.

.j ..... . ....
 FAMICATICN -TWAIKLO~
i- 7_

Tae extent to which the service Perfore of bsllws. eM. 4iqhmi

achieves the~d~i t~y
the materials frcmi which they refbict&and.b-he.cm*44nthefii
facture. Material defects, weld dicniute -cnt
and postfabricat ion damage car- all result in locally hioi stresses that =W
lead to premature failure. In this section, the present state of the -s~rwith
tomaeral manufacturing methods, and'inspiiiioa wifl be dicussed.
resec

Mterials

Typical materials for formed, welded,. and deposited bellows are de

cussed. So many materials are suitable for machined bellows that .they are
not included. Diaphragas can be made frcm any of the zaterials listed' icr
formed, welded, and deposited bellows.--

Formed Bellows- Materials

Materials for formed bellows, shown in Table -3, mast be veldable

and formable. Althwgh smaller bellows are usually made fromssaliqns tubIng,
mzst bellows over an izrch or two in diameter are ade from sheet or strip
formed into a cylinder and longitudially seppi vede-I. Whether the str~p- or
sheet is purchased to any special tolerances depends upo -the end application
of ;the bellows. When the spring rate is not. critical--bellows intended for.
eypansion joints, for example--the custcmgary 10 percent mill thickness toler-
ance is satisfactory. When the deflection characteristics mist be more care-
fully controlled, materials may be selected from warehouse stock. In this
way,, thickness my be controlled to within about 5 percent on a given order.
fterolleld materials from specialty metal fabricatfors provide the be'st covwr.-
cia2lly obtainable thickness tolerances. Bellbtf indd~turekz using rerofled
materials claim thickness tolerances of ±0.0001 inch. A more c~mmoly qwed
tolerance is ±0.00025 inch over a 20--inch strip width.

45
 TAK 1E 3. TiPfCAL FMMUD.LMI NAIUIAWS
- t

AlliersaLtz~tw -~

Solutidn-strergh4e.~

Alpha brass Face-centered cubic

Copper ~
Stainless steels, Types 30,321,31
Titanium, commerele.Uy jpure and Plpha
alloys Hexagonal close packed
Theonel Face-centeredi cubic

Pi-ecipitation-har&ened alloys
Beryllium copper Face-centered cubic
In:!onal T18-
Aluminum, 7075 alloy

Trnfrainhree lo

Titanium, Alpha-Beta alloys Hexagonal close packed,

boy-c~~~c

iaporoveaent in the fatigue life of bellows produced frcn bright-finished

mae . (IM). 2-3 finish), while others see no difference. Sin maufacturers

Thre twiw th us ofbright-finished material arising from

forms atred
the allegedi tendency of Ins'etds to pass shiny bellows and reject dull
bellows.
s ome diagremet.most manufacturers see no
Altaag thre
effect of "grain", or preferred orientation,, In their starting-maLterial. The
common materials from which formed bellows are fabricated. when produced
under good control, are nearly Isotropic, so pre1rprrcd orientation does not
appear to be a serious problem*

46
 61d --- i a-tiai

.14WeI&r-3eU"
terial
fbi - d medIatIWt-
- -f- 4 46

materialsfra for l edzsb~eftws V2 id''tiatex iof-&t--

ormb~iiwty

above for formed bellows, a variety of less tra-taObbe alloys i's cdWoqty u*&
for veldedbellows:Table 4shows tpical velded1 ~
lr4 TA ~
Welded blosaepoue ncnieal mlrta

formed bellows and tend to be used in kom specialized applicatkzas., AU -of

the alloys itemized in Table Z4 for exmle are hesttresistnt. ally4 Ue.
of rerolled, bright-finished starting materials is standsmrA in the wle.

I
bellows industry.

TAKE 4. TYPICAL WELDZD.R&LGIWS ALLQY5

Alo - -rsa rutur -

I ciitatin-arenedl. cf
Inegnel X ft-ce-centeredcui
Renie 41
-~]Rei4 62
17-7 PH1
1T-4 PH
PH 15-T X3 I
14-252 t

Waspalcsy
UIimect 70O

Souin nL~hnd!~~

19-9 IL ?ae-centered cubic

A-286, __________ eterel cubic

Depo2sited-Ballo-is Xaterials

Materials for deposited bellows are made by electx'olaing: or chemi-

4cal depos.Ation. The most ccuon at-rial for electrodeposited ballows is
nickel. Copper is also used. Ciieiy deposited bellws~ can be made from
%.1oys which, though still over 90 percent nickel, contain sa41ficant
 '-

percentages of other strengthening elements. Both types of deos ed bellows--

can be made with ::omposite metW. walls consisting of layers of different -"
metals. The deposited-bellows industry is relatively you, andf;ther
developments in deposited-bellas materials can be expeted., - ,

General Material Limitations

All manufacturers- of bell s and diaphragds use 3Felat±*4 s-A&alI

tonnages of mateiial. Consequently, they are able to purchase ony i*terUs

melted according to usual comercial practice. Satnless steel*, foi -ixp e,
are from electric-furnace melted heits.. Although the ell {i{dTal-ii
of stainless steels is good by ordinary-standards, it is lkeIythdt'sm"e
improvement of bellows parformnce could be achieved by t euse of vacuum-
melted materials with their lower inclusion contents. Such mterials have
become standard in bearings. Quantity p'rchase of heats of vacuum-meted or
otherwise specially treated mterials by large users of beUo(s, and release
of these materials to bellows manufacturers might-be a method-of-significantly
improving uniformity and quality of bellows throughout -the iztdudtr.-

Manufacturing Processes

There are four primary kinds of bellows: (1) formed, (2) welded,
(3) deposited, and (4) achined. Typical processes for producing these ballows
and for producing diaphragms are discussed briefly. Pecause the coaoeitive
positions of bellows and diaphragm manufacturers are significantly affected
by differences in manufactdring techniques and .equipent., it is .not-possible
to determine the details of the manufacturing processes. HIfoever, it is believed
* that sufficient information has been obtained, for the purposes of the.progra2
and for this report.

Formed ?ellows

A flow chart -for -manufacturing -for d.-b .lows-is swn in -Figure 3.

The process begins with a thin- netal cylinder.. For small bellows, the mtal

48
 . t- -

Clean - . :. "
Weld- .
(UJ21e18a seamless)

Planish Weld Sean.

• ( O ti al)° .." ' ..':

Clean

Assemble Plies...
(if Multi-ply)

Form Convolutions

Repeat if necessary

Process-Anneal

Cleani ""' ,
I Set Convolution ;
Spacing
(if required by process)

' L
Partilly Eid
6lj
4_____
•(if re t.ed) +.

I
'!_. ! C
cq~l ete
En d We
ld1

] Clean ---

FIGURE 3. MAqUFACTuRING FLw SHEET--r)E iLw

149
 cylinder is usually a seamless tube. For bellows hs.'ving diameters much in
excess of an inch, the cylinder is fabricated from fiat; sheet, or str ,k.
having a high-quality surface ar.d coiitainiug p6 visible damaage to..h,- eddes.
After the sheet has b' en cut tt size by a shearing operation, it isa roll-form.
to a cylindrical shape. Typically, tChe cyJlnder is some'what overformed in
order to assure that the edges will meet satisfactorily. The formed cylinder
is then placed in a welding fixture consisting of a rigid backup and hold-
down clamps, and a butt weld of the gas-tungsten-arc type (GTA, also !cnv'n
as TIG) is made along the mated edges of the sheet. The technology of making
such welds is well advanced, and manufacturers are capable of making welds In
matecial as thin as 0.003 inch. Most welds are longitudinal, but one manufac-
turer uses a helical weld feL- =rAll be'llows.
Before the GTA weldftng;process technolgy had been developed siri-
ciently to make w~lds; in very thin umaterial, longitudinal weld seamns were
made using a resistance-lap seam-welded Joint. Although some- bellows are
still made with this type of j ,int, the overlapped region is a stress raiser
and is a potential source of : apurity entrapmenxt.
YAny manufacturers coldrwork the weld zone with a pair of crowned
opposed rolls in a planishing operation. Planishing must be carefully con-
trolled in order that the wall thickness In the vicinity of the weld zone is
not reduced below the base-metal wall thickness. Some mnfacturers do not
use planishing because of the danger of wall thinning, while others use It
only for cettain isterials.
In the fabrication of multi-ply bellows, a series of tubes,. sized
to fit one inside another, are cleaned %nd-assembled ready for the forming
operation. The cleaning at this stage is particularly important since it Is
exceedingly difficult, if not impossible, to removei contaminating materiatls
that have been trapped between the plies once the bellows has been formed.

Almost every manufacturer uses a unique formin# machin e oftpro-

prietary design. Although these machines fall Into sevei#l basic categories,
there are differences in detail which may significaritly'aifect the performance
of the fabricated bellows. The basic categories~ of formink maciines are as
* follows:

50
 - - ? - - - !-

-. traulic,, individually formed convolutions

". drostatic, rubber pressure medium .

j 4. Mechanical rolls
5.
Mechanical e x ansion tools.
,

In the hydraulic process with simultaneous c6nvolution-.formation,

the ends of the tube are first closed by movable platens. The end sections
of the bellows are constrained in cylindrical dies that may be part or the
platens. A series of split rings, one less than the number of convolutions
desired. is carefully spaced along the length of the tube. Wdrdaulic pres-
sure is then applied to the interior of the tube., causing the tube to bulge-
outward between the split rings.

iRrom this point, processes of different manufacturers differ.

SoCm manufacturers leave tha rings in place throughout the entire convolution
formation operation. Some manufacturers attach the rings to a pantograph
*- during forming to maintain uniformity. Others remove the rings completely at
this point and complete the convolution formation with the tube eairely free
. of restrictions except at the ends. This latter method is claimed to
be advantageous, since it requires a minimum of contact of the tube with
metallic tooling. During the formation of the convolutions, the platens mist
be moved together to accommodate the shortening of the tube. Some manufac-
turers acco*plish the movement of the platens and the regulation of the by-
draulic pressure by hand, whereas others have applied automatic controls to
the process. Automatic controls are desirable from the standpoint of product
uniformity.
It may be necessary to form the convolutions in several stages,
depend-ng upon the material and upon the depth of convolution desired ?ela-
tive to the tube diameter and wall thickness. Some manufacturers process
anjeal their tubes following the initial bulging operation. Others fina it
necessary to stop several times during convolution formation, remove the split
dies, process anneal, and reassemble the tube in' the forming machine. Still
other manufacturers restrict their product J.inp to convolution depths that can
Sbe formed in their materials using a single operation, thus eliminating process
annealing.

51

04
 yvee 8- bebhe..- -
, -.

equipment of somel mwAi.scCur,_. . i. -f. '~ da

A of forang -- *en
process anneals than does equipment o, '.hlr ;turers.
Some manufacturers form each. convlrkion individually uhinig essen-
tially the sawe pro:.ess as described above, but with the hydraulic fluid
confined to that region of the tube where the convolution is to be~ formedi.
The tube is first bulged. Then the external clomp holding the unformed por-
tion of the tube is moved forward a preret distance to form a ccnvolution.
The operation ie repeated after the tube is indexed zo.the next. convolution
position. It is claimed that this forming methol.gives more uniform convolu-
tions than the methods in ithich convolutions are farmed simultaneously.
A variant of the hydraulic process is one in which the hydraulic
oil is replaced by a rubber form. Under pressure, the rubber acts as a hydro-
static fluid. Its use eliminates the presence of oil. If rot completely
removed prior to process annealing or final heat treatment, oil cacus
carburizr -. 7 n and possible eirrittlement of the metal. Residues from uii have
also be,,:, kr._4i to cauise pit-type corrosion.

Perhaps the oldest method o.; forming bellows, is that of shaping the
convolutions by mechnical tools wbile rotating the tube. As ba the bydraulic
processes, there is considerable variety awons the machines for roll forming.
Some roll-form tooling resembles a lathe on which the tube to be ford in
slipped over a centered rotating grooved diee An external tool is then used

5' I to press the tube into the grooves in the die, one groove at a time. Another
type of tooling makes use of two small coaxial wheels over which the tie
placed. While these wiheels are rotated,, th'us rotating the tube, a third wheel
is

is brought down between the other wheels, thus foraing a convolution., The
tube is then indexed one pitch distance, and the operation is repeated. Con-
siderable inf~vnuity by the manufacturers who use the roll-forming process
has led to the ability to roll-form the convolutions cutvard as well as inward.
However, roll-formed bellows are currentlyr in disfavor because of the danger
*of creating surface defects, and of sakearing mtal over these defects In such
a way that they are hidden. A second objection that is often cited to roll-
formed bellows is the excessive wall th.nning at the roots or crowns of the

52
 convolutions that may be encountered if forzing in Dot d~ne carefilly enObiA

Iny ive ete ofabelcate faiimes of~T th inl Moais mccc s ieiy .

strinshadtingnorfue .toe oll-girme~dix'ely atd -o s:

ftrmngh theramont.o terai hardv'_6totg mi -t i6

strngt.
Te ao~m o stainhar! ~gfor a give! mount of -thinning
depends in a subtle yqay_ on the -precis*,;&_ta4U1 6f *~metal flow during
forming. The relationshipqOf f thinii-mig,2stii tardening, formability
!3imits, and metal-flow directions amre orly underc ad as they apply to
bellow fabirication. -

When *Amulti-ply bell.Wi Ii9biafrmed. oe current -practice con-

sists of partially closing the ends of tbd ~fl~IME b~ss~i em
Si1
welding or GTA welding. The- iys i_ aen given an elevated te.-
parature treatment in an efrtrto. remove Iai* ccki: mifmticn from betVieen the
plies. This heat treatment may also be the final Ig1ag: treatment for the
bellows material. After heat-'treament--the-end- zelds are completed, n h
bellows is ready for inspection,.

Welded Bellows

The- manufacture ofwelded beLlfokms, as sown in Figure Ibegins

with the blanking of disks, ar di~ap 'frcu sheet- material. Ma blanking
operation must be carefully done using dies4that.-arein-gooizajstment to
minimize the formation of burrs.. Any-Miars-which--are formed on the edges of the
diaphragms must be rewyoed. _Dendin1QLbe..Asgn of the velded bellows,
the diaphragms may be formed into a shape containing-,-me or more circumferen-
tial convolutions. Tbe forming operation is carried mitl using conventional
metal-working te ~q 2 zf
U ith ca~~d_ te~ilbl4' gfi to maintain-
ing the surface finish. Diaphraga thickness is amitored, and diaphragms for
- - a particular bellows may be preselected for uniformity of thickness.
Pairs of diaphragms are placed together in a welding jig with the
inner diameters in contact and clamped with chill blocks on either side of-

53
 Clea

IAse-abie Diapbragm Ptir La,

In ID Welding Fixture Rpa o
-Requirea haer
of Convouions
Mkec' ID Weld to
Formmv~oltioni
Clean

-in OD ielding Fitmre

Make ~5~d

Weld Ezii Fittings

Clean*,Tea

FIG;MR II. XAWJFAC-iIff.?LOWSHT--:YMDeD 3=.MS

i-I -5-
 the jIoint. An edge weld is-then made around the inner circumnference usingZ
the ga-tungsten-arc process (GTA). The wlded-pa -,of &iapkra -trer j~d
* ~~ to as a convolution. Tne welding. orertt~on 1s, repeated for. -the n rof~con
volutidns desired in the bellow.. The convolutions are then stacked 4n-am*e
* welding fixture with the other diaaetera of adjaent convolutions in _cppnt,-
J chill rings being used between the mated surfaces, and the outer .di~t~ers -

are welded in the samef manner as, the Inner diamtersc!

I An alternative isethod of making'velded bel-Lows, adding 'diajpwa*s

one at a time, does not differ materially from the previously descrlbed--method
except that it might be somuewhat faster. In this method; alteknat-e Uixe-
diameter and outer-diameter welds are made until the desired iuer of c~nvo-
lutions is reached.

I Alnost all welded-beliovs manufacturers use a semi-autmtic form-

of the GTh welding process in which the material to be welded Is rotated
beneath a stationary torch. Up=n ccinpletican of the we if thefxuei oe
I. to the next weld posit ion and the process Is repeated. An automtic bellows-
welding machine has been publicly announced by G. H. Silver and Associates,
Nawtonville, Massachusetts, %7nd it is expected to be available for public 'iale
* by the end of 1965. It is claimed that this machine -rill be able to produce
welded bellows priced c=tktive1y with formed bellcws.
Welding difficut-tes that occur in iwelded ie~llows are related to
the bellcwis materials, s-e: of wbich are not as weldable as the alloys used
fof- formed bellows. Hiet-resistant alloys, most of which are vacuum melted,
typically contain two or =are phases and undergo various solid solutionm and
precipitation reactions &wring the thermal cycle associated with welding. In
some alloys these reactionas may result In loss of dutility or stren.3tb in

the hsant-affeCted zone.

A,, least one welded-bcallows manufacturer has introduced a two -ply
welded belaois. Althongh the marsufacturer declined to discuss the details of
~ I !manufeture, the isleaning reqliircme-nts for the &Iaphrawns are sore stringent

I than thone fwc~ onvez-Ftioal welded bellows, and care must be taken to avoid
leaving contAamiiwtiug ma.-terials trapped between the plies.
are similar to those us~ for formnd bellows.
Methods of cleaning

Making of the first weld in a multi-ply welded bellows would appear

to preserit no problems otber than establishing proper alignmtent of the diaphlragms

55
 and adjustment of the cick-6ut 6f the e4es 'beyozid the chili b1ei.toice
the vald Is phys~ca1. lareer thkn tor -a inge-] bellowlisi ~ irr~
* must be greater. Ce'-. Letion of the second weld, vibichl biimhe.V@
tween the two jprei~ r,C diaphrir' May' p n t czigrc-Prleis, Ifoeer u
to blowing of a hO* in the wiild Puldle doi' letvii a tus~sufae ''oi-&at the
point of weld clcz' re. Among the sevrei'l *61%-hoda of overcufr-g-"hitsA fcUty
are (1) making certain that there are- -no 0r'g~nic outgassing: mterials present,
(2) increasing, th-- chill-block, clamping forde -to -seal gasers bttee&-the -T11es
£so they cannot reach the weld, and (3) use.'of electron-beam -welding-in vacim'
frtefinal weld oa each diaphragm.
jo ave made welded bellows using electrn-beam
hnfcurr
welding, but the reported results are not consistent among then. Electron-
beam welding is potentially a cleaner process than GMIA welding.. -It v&Uld
seem, however, that difficulties In making welded bellows are- due..ore to: the
difficult materials being welded than to the welding process being used.-

Dapsitei Bellowts

Deposited bellows are produced, as outlined in Figure 5, either bzf

electroplating or by chemical deposition. Both of these processes offer re4--
dcx2 from several of the restrictions on formed or velded bellows. Very thin
walls can be achieved (0.0033 Inch), thus resulting in bellows having extez'mty
low spring rates. %Ellowrsneel not be axisymaetr~c and ay be of variable
cross section along their length. Althouv the ba..e material for del-zS-Ited
-bellos is ueuaflly nickel, it is possible to produce chumical ly depositei
f
bellows of a nickel alloy, and it is possible to uae composite wail structures
in electrodeposited bellows by plating successive layers of different metal-=.
Deposited bellows are made on mnandrels that have been machined to
the precise shape desired, and apparently 6"6 alusirnm alloj is universally
used for the mandrel material. A sepsrate mandrel must be used for each
bellows, and the mandrel will be reflected in the structurc of the bellows.
* Tne expense of machining and finishing of the mandrels to the rigid quality
requirements necessary iz a major cost item in manufacture of deposited bellr.s.
After the bellows have been deposited on the mandrels, the miandrels
are dissolved away, presuisbly In an alkaline solution in order not to damage

56
 machine els

F Polish Mandrels
Cleam

Dimensions and Finish

F Prepae Solution

FDewsit tiellows

IClean -
I.--
Solder EdFtig
Clean4

Inspect

PIGM~ 5. KMA7M~RING FLGV SHE-1-DEPO6IM ELmc

5T
the bellows.. The bellcvs may then be gold plated. although this is not e~sen-
tial, and end fittings are soldered in place. Soldering is used because of
the thinness of the bellows walls.

Diaiphrams.

Diaphragms are constructed by spinming,.. ta&mVng, or hdrst Ic

pressure. M spinning is done in a lathe by pressing the metal against a
Te
corrugated form. This results in a certain amount of cold working which il-
proves the life of the diaphragm. Som manufacturers stamp the diaphragms
first and finish them by spinning.

In the stamping process usually two mating steel dies are used.
Some dies are made so that they make ;ontact only with the material on concave
sides of the corrugations. The depth can be adjusted thirough s wide range.
The die can be made such that the corrugations are formed in succession from
the inside to the outside, thus drawing the material gradually from the out-
side. In order to reduce friction, a lubricant may be used btteen the mater-
ial and the polished die.

The hydrostatic process is similar to the pr6cius used to form

diaphragm for welded bellows. -A metal blank is clamped rnainst ths die and
hydraulic pressure or pressure frc-cm steel-backed rubber forces the blank
against a corrugated die. Small bleed holes relieve the'pressure tetueen the
blank and the die.

The .material may be heat treated prior to forming to make the material
more easily worked. After formation, the diaphram may be heat treated to
reduce the residual stresses created by the forming operation. The ty'jr of
heat treatment required before and after forming is a function of the mterial
and of the diapragm shape. "
A single diaphragm may be used or two diaphragms my be fastaned
together to form a capsule. In either case the diaphragm will be fastened at
the outer diameter and suitable fittings will be attached to the center of
the diaphragm. A diaphragm used singly may be clamped at its outer edge.
Fittings are attached to the centers by a variety of processes including sol-
dering and spot welding. Welding of the diaphragm edges is usually accomplished
by O1TA welding, although considerable experimenting is being done to adapt
electron-beam welding for this purpose. Electron-beam welding is particularly
attractive when the diaphragm capsule must be evacuated for service.
58

I _________________________ ________________
 Bellowsr an iphaa re inspected diring and after ftctin

fr-process inspection is fttendtnd to :deteoit !a"Ity.a"~;al

q-1144)g

Material Xnffpect ion

Ae described previouuly. zmterials for biellows mn4 dlxaars mist

be very uniform and free frca defect~s. W~hen material Is received, it is
usual3ly measre~d and inspected careful..y~to mae mire tbat ohre no laxrge
ire~mlirities such as tbinnfrE, bules, -wrinkles,, and &,ate,, azd no Suxrfzc~e
dama~ge s
__- ik, cace. ndpt.Otn cmsfa a- sheet
are used to determelueI the yield strength, ultimate strength, h4Lrdnqs, and.i
chemical ccoposition of the uterWat.

Weld IngEction

Ywmanyrinufatrers consider the formiQ& process to be an adequate

check on the quality ot the seam wo-ld. tenis is paticularly true of bellci
formed by bydraulic or rubber pr~essfure. The appearane of the Weld is usually
checked for size sad =fforaft7. In s~cates, the vold 1.6 thecked by radio-
grephy. As is~ explainedl In the following secticun, there in considerable dis-

agreement concerning the value of radicgraphic inspect ion.

Convolution Fcrtian

Althotgh it is know~n that seadowrph equipaent Is uged by acm con-

panlos for checking convolution shape, VA means of checking convolution
shape and meterial thickneon are no'o, known in detail because this func'tion Is
a part oft the proprie-cary aspeart of befloe~w and dispbragn manufacture. Now
ever, it ia apparent tU0 visual cbi~aka and meeaent ofciia iesions
are used in-all oa~es.

1 ____ _____59
Handling Damage

Extreme care must be taken to assure'that the thin beflogs and..

diaphragm material is not damaged during manufacture. Inspection for such
damage is done visually.

Cleanliness

Extreme cleanliness is required for aerospace b.0lows andd iaphragms..

particularly for multi-ply bellow~s. The procedure for checking cleanliness.
depends on the type of cleaning process being used and the cleanliness level
required. With multi-ply bellows, a reasonably good check is provideA by-Ia
postforztion heat treatment.* If moisture is trapped between the plies, gas-
expansion causes a bulge in the bellows. However, cleanliness is so important
and so difficult to check. that rigid control of the cleaning processes is t
best means of achieving satisfactory cleanliness.

600
 EVALUATION TEO===0

Bellows are both inspected and evaluated in"performance after fabri-

cation. Some or all of the following procedures may be carried out by the
manufacturer as well as by the purchaser.

Evaluation Procedures for Bellows

The primary fa'ilure modes for aerospace bellows are: (1) leakage,
(2) improper spring-rate characteristics, (3) squirm, ( 4 ) fatigue due to
excessive magnitude and/or number of applied loads or displacements, and (5)
fatigue due to excessive cycling because of vibratipn. Because failures mey
be caused by improper design, by material and manufacturing variations, and]
or by unanticipated operational modes, the evaluation procedures are concerned
with one or more of these sources of failure. The types of procedure used
for any given bellows are determined by the manufacturer and the user.

Physical Examination

Although aerospace bellows are usually nufactured carefully in

small quantities, it is still customary to submit each specimen to a physical
examination after it has been fabricated. 'Such an examination is desirable
because premature failure can be caused by very small material and manufac-
turing variations, and such variations may occur even in the most closely
supervised manufacturing process. Bellows se ccuonly inspected by the fol-
lowing .procedures.

Visual Inspection. If an unanticipated stress raiser occurs in a

highly stressed area, the bellows my fail prematurely. The stres' in
bellws are greatly affected by the thickness and shape of the convolutions.
A visual inspection of each bellows is usually made to determine that the
bellows has been made to the proper dimensions, and to detect changes in
bellows shape such as bulges, dents, flats, ridges, and grooves. In addi-
tion, measurements are often made to verify that excessive thinning of the
material has not occurred at the roots and crests of the convolutions.

Visual inspection is also used to obtain an indication of stress

raisers occurring in the surface of the mterial or in the welds. Because
fatigue failures can be greatly accelerated by surface irregularities. the

61
bellows surface is closely inspected for pits, nicks) and scratches. EL m-
portant are surface effects that a bellows may be rejected if the appearance
of the bellows surface is significantly different from the appearance which
is known to be normal for that particular manufacturing method. Welds are
inspected for size and uniformity of bead, and for color.* Although it is
not always possible to detect a poor weld by its appearance, experience has
shown that a weld which is irregular or which shows evidence of improper heat
application may cause premature failure. Mhe inspection of both the material
surface and the welds is often con.uted with a liquid dye penetrant to assist
in the detection of cracks and scratches.

Visual inspection is also used to verify that the bellows had been
cleaned properly. Usually aerospace bellows must be extremely clean and ar
evidence of dirt or liquid film may be cause for rejection.

Radiographic Inspection. There is considerable disagreement con-

cerning the value of radiography for inspecting bellows. Some companies
believe that radiographic inspection of bellows is so difficult as to be in-
effectual, while other ,-tVaieu believe that it is a usefu.l, if iperfect,
inspection procedure. Ihe value of the procedure depends to a great extent
on the size and shape of the be-llowe being examined.

The large surface of thn material and the number of welds make
bellors susceptible to leakage. Leakage is usually checked during the prolo-
pressure test because this is the condition most likely to reveal a leak.
Three types of leak tests are comon; (1) a bubble test in a liquid, (2) a
soap-bubble test, and (3) a helixu-mass-spectrometer test.

In the bubble test in a liquid, a gas is used to pressurize the

bellows and the bellows is placed in a liquid to determine whether bubbles of
escaping gas can be detected. In the soap-bubble test, a soap solution is
painted on the surface of a pressurized bellows. The bubble tests are parti-
cularly good for liquid systems because it has been shown that a sotirce of
leakage which cannot be detected by these methods will rarely leak a liquid.

62
 IFor gaseous arysteas,. and for liquid sysae aperatIn uni strisomt
-e oni.ns, :the h~lium mass spectrcmeter is nau7l use. -Thli 'tet xW be
conducted either by evacuating the bellows and swroubdintetbe11l a-with
heliusm, or by presw~iJz4ng the bellows with blm ad *snftti3W 1 taew mkoif

I of the be11ovs- In either case, the rate of leakage Is ua1y,,zdt-=9swo&--

any leak being considered cause for rejeect.

The spriNg rate of a bellows is the summation of the spring rates -of
each convolution. In addition, the spring rate is apoxaste3. proportiowa
to the cube of the thicknese of th.'e bellows mterial. Canseqitly,, excessive
spring-rate variation is one of the wat sensitive imans of iricating iiam
properly formed convolutions or ecesove variation In mttria3. thiknes5,. Mid
t spring rate is a ajor criterion for ccngpring the expected performnce of a
.* beflova with the laboratory perforinne of test specimens of the mme

I . For most bellows operating In the elastic state, the axial spring
rate of an ungpreiosiwized bellows is linear.. If the requied. loads are =all, the
spring rate may be meaured with dead wilghts or a load cell. If the lo.a8s wre
large, a univrsal testing machine is camanly used. It in desirable bafoe
taking mrnmets to "exercise 3 the bellows several time to tftaln a re-
producible deflection-versus-load relationship.
Since the axial spring rate is sigeificantly changed If a bellows
is pressurized, the spring rate under operating pressures and deflections is
also often deterained by experiment. 2his is a more colex test than an
unpressurized spring-rate determinaticu because the bellows st be sealed

properly and. restrained, and accurate preisinizIng apparatus is required.

I offset or angular d~flectioin. It is usially neces=7' to design spacial

fixtures for each application. In general, the forces reuIred for this type
of test are sufficiently low that they ane applied with dead veights or a
load cell.

63
Operational Evaluation

Operation~al-evaluation procedures consist Of cmqating the bellcvu

under conditions which simulate the most severe operating conditions.. fpi-
cal tests s:.'e concerned with deflection,, pressure,. vibration, -and. flowr., _Te.
tests are oftfta conducted at ambient temnperature first and then at operating
temperature.#

Endurance Cycling - Deflectionm 0& Mny bellow operate with fluid

pressures which are sufficiently low that the stresses due te pressuz,* are
insignificant. For such bellaws, it is possible to evaluate their operating
life by unpressurized deflection tests. It is necessary to duplicate the ex-
pected deflection modes and to deflect the bellovs a sufficient number of
cycles to demcmtrate adequate life. It is very difficult to determin'i how
ay cycles beyond the expected operating cycles are necessay to demiotrate
an adequate design. This is decided for each application by the manuflacturer
and/or the user.

Endurance-Cycling - Pressure Only. In sm bellows, for ej~aple

flexible metal hose, the stresses caused by pressure are the only significant
stresses. If pressure pulsations are expected in the fluid system, tests are
conducted to determine whether such pulsations will cause fatigue failure
ia the bellows.

Endurance Sircling - Deflection With Pressure. The most cmn

operational tests consist of deflecting the bellows in the expected opera-
tional modes while the bellows is pressurized to the maxinua Internal or ex-
ternal system pressure. This test produces the highest cambled stresses.
Cycles equivalent to the expected number of life cycles plus a certain umber
of adit ional .yles rcre Imposed to demonstrate adequate life expectancy..
When only aasl deflections are required, these tests are relatively simple.

Vibration. Because a bellows behaves like a spring and responds to

certain frequencies, it is necessary to demonstrateftt a bellows will not
 reepond inproperly to the expected vibrations of the system structure 4Wafi
prematurely from fatigue or overstressing. The belows is,,ususl ~ubotd
to the amplitude and frequency inpuis which represent .the cniin~~~~
* ~in the system. A range of frtequencies wan amlitudes -is. usvally sdcu
of the difficulties of'predicting the-expected operatingvausacrey.,
Occasionall~y the behavior of the stru~ture is put oni tape ad4 tet=p
used to drive the vibration equipment.;

Flow. The effect of fluid flow through the bellows-mist, smftlmm

be evaluated. One criterion is the mount of pressure drop ceased by the
bellows. Another criterion is the susceptibility of the bellovs to flow-
induced vibration.

Squir_

A bellows with a length-to-diameter ratio greater than one May fail

as a colun when subjected to internal pressure. The determination of this
possibility is usually made during the proof-pressmne test.

Proof Pressure

As with all fluid-system, compunents, be!llows are subjected to IL4/2

to 2 tizjes the maximm system ;.ressiue to demonstrate a pressure-salfety factor.
T7= proof pressure is usually utintained for a period of several minutes.

Pressuire Capacity

It is cUstmry to subject a representative bellows to Wuficient

pressure to cause rupture. O~casionally, note Is taken of the pressure at
vhich severe deformation of the bellows occurs, because only large deforms-
tions change the subsequent response of the bellovs (e.g., spring rate,
stroke capacity., fatigue life). Thus the deformaftmo pressure rW be the
practical mimu pressure capacity.

65
 If a beflove fails to pass ona* of the above teiWS-, *(wtieulk-4.Ftfie
operational tests, it is neceisary to determine vhetbtr failure waSs diusid-by
circumstaes peculiar to that b*3llows, or vhthebr ai design or m-fituing.
chane should be made. The postexamination donasis bf itilizing tbht pioceaures
described under Pbysicil, Izmination, to determine the cause of ;filurjj -If
addition,. it is comon tractice to atake a microscopic examination of'-the failed
areat, and to make a metallogr&phic examination of polished cross sections of
the failed material.

Ivaluation Procedures for Daham

Diaphragas are iisually evaluated only by the amaufacturing facility.

The component containing the diaphragm may be evaluated by the purchasing
facility, but the purchaser is usually not concerned with the diaphra per
se.
The evaluation of diaphragms as instrument devices is described in
considerable detail in eWdell's "Diaphragm Caracteristics, Design, and
Terminologa".(i T1e techniques and equipmnt described should, be adequate
for measuring force, pressure, and deflection characteristics for a-'
diapbragms vhich are to be operated entirely within the elastic atate. Many
of the procedures described could be used for evaluatin smll instruz-ant
bellova.

f
The evaluation of diaphragms for other operational modes is not
veil reported. In the absence of sufficient information, it wast,be concluded
that expected operational conditionas are simulated, and test specimens are
cycledi in an attempt to determine operational adequacy. It is asmined that
asz , of the techniques used for beflows are also used. for diaphragms.

i~66
11. * .
Newll, F. B., "DiaphagCa
AM, Nev York, 7i Pp, 1958..
raceie s Design, and- ex=
-:
W

2.Widhpk W. A., t ml.,p "Investigations of the Propertiies. of- Corg~ed

3. "Pressure Traneducing and Instrumentation Techniqiiesi, L -A 1p.Ga~n

Controls Corp., Wharte, Calif., August 1, 1960 (AD 25i&1b

1&. Grover, H. J.,p and Bell, J. C., 'Some Evaluationsiof 'Stresses in Aneroid
Capsules",. Proc. of Soc. Exptl. Stress Anle, lp pp 125-134, 1948.

5. Laup, A., and Weil, N. A., ."Analysis of U-Shaped Expansion Joints,.. J.

Apple Meech., pp 115-123, March 1962.
6. Anderson, We Pe, "Antlyttis of Stresses in Bellaws"p. Part 1: Design Criteria
ila 'Test.Results, Atcuics International Rept. _XRmi,toe1, 1964.

7. *And~racxi, We Pe, "Analysis of Stresses in bellows%, Part II,xwbhtical)

Atomics lniternational Report XAA-M- 1 152T, )6Ay 27, 1965.

8. Winbornep R. A.,. "Siplified Formzulas and Curves for Bellows Anaysis".

Atomics Internatioml-Rept. 1AA-SR-9849p Auust 1, 19611.
9. Winborne, Re A., "Stress and Elevated TeuPerature Fatigue Characteristics
of 'Large Bellow", Atowl"s Mzzermtional, Septdei 15, 1964 (N11.i935).e
1.0. Turner, C. Be.. and lord, Be., "Stress and Deflection Studies of Pipeline
BxpasiLon Bellius" Proc. IEnst. of Mich. Ing., M (15), pp.526-552,y 195T.
1.1. Penny, Re.14p, "Axismtric Bendi' of the -Geiral Shell of Brol1tition
Ain.rrg Creep", J. Mach. Ing. Soti.,. 6 (1), pp 44-4Z6, 1964.
12. Dadianiy B.,, ari Radkowski P. Pe, "Nmrical Anlysis of Vnsimetriciil
Dending of Shells of Revolution", AIM Jc ,2. (8) August 1963.

13. Greenbamu, G. A., wCinmnts. on 'Nirical Analysis of Ilnqyinria BI'ending

of Obells of RsVOlutio" AIMA -ournal, 2W(3, PP 59059,Mrh16.

1~4. Sepeitoski,-Ve. K., at fi.i ."A Digital Cqutw Program for the GeUeral
Axially *mtric Thin-Shell Problem, AMK Trams., iii pp 655-661p 1962.
15. Krausp Rep -A Revi and Evaluation of Computer Progaxs for the Analysis of
Stress In Pressure Vessels", Pratt & Whitney Aircraft, OLto 1963-Se~t. 1964.

* . 16. Bell Aerosytems Co.., "A Study of Zero aviY Posi Exulsion
Techniques',ept* No. 8230-933004., June 1963 (N63-199&4)

-61
 awmm~z0 (C~zm
.1T. Stricklin, J. A.,. et al.., "Large Elastic, Plasticp Wa Creep bfInitions
of Beas and Axiuyntric Shells", AIM Jounal,. , pp 1613.-1610,
September 19611.
18. Wigotaky, V. We., -Volum Compensating Welded Bellows", Design News, j
(10), pp 6-9, May 9, 1960.
19. Glletly, G. D., and Radok,, J.R.M., "On the Accuracy of Somw il
Solutions", AMI Trans., 81. pp 577-583, December 1959.
20. KAlnins, A, "Analysis of Shells of Revolution Stjected to Vjxstrical
and N= ymtrical Loads", ASU ras. pp I46T-1176. Septmer 19a1.
* 21. Goldberg, T. Be., Bogdanoff, J. L., "Static and Dy'namic Analysis of iNopunifors
Conical Shells Under Hgmtrical. and Unsymtrical Conditions%", 441
Missile and Aer2rae TecbnO~oa7. pp 19-238,, August 1961, Acadami& Presus.

22. Mfrsbal, J-. A., and Dight, D; G., "SCR-II, A Progrmto Perform Stress
Analysis of Sb1l of Revolution", General Electric Copany Rept. KAP-aN-
EC-19,- UC-32., Jume 1, 1962.

23. Cohen, G. A., -Computer Analysis of Ayintrical- Defointci of OrthotrcPic

Shells of Revolution", ALM JO, 2 5) ay 196k, P 932-93k1.
211. Thlnins, A., and Lestingi, 3T. F., "Oni lminear Analysis of Elastic Sbells
Of Revolution", submnitted to 3. of App].. Mach., 1965.

25. Clark, Re A., -Auypttic Solutions of Elastic ~1501 Piblm',Amol

:6
Sltosof Differentia E ati and Their Auicios John8
& Sns, Nw York (1964)., pp 105-209-

Foj .,Nurical. ouino riayedPri1Dfeeta t li S

tAdsnWse Machie sign
1962.161

29. Daniels, C. M., "Pressure Losses in F1exiblee Mtal, Tabhiag, Prod. kg.,
April 1956, PP 223s 225, 2T..
* 30. Hawthorne, R. C.,. "Flow in Crru@Lted NOse" Prod. Nag.,9 pp 98-100.,
June 10, 1963.
31. Danielsp C. M., "Designing for Duact flexIbility With Usilows Joints",
machine Design, s, (21), oatber 1959, PP 14&6155.

32. Naringi, J. A., "Instability of Bellows &ibected to Internl PressrO,

Philips Research Repts., j(3), PP 189-196, June 1952.

68
I 33. el-d D. 3.., "packling of Doube Bello- Expasion Joints Unider Internal
-i Pressure", J. W~eb. Ing. Sci., 6 (3), 19A-1
314. Seide, P., -Thek Effect of Presure on the Bez.Ing Caracteristics of an
Actuator System1 , AMU Trans., 82p pp L&29-437, .'960.

351. Vail, D. B., "Daternination of the Natural Frequen:ie: of Vibratioa in the

Breeze Bellows". Knlls Atomnic Paver Lob. Rept. KAPL-M-M!V-&, March 4., 934.

136. Bell Aerosystemis Co.,. "Studiy of Zero-Gravity Positive ibxp ilsion Techniques",
I Interim Report No. 8230-933007., April 1961.

37. Rekate~, ff. L., and Schvartz, J. I., "Behavior Stadyt of a Curtias-Wright
Snorkel Flexible Exhaust Connection for Suibmarine Aplication3 ,' U.S. X&TY
Marine Eng. Lob., Nowe er 13, 1963 (AD 43023).
38. Langer, B. F., -Design of Pressure Vessels for Lov-Cycle Fatigue", J& of
Basic Eng., pp 389-Io2, September 1962.
39. Dzessler,. R. F., "Bending and Stretabing of Cor-ruated Diaphrages'
(J. Basic In3g.) ASW Trans.,. 8.1 pp 651-659, Deatmer 1959.

40. Cihrron, P, nDefomtim des Hmbraaes Anieroides", Cmtea R~end=s

Hebdcawdafrs des SLar-ees de L'Aaad~zle des Sciences, 209 pP93985,
1939 (in Frenchb).
41. Pfeiffer, A, "heory of Cor-ngated Diapbraws for Pressur Wastning
=nstruments", Rev. Sci. Inst., , pp 6Wo-6&,4, 5-edteer 1c.Ji7.
- 42. Haringx, To A.., mThe Rigidity of Corrugated Diapbragas", AppI. Sca.. Res.,
Sect. (a), , pp 299-324,. 1950.
A 4i3. HAriigx, J. A., -Nonlinearity of Corrugated Dispras", App].. Sci. BasIL,
I Seat. (a), 61 pp 45-52, 1956.

* 44. Bhrfrn, J. A., -Design of Carrugated M~aphragas" Aga Trans., 12 pp 55-

61, 61-62, 62-&%Janumzr' 1S5T.
k15. FMint, C. 3,. 'Theory of Rlestrained Cocrugated Dipram" Y gineer
(Load=n) RE (5246) PP 193-195, kguut 1956.
li6. Akasa, ?., wad fkagisbiv To. -Vibrtica of Ccrrugxted 11ia~pbhrag
Bull. Japan Soc. Nech. tg., .3, PP 215-221., August .1958.

69
 ___

- ~

*1
I
I

&

Mmo1~A~D E~Q~AP~Y

II
-1
 Adachi, J., and Benicek, M., "Buckling of Torispherictl Shells Thider
Intrna Prssu E xperimental *ishanics, ugust 1l6ad, p'2LT-222.
_,()
Duckling experiments were performed on a series of acdel torisprical
bulkheads loaded by internal fluid pressure. The models were rigid poly-
vinyl chloride with a base diameter of 10.52 in. Parameters which were
examined included thickness, central angle, and toroidal radius. The
critical buckling pressure varied as a power of the thickness and almost
linearly with the central angle. The buckling 1pressure was found also
to be very sensitive to a change In'toroidal radiusa.

2.
Aerospace Fluid Comment Desinrs I Handbook, Vol. 1, Section ' .Op
Deach, CasifeaMy 964~

Section 6.6 of the handbook presents a review of design Information on

metallic bellows and diapbraga. Several design formulas and charts are
given to aid the designer.

Akasaka, T., and Takagishi, T., "Vibration of Corrugated Diaphraga",

Dull. Japan Boc. Wech. Eng-., 1 (3), August 1958, PP 215-221

T. Akasaka formrly studied the elastostatic properties of the corrugated

diaphragm, considering it as an orthotropic elastic plate by smearing out
the effect of the corrugations over a wave length, and foind that the
calculated results agreed well with those of experiments. In this paper,.
.. an analysis is made of the vibration properties of the corrugated
A -. ,.diaphraga with the sam assi~tions as for the static case. Approxi-
mate formulas are obtained for the natural frequencies of syinetric and
aqnamtric nodes Including the effect of concentrated. mass Further-

I
more, by the enerW method, the effects are studied of initial tension
and concentrated mass an the natural, frequencies. 'In order to check
this analysis, some experiments were performed. The experimental
results shoved good coincidence withi the corresponding formulas, when
the anisotropy parameter a is not- so large.
 -~~~~1 - - -- VT.

Aksel'rad, E. L., "Equations of Deformation forIbhells of-Revolution

and for the Bending of Thin-Walled Tubes Sujected to Lorge eaic
Displacement", ARS Joura., 2, July 1962, pp fLhi-ll5i.

In this paper the author extends A. I. Lurie's equationg.tOr a shell 6f

revolution to large elastic displacements or an anliotropied,~ ne~
shell. These equations define the bending of a curved tube provided the
tube X-emins a section of a shell of revolution.

-Asel'rad, E. L., "Calculation of Corrugated Members as Nonahallowav ls m

Izv. Akad. Nauk, S, Qtd,. Tekh, Nauk,. Nekh i Mash (5),. 1963, Pp 67-76
(in Russian).
Equations were previously developed into a form suitable for the calcula-
tion of large buckling of membranes of an arbitrary profile. A solution
of this type is analyzed by the ThMnov-0alerkiz method to a linear appoxi-
mation. A solution is given for membranes for the general case of an
incomlete sinusoidal profile without a larger edge corrugation; it
corresponds satisfactorily with published experiviental results and it is
suitable for practical calculations of membranes with a liea deflection
curve. This solution takea into account the effect of the type of fastan-
ing of edge corrugation and the effect of corrugation pitch. The method
used can be applied to the calculation of nonlinear characteristics of
menbranes, including membranes with edge corrugation.

6.
-Akselarad, 1. L., "Large Dforntions of Corrugated Membranes as Sonecallov
Shels",Izv AkSed huk, 5331, wakh i Nash (1), 1964, P 1&53 (in Russian).

In a previous wik, the analysis of in3.l deformation of a membrane vIth

sinusoidal corrugation was based on the solution of equations of a shell.
In an expansion of this solution, the ]problem of cilculating nonlinear
ebaracteristics is analyzed In this paper. The produced formtulas for
calculation of deformation of membranes with uniform sinusoidal corra-
gation are compared with experimental results produced by V. Ta 11l'm1%skid

T:.
Aleach, C. W., et al., "Prediction of Creep Effects in Aircraft Structures",
General kbramics/Convair Report,. Yebruary, 1962 (AD 27163).
The method presented in YAM TR6O-I4fl Part I for predicting 19formations
In aircraft structures undergoing creep is reviewed. Modifications and
sImpUifications of this method are considered, and a sivolified version
of the original method is presented. Box been tests were used to investi-
gate the simplified method. The outcome of these tests indicated the
accuracy of the method in predicting deformations In aircraft structures
undergoing creep was in the same order of accuracy asthat usually

2
 experienced in redundant analysis procedures. The power law for exprssizg
creep relations is found redundant in solution. Its usefulness appe"s
chiefly ia the display of creep test data. Currently ceppeitc
methods, such as the Larson-Miller and Manson-Hferid methOds, 4MWa £
*linearizing approximations for creep-relations withb the result tia iinc-
curacies in predicting very long and very short creep life restict thier
usefulness. An approach to creep prediction based on creep-ruptr iuory
* for establishment of creep laws is proposed. Relatins'hips between tens oh-
and creep-test data are examined experimentally with respect to'single
and repetitive load applications. General relationship: are exined
with the outcm that a general approach to creep prediction in all1
metallic materials appears improbable.

8.
Anderson, W. F., "Analysis of Stresses in Bellows, Part I: Design
Criteria and Test Remalt 'l Atomics International. Report NM-SR-452T,

Design charts and systematic design forms are presented for simplified
calculations to cieck the number of convolutions and thickness required
to limit the deflection-pressure stress range in three types of bellw .s:
(1) convoluted bellows, (2) convoluted bel-lows with reinforcing rings,
and (3) toroidal bellows. The design charts are based on the equations
I *~ for stresses derived from an asymptotic solution for the equations of
toroidal shells found in previouslyr published literature. Proposed stress
limitations to be used wdith the calculated stresses are based on those
* of the ASA Code for Pressure 7iping. Data from 108 fatigue tests at
T0 P and 18 fatigue tests at 1200 F for all three types of bellows are
presented. Thse data are evaluated statistically and justify use of
the proposed allowable stresses. The conclusion is drawn that expansion-
Joint bellows can be designed to the sme stress levels as other comp'cato
of a piping system with equal confidence in the reliability of the deiigz2.

Anderson,, W. F., "Analysis of Stresses In Bellows, Part IX,Mathematical",.

Atomics international Report mA-sR-i&52T, may 27, 1965.

This second part of a tWo-part report presents the mathematical juetifica-

tion of a proposed dectign analysis method for pipeline expansion bellows.

varitles td simpiyiad thee

qatosaotheatia sltn cud be nbey
(2) Frnae esinite difference,
bsoutd seeme torqirec anadeslytical
nthe oiffeent eauios the endingstesseshanr (tta)Therunpead

jI 3

4L_
_ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _
 values and start to decreue within 1/6th (15 dog of the 90 di4)-.of the6-
segent of the toroidal shell section. Also, with the cc*'ter i -
able at that time, costs of programing the finite diffrece eqatini
and of computer time promised to exceed those for the techniqaw slectied
(A greater number of important variables, hovere, itht b've been
reliably investigated by this finite difference techfiftie.) (3) The,
possibility of oaining direct solutions of the tcroidal ihell equations
as a boundary value problem, using published values of Henkel functiino,
had been demonstrated by the work of Clark and Hetelyi and Timts. This
technique requires solution of only four smultaneous equations for each
combination of variables, and allows inclusion of one Imortant variable
that must be omitted by the truncated series solution. Since the Henkel
functions are solutions to equations which represent the shell equations
asymptotically, a certain error was unavoidable in solving the asymptotic
equations at points far from the as.yptote. Nevertheless, the direct
technique was selected for use when it was observed that compensation for
this error could be introduced, to permit reasonably accurate results.

10.
Andreyeva, L. E., "Calculation of a Corrugated Membrane as an Anisotropic
Plate't, Inzh. S., Akad. Nauk St, , 1955, pp 128-1i1 (in Russian).

Existing methods of calculating corrugated membranes have limited ranges

of application, since they are applicable only for ccuparitively mall
deflections (of the order of 2 - 2.5 times the thickness of the membrane).
However, most mmbranes are deflected an order of magnitude greater than
the thickness of the membrane. The thinner the mebraae, the greater
its relative deflection comared to the thickness and the sller the
region in which the calculated x-nults are valid. In this pPer, an
approximate method is presented for the calculation of a thin zembrane
for deflections considerably exceeding its thickness. This method mwy be
used for a membrane with arbitrary periodic corrugations. The method is
essentially the same as that used by Haringx.

I 11.
"Application of Metal Bellows", Machinery(London), 73 (1888), December 30,
194f8. pp W8-894f.
Brief descriptions are given of the design, manufacture, and use of formed
and welded metal bellows. Causes of failure are given, and a chart is
presented for estimating the possible cyclic life.

I.
L

. . . .. - -4- . .. . 1-.--
 12Archer, Go R.., "SOcessful Welding of Foil Gages", WeldinM;Jo,.$2'(
April 1960,. Pp 34I3-34T7.

-' Time-temperature char~ts are given for the heat lag of electritaI-fiuiftaince
spot welds in stainless steel 0.010 In. or less thick: The maximm
temperature is usually rreached wthin the first .1/2 cycle of 6D cps current
I . flow; a weld forms in the first half cycle and the weld nugget Increases
'I in size for each subsequent Walf cycle. Variations In welding time (number
of cycles) do not produce pronounced effects on weld st'ei~gth budt v~fari-
tions in heat (setting of the welder) are extremely impdrtiit." Expiilsion
of molten metal, if it occurs, usually occurs on the first or second half
cycle. Peak voltages and currents, rather than rue values, aire significant.

ZArcher, R. R., "On the Nimrical Solution of the Nonlinear Aquations for
Sbells of Revolution", J. Math. hys., i4l, Septeber 19W, pp 165-178.
Although the general equations for the finite axiyetrical deformations
of thin shells of revolution have been available for som time,. it is only
very recently that computational procedures have been designed which have
produced detailed numerical results with the aid of digital computers
for significant nonlinear shell problems. It Is the purpose of this paper
to st up a procedure for the numerical integration of the nonlinear
shell equations as derived by fleissner, and demonstrate the effectiveness
of this procedure In handling nonlinear shell problems. In particular,
detailed nimerical results are obtained for the clasped spberical shel
segiment under uniform pressure and the spherical shell seamsit witholut
edge noment or horizontal restraint subject to a point load at the apex.

14.
Arkilic, G. M., -Analysis of Toroidal Sblls of Sei-elliptical Cross
Section", Ph. D. Thesis, Northwestern University, 19514, 88 pp.
The problem of toroidal shells of semi-ellipticc-l cross section under
axially syeotrical loadin and given boundary conditions is treated by
deriving the differential equation from the fu~t1equations of thin
shell theory. The colete solution of this equation Is obtained by
utilizing the asmpotic forms. After baying wArayzed the general pro-
cedures of several types of loadings and boundary conditions, some prob-
lems o!' practical Importance are worked oat numrically. OE these, the
atmysis of stresses and deformations of bellows my be mentiored. In
addition, sppraximte formulas for maxima stresses and deformations ale
established for general engineering use.

C*Au, T., "Equations for Thin Toroidal Shells-, J. of the Aerospace Sai.,
26 (6) JTune 1959, PP 39l-392.
In the analysis of thin elastic shells of revolution having Constanit
raaUus of curvature along the meridian, the fundmimtal equations of equilib-
run and deformation based on Love-Heissner' s assumptios can be reduced

5
 0, two simultaneous differential equations of second order-6. ntbe case&_-
of conical ~d spherical shells, such reduction Uies'tb rattesiag
results. F:i the portico of a toroidal shell whieb semwess ts-,ibe
of a toriconical or 'toorispherical, head of a pressure vessel, sia'
results can be. cbtalmed.

16.
Azzi, V. D.,I"Plasticity - A Burvey of the State oft ' be Art", General
Dynamics/lectric Boat Report No. 1413-62-006, raiwv~t 15, 1962.
This report reviews the available literature pertinent to am aspects of
plasticity. Pist, achievements,, present areas of investigation, and possible
future areas of interest are considered. The emphasis is,on the present
state of the art of those phases of plasticity broadly 4asiie s
limit analysis, dynamic plasticity, and yield criteria. Research in plas-
ticity has been most active for little mawe than a decade. It Is concluded
that altbougb considerable effort bas been expended In the Investiation
of those phases of plasticity of interest, and extensivc literature is
available as a result of these Investigations, mobh additional research
will be required to bring this relativel~y young field to a desired stage
of maturity. A bibliogapt of 2614 references Is Included.

Badavy, 3.8.11., "An Investigation of Deflection and Stress Distribution

in Corrugated Diaphragms", Ph. D. Thesis~ University of Minnesota, June
1959, 132 PP.
A general analysis of corrugated diaphragm is presented. Expressions
for deflection and Stresses have been derived by differential equations.
Additional developments have been made for the deflection of corrrgated
diaphragms, subjected to central loads, by dimensional analyses. A com-
plete stress analysis wes made for a diqbragm subjected to a central
load. Using the results of this calculation, further recommndations are

made for other diaphrbagmsa.r, ul, 96

18. is n , .
LiL'ear Differential Transformer". sl eo.u.,BcoeTy,16

A pressure transducer Is described which uses a liervariable differen-

tial transformer to sense the displacement at the centre of a diaphragm
subjected to the pressure being measured. Design details of the transdxcer-
are given Including complete specifications for the differentia trans-
former and the electronic circuit used for Indication. The diaphragm
of the transducer is changed to get different full-scale presture ranges.
The accuracy of the transducer in all ranges is better than *1 percent.

6
 barker, R. B., -Pipsa Welding - A Method for _Xoding .Weld Po9rpiy" Wding
Z., !A, JTanuary 1965, *p lss.
Porosity in welds.-is greatly redu~ed when Al is welded with, the pas-metel-
arc process in a- chamber filled with an inert gas to a pressure of 50 Psi-

20.
Barnby, J. T., "Effect of Strain Aging on the High-Temperature Tensile
Properties of an AIBI 316 Austenitic Stainless Steel%, frozu and Steel Inst.
J. M (), April 1965,. PP 392-39T.
Thepperane of jerks in stress/strain curves during high-te ertr
Tehe estnad fiiia n il it romt~ e turesl
tesitssos tngt stnd o of ilocions
dnta n occuspr in AI 3'0esl6
aestes.ti stainlesst steengl.ockneing soateof canocrking ith31
iuncreaatsins stemeaue. conient i jraytess/sranres, wish
interreitepasue. strj
odniden agn fdsoain
tes/tain chrvbecomi
strongly locked by carbide precipitation upon them.

21.
Baskevitch,. N., "Effective Area of Diaphragms", Machine Design,., August
7, 1958, P 12 1 f.
Formulas are presented for determining the effective area and effective
diameter of circular diaphragms. The formulas are derived by-treating
the diaphragm as a uniformly loaded trapezoidal bean.

22.
Beard, C. S., "Positioners for DiaphragK Actustors", Instruments and Auto-
mtion",. 29 (9), September 1956,. pp 1782-1.783.
Valves must be positioned accurately to functioa as Precise control elements.
Brief descriptions are given of valve positiocxers using bellwsa and diaphragms
to impart sensitivity to the valve and to insure accurate positioning as
dictated by the control signal.

~BeUl, W. J., and Benhem, P. P., "The Effect of Mean Stress on Fatigue Strength
of Plain and Notched Stainless Steel Sheet in the Range from 10 to 10T' Cyrcles",
Ba4I5163, Battelle Manorial Institute, October 1962.
Stainless steel sheet (18 CR - 9 Ni) vas tested in fatigue under axial-load
:1 cycling in plain and notched conditions. Various stress ratios R were
used, rangin from R a -1.0 to 40.91, and cyclic lifetimei of from 10
to 10T cycles were covered using testing frequencies of 5 to 15 CPR and
3,000 cis. The effect of sean stress on notch ftigue streng+.h could not
be predicted empirically tolely from unnatched material data; -it least
one notched fatigue curve is required. A fatigue strength reduction factor

T1
 1 based ou maximu stream for a particular mum stress =ud eanduratweVprovlt
the most reliable correliAon bertwien uhmot~hed an& Dched Wia i
functions existed in the low-cyqie region between stidisrabige and plstic
strain range, and between total energy and cyc~es-to-frac*tir*; both
functions were largaly Independent of stress rio.

241.
Beli Aerosystems Co., "Research on Zero-Gravity kpulsion Techniques,
Final Report", No. 7129-933003 Mireb 1962 (AD 2TWOi4)
Design concepts are presented of zero-gravity expulsion- devices on an
extremelyr broad basis and a compendium is formed of &%Acb device configura-
tions as an aid to system designers. in the selection of expulsion systems
for particular applications. As a supplement to the concept presentations,
the report also presents some of the awre liuortant data necessary for
design with emphasis on information not readily available In current litera-
ture.

Deli Merosystems Co., "A Study of Zero Gravity Positive Expalsioa Techniquaes",
Report No. 8230-93300i1, June 1963 (N63.199611).

This report summaizes the results of a 10-month propem condueted by

the Deli Aerosystems Ccmpan to fulfill three prim objectives: (1). To
evalatme positive expulsion devices alIcuble to rocket propul-

pesn emaion ra()tes

conudtspiald inotetaicn atralyss -oste inxplsion
teie expusiondevtcfes. th has est potentialthat apsubstantialobo annod
espacerfighdei and hia t intada of nirnein data werealedinestc,
were
lacr n being
titeprcsso ohfrgatee m eerae ex uin dev an r nno-
fh

conventional systems such as those using surface forces, dielectrophoresis;,

and acoustics. On the other bind, the ez..sting booy 6f engineeing data
covering metallic expulsjion devices was found to be rather meagqr. How-
ever, metallic expulsion devices offer a greater potential for manned
space fligh because of their ability to withstand a wider range of operating
tempeiatures and raitondsgs
LM hnnnmtli sses

* 26.
Bell Aeroirjstens Co., "Study of Zero-Gravity Positive Expulsion technIques",
Interin Report No. 8230-93300T.9 April 1961.
This report smmarizes the work of a 10-month continuation effort under the
basic NWT-149 progran. the major tasks considere for this follow-on
effort we:(1) structural Investigation of mtallic bellows; (2) ynamics
investigation of metallic bellows; (3) permeation technology; and (Ii)
optimm envelope investigation. The report discusss the modifications

8
 . .. . 2'

incorporated into the structural analysis of belloas which was stated dutrng
the basic program. Also presented are results from a prellminary stlu: of the
effects of variation in geometric parameters. The dynamics anelyses of bellas
include a comparison of the bellows. accordion mode and the gteral case of
a longitudinal homogeneous bar fixed at both ends, A mathmUtical mode for
the liquid modes was also developed. This porous-bar concept demonstrated
good agreement with test data. Results of an experimental test program to
determine the dynamic characteristics of double-ply bellows are included.
The permeation work was a continuation of the investigations initiated during
the original program on detectors. As a result of this work, the thermistor-
type detector was selected as the most practical. Both the design and the
application of this detector are discussed. A new permeation chamber, using
a corrugated diaphragm in place of a bello-ws, is discussed and the results of
development testing with this apparatus are presented. A series of graphs is
presented for use in selecting the lightest weight bell(rs assembly.

2T.
Bell Aerosystems Co., "Study, Design and Test of functionally Integated
Pneumatic Components for Rocket Propulsion Systems", Edwards Air Force Base
Tech. Rept. No. AFRPL-1I-65-93, June 1965.
An investigation was made of helium pressurization functions in propulsion
systems for several classes of vehicles. Techniques of performing the funz-
tions of fill and drain, filtration, start, pressure regulation and relief
were studied and concepts of integrating and combining these functions were
evolved. The concept of integratiom of functions vae implemented through the
design and fabrication of two integrated pressurization components for service
with fluroine and methane gas pressurants. Design characteristics included
nickel-berylliu seats and poppets, monel cladded solenoid with solid-state
current limiter, metal dynamic and static seals, common relief and regulator
spring and bellows, and Monel 5 micron filter. HF corrosion tests were con-
ducted on nickel-beryllium and Inconel specimens. Leakage, response, and
endurance tests were conducted on the solenoid start valve which incorporated
a nested plate bellws made of Berylco Nickel 440. Performance tests were
made on the relief valve which contained a bydroforned bellows made of 3-ply
0.003-in. Monel. Steady-flow and dynamic-flow performance tests were con-
ducted on the pressure regulator which incorporated an externally pressurized
bydroformed "Omega" bellows made of Inconel X. Finally, performance tests
were condu-ted on the prototype integrated pressurization components to verify
their design and ascertain achievement of performance objectives.

-"Bellows Give Flexibility With Strength", Engineering, 1, March 23, 1952,

pp 36-387.

A brief description is given of the use of metal bellows for providing the
fle ibility and strength needed for expansion and sealing of piping systems
and processing equipment operating at elevated temperatures and pressures.

9
 "Bellows: Improved Structure Reduces Flex Failures",, Iran Age, 172,
December 24. 1953, PP 86-87.
It is claimed the, a welded-diaphragm-type bellows developed for atomic
energy applications offers long life under severe c- litions of corrosion,
vibration, and high temperature.

"Bellows is Redesigned for Welding", Prod. Eng., 27, December 1956, p 161.

A change from bronze to stainless steel required thiner ections at all

joints for welding.

31.
Bert,, C. W.,. and Crites,, N. A., "Experimental Mechanics in the Develop&nent
of a New Miniature Pressure Transducer",. Experimental Nechanics Perwigon
Press, Ltd., 1963,1 pp 307-322.

This paper describes the development of a miniature pressure transducer of

the diaphragn type, with strain elements based on a new concept. The dia-
Vhragn was only 0.0025 in. thick and 0.190 In. in diameter and it was
necessary to develop a special microanalysis technique, using ultrathin
brittle-lacquer costing. Special procedures were required for dilution,,
spraying, and curing. Consideration of the train distribution on the
diaphragm surface showed Ot~w the Otium arra-Sngemen-t bad radial strain
eleiments in the outer region and tangential ones In the central portion.
A unitized etched-foil design was developed in which all four eras of the
bridge are active ars on the diaphragm surface.* The principal advantages
of tiifis design are its internal temperature copensation and its high voltage
output per unit pressure change. Performnce data for the transduc-er were
gathered.

Bervald,. W. B., Bass, H. A., and Reistle, C. E. Jr.,. "koreau, f Mines

iMutiple-Diaphragit Recording Subsurface-Pressure Gae".. Bueku of Mines
Rept. of Imaeti~tions 3291, Nqveber 1935, 19 pp.

Accurate measurement ofsbufc rsue n evrtrsi rdcn

nd ell isan important step +,o place the practice of petroleum
oil as

ioc-ttin
nai xat cenifcbai. ii eprtpesns r10y h

- rtof___the__Bureau____of__Mines___has___played___in___te__dev
-- -utal
 Bhuba,- P. C., "Analysis of Toroidal Shells of i-d~el4itie4 Cro*o Seztibn
Sujected to a Uniform Internal Presse' N . htiPsle"e
pi
State University, 1957.X.9"h
shelsofsemi-4fliptical cross section. The shill is suicted to.&a uni-
I .form internal pressure aknd the bomndies of the ahill. wre assumed to be*
fixed. The analysis of toroidal shells of semi-elliptical cress sectilon
wa~s ade by G. 1k. Arkilic, for certain types of loadin a and boundary
conditions; hogVyer, the solution *for #Lshell with fixed bounftries,
subjected to a uniform internsa. pressure alone, has, not previously been

developed by Arkilic by using the method of asymptatIC integratioet. The

solution is valid for large values of the diniiuionO parameter Ii. A numerical
problem is solved.

*/Black, H. L.,. and Lbebier, L. 1W., "Press Frmility of histenitic 'Stainless

Steel', Metal Progr., June 1965, pp 62-6T.

*To obtainL satisfactory resu~lts dur-ing press foiuizig of austenitie stain-

less steels.., the material ahould have the lowest possible yield stress and
a -zoerate rate of vdrk bordening. This article discusses how the ccopo-
sition of a stainless allo7 can be modified slightly within AISI specifica-
tions to insure cptimu performe.

ml~r, R. 'd., Johnso, D. L., and Morley, J. P. "Neal Bel~lows Seals",

Lubrication Ing., 17 (10) omtober 1961, pip 47o:4~75.
Sometims temperature extremes a: g =-radiation dictate the use of a
metal bellows In an .end-face-type seal. The purpose of -this paper is to
acquaint the seal designer and the seal user oith the varimus pwamters
associated -oith the two basic types of metal bellows. Brief descriptions
are given of the parameters.

36.
Borik, F., and Alersj,. A.,, "Mieasurmnt of the Ilastic, Properties of
Rolled Sbeet", Xetallmrwjicsl. Soc. of AIM, Transactions, au (W) January
j 1965, PP T-11.
This article describes a simple masurement technique to obtain the nine
elements of the elastic modulus tensor from which all the elastic proper-
I -. ties can be deduced. Variation of Young'sa modulus and the rigidity nodulus
in the principal planes c0 ausformod steel iheet is discussed.
 - -V

Dotosan , R. A., and Sorensen, A. E., "Electroforming Large Diaphras",

Paper for presentation at the Space Power and Propulsion SMposim,
Philadelphia, June 3-5, 1963.

Tie successful use of vacuum-formed plastic molds as mandrels for electro-

forming of large parts has not been previously reported. Three advantaes
are associated with the manufacturing process: (1) the wood molds required
for the manufacture of'the mandrels are inexpensive and may be quickly made
by any pattern shop, (2) the plastic mandrels are inexpensive and may be
produced at a high rate, and (3) three-dimensional objects having back
angles may be manufactured in one piece.

Boyd, W. K., and White, E. L., "Compatibility of Rocket Propellants With

Materials of Construction", Battelle Memorial Institute 1t4IC Memorandum 65,
September 15, 1960.

This report summarizes the available information on the compatibility of

liquid rocket propellants with prominent materiali of construction. Compati-
bility data for materials not ordinarily covered by the Defense Metals Informa-
tion Center are included. These data were found during the search for in-
formation on materials that are within the scope of the EKIC, and are included
for convenience. Fuels and oxidizers of current interest are discussed.
The corrosion data which are presented apply to storing, handling, and con-
trol equipment outside of missiles and to missile components up to the
combustion chabm.er. The compatibility of materials with reaction products
in combustion clsabers, nozzles, etc., has not been conaidered.

Boyd, W. K., Berry, W. E., and White, E. L., "Compatibility of Materials

With Rocket Propellants and cidizers", Battelle Memorial Institute 21IC
Memorandum 201, January 29, 1965.

This memorandum summerizes compatibility data on ammonia, boron-fuels,

halogen oxidizers, bydrazines, hydrogen, hydrogen peroxide, methylene
chloride, hiti. acids, N2 C, oxygen ozone and solid propellants.

4o.
Brobacher, W. G., and Lashof, T. W., "Bibliography and Index on Dynamic
Pressure Measurements", Natl. Bur. of Standards Circ. 558., February 14., 1955.

This bibliography and index is one of a series of reports intended to sum-

marize the state of the art in various areas of the field of instrumentation.
The circular contains 850 items on dynamic pressure measurement and related
subjects such as static pressure measurement and general information on the
comppnente of instruments. An index of the bibliography by both subject
and author is included.

12
 Brcebacher, W. G., "Bibliogrphy and Index on Vacuum and Low Pressure
Measurement". Natl. Bur. of Stafndards Monograph 3, Nveer 10, 1961.

This bibliography and index on vacuum and low pressure was prepared for
scientists, engineers, and others. The bibliography contains 1538 refer-
ences, of which 52 are on books. About 550 of the periodical references
are specifically on pressure measurement, including both vacuum gages
and micromanoeters; The balance are on vacuum technology, including
adsorption, degassing, vacuum pumps, controlled gas leaks, vslves, seals
and vacuum systems, all of which bear on the technique of vacuum measure-
ment. The indices consist of an author index and an index of the subject
matter of the listed references.

42.
Brombacher, W. G., Goerke, V. H., and Cordero, F., "Sensitive Aneroid
Diaphragm Capsule With No Deflection Above a Selected Pressure", J. of
Res., Natl. Bur. of Standards, 21, January, 1940, pp 31-32.

A sensitive diaphragm capsule has been designed in which the two corrigated
diaphragms nest into each other at external air pressures above a desired
value. Evacuated capsules of this type have particular application in
measuring pressure with radio sondes of the Diamond-Hinman-Dnmore type,
where the deflection of the contact arm is fixed. At the ground level
the diaphragms are designed to nest into each other and deflection does not
begin until the air pressure is reduced to the value called the cut-off
pressure. Several elemnts of two capsules each were constructed with a
S*cut-off pressure of 140 millibars. When installed in a radio sonde in
place of the usual type, which is responsive over the entire range of
pressures, a sevenfold increase in sensitivity in pressure measurement
was obtained at altitudes above 46,000 ft.

rombacher, W. G., "Some Problems in the Precise Measurement of Pressure",

Instruments, j& April 1949, pp 355-358.

Engineering and scientific literature reveals a steady stream of articles

on new or improved instruments for measuring pressure. Broadly speaking,
this field my be held to include both indicating and recording instruments,
remote indicating instruments, pressure regulators and controllers. Mea-
surements may be made dynamically or statically. Oly indicating instruments
primarily for static use are discussed in this article, although the dis-
cus ,.on may have application to other fields.

Broier, D. F., "What You Can Do With Magnetic Pulse Forming", Metal Prog.,
8T (4), April 1965,. pp'79-83.-
Magnetic pulse forming is used in mwn assembly operations in the automotive,
aircraft, and aerospace, electrical and process industries. Typical ap-
plications include joining of power transmission torque tubes to end fittings,

13
 45.:

Circhrirk1fitting oflgiffierltfoumed fro a gou

tof ntesial ohprfaving

widely varying strain-hairdealng rates. The coag~ete develount of the

shapes and strain dietributions was datermined experimentally, azd the
stress anid radius of eurvature at the pole were calculated as a function
of the =mam~ stratz. An analysis off the data revealed that strain
gradients and, therefore, the bulge heights were influenced by the stress-.
strain abeac'ateristics of the mettal. It was also found that the bulge con-
touar mas c1.orely apprximated by a sphere only at strains In the vicinity'
of the instability strain. Instability was exhibited by all materials having
a sufficient ductility at strains varying froct 6 - -0.T for annealed 7075
aluzinum to 6 -0.&li for annealed low-carbon steel.

46hBown. We F., Jr., and kebs, Go,, "Strength and Failure Characteristics
of Thin Circular 81men, AMU Trani., 70, April 1908, pp 241-251.
The problem treated in this paper comcemns the-deformation and failure
characteristics of thin circular metal memranes. The Instability pheoena
encountered in the deformation of such shagpes in analysed In terms of
strain distribution and also by previously dcreloped equations relating
the stress-strain and radius-strain functions. The paper also develops the
fact that the circular bulge test appears particularly suitable for deter-
mining basic stress-strain relations to much higher strain values than are
obtainable by conventional methods.

Budiansky, B., and hdkovski, P. P., "Numerical Analysis of Unsytical

Bending of Shel 3 of Revolution", ATAA Journal, 1 (A, August193
pp 1833-18142.
A general nmrical procediwe,, based on the linear theory of Sanders,. is
developed for the elastic stress. and deflection analysis of a shell of
revolution subjected to arbitrary lo&as and temperatures. The shell may
have variable and discontinuous, but aiyteric, geometrical and mechanical
properties. The procedure involves the expansion of all pertinent load,
stress, and deformation variables Into Fourier series in the circumferential,
direction; the individual Fourier com;onents of stress and deflection'then
are found separately by matrix solution of the f inite-difference forms
of appropriate differential equations in'the meridional, coordinate.
48. -

Dulgakovy, V. N.X Sbtics of Toroidal 82ls, Kiev. AM~. AM&'MWiB i9ft-

Thi6 Is an extensive monopaph on the analysis 'of t~woidal shells. Afteew'

the deveomnt of tba theory of shoe of rotrtion and of torbFoashels
the author discusses numerical methods of solving*the "ro6lems Vh.~ol,
methods discussed are the Runge-fttta integration and the finite difference
approach. Tables of influence coefficients for torus shells are presented
and a number of exasples are calculated.

Bulgakov., V.. N.,, "Application of Therical Yatbods to the Calculation of a

Toroidal Shell" 'Trudy los. po Teorli Plastin i (boloobsk, Kaia'
1960, pp 4-145 (in Russian).

This is one of the earliest Russian papers on the ma-wical solution Of

toroidal shells. Both Runge-mkitta, Integration and finite-difference approaches
are discussed. No numerical result~s are given but the author states that
two toroidal shells were solved by the Runge-Mrbt approach on the comabzing
machine, ' tftrella", which coincided with the solutions given by Clark.

Blgakov, V. N., "Calculation of Torus-Shaped hflaws for Axial Load and

fr-ternal Presmire", Sbe Trudoy Lao Gidravl. Mash*, ieyv, Izd. AN DkOM (9)
1961, pp 94-102. (in Russian)

A cauter progran had been written earlier for the calculation of stresses
in shells of revolution using the finite difference approach. This progran
was used to calculate stresses In seven omga.joint billows.

BlgakoyV ., and larpenko, 1. V. Nmrical Solutions of aooeneous

Eqainof a Toodl*I, b rdvLb Gidravl. Wuh., Ke. AN UkBM
(9), 1961, pp 89-93 (in Russian).ev
This paper discusses the application of the Runge-lutta. technique to the
nuerical integration of the equations for a general @bell of revolution.
A coiaputer Progei written for the Russian mahines "trefla was Used
to solve two toroidal shell problm and the results are cooared with
Clark's analytical solutions.

52.
1hulgakov, V. N.,. "A toroidal SbeU Under the Effect of Centri1fupLI Forces",
Prikl. Nekh.. I (2), 1957 (in Ukranism).
This paper deals with an exmnsation of the calculation of a rapily rotating
toroidal shell filled with a fluid. V. Novohilov' s cat] theory of thin
shells is utilized. The principal equation is deduced for the tooul
shell at a given loading. The solution of the hacoeneous equation.As .
given in the form proposed by Zen6v - NrvozilV." The jstial solution
of the nonhoaogeneous equation is found by the method of variation of
arbitrary constants. The special functions entering. into the solution are-
tabulated. The effect of boundary conditions upon the stresss
s in the
toroidal shell is considered.

Burry, P. E., "The use of a Nonuniform Diaphragm for the Reduction of

Acceleration Errors in a Pressure Transducer", Tech. Memo 3o. T.D. 3T,
Rojal Aircraft Eatablishment, Farnborough, September 1958 (AD 207266).

Several types of pressure transducers now available or under developsant

use integrally 'mchined diaphragms. It is shown that by reducing the thick-
ness of the outer portion of the diaphragm and increasing that of the central
portion, a redu-ed sensitivity to acceleration may be obtained while retain-
ing the original pressure sensitivity.

Campbell, J. E., "Review of Current Data on the Tensile Properties of Metals

at Very Low Temperatures", Battelle Memorial Institute MNIC Rept. 148,
February 14, 1961.

In reviewing the current data on the tensile properties of metals to liquid-

bydrogen temperature (-423 F), some results may be smmrized as follows:
(1) face-centered cubic metals, some cobalt alloys, some magnesium alloys,
acme titanium alloys, tantalum, and zirconium tend to retain zonsiderable
ductility at very low temperatures; (2) to minimize brittle fracture tendencies
at very low temperatures, special precautions should be taken in design and
fabrication to avoid stress concentrations in equipment for low-temperature
use, and (3) qualification of welding or Joining techniques for equipment to
be used at very low temperatures can be accomplished only by testing appropriate
specimens at the service temperatures. Specific recmmeeations are given for
a variety of metals.

Cardullo, M. W., "Metal Expulsion Diaphragm Development for Sherkal Pro-

pellant Tanks", Minneapolis-Honeywell Aeronautical Dirn., Decemer 12, 1962
(N63-19993.).
This report describes the design and development of an 18-inch ueta expulsion
diaphragm for spherical prop.llant tanks. Ten of the convoluted diaphragms
were tested with water. Units evaluated were of 0.010 and 0.015 inch thick-
ness. Tests conduated on both sizes indicated that over 90 percent of the
volume above the diaphragm would he expelled with a difrerential pressure

16
 of less than.,4 psi. However, expiulsion efficist or
o 98 pemat 0-
upward- of 25 pat,. Rteslts of this progpa indicite that'the fN iqn -

* ~method evolved* wwtld be applicable to a vide ranse of -4iPbrsg_ izW

I. Carey, R. Nt elded DIragA Metal' Bellows", KM

and Systems Design, 1 (8), August 1963, pp 22-25.
tr chaical, Co~ironts

welded-diaphragn metal bellows consist of metal contors., called disphragas,

welded in series at the inside edge to form convolutions. A series of convo-
lutions are stacked ad welded at the CD to form a capsule. h fitt!Ms
are then attached tb the- CpsUle to form the z4inal- asembly. A few typical
applications of velde&.-diaphraga metal bellows together with. th1e adVIAnt~e's
for each aplication are described in the article.

57.
~Carey, R. N., "Eight Applications for Metal Bellows', Prod. ftg.,
November 26, 1962, pp 58-59.

Brief descriptions are given, of metal bellow$ used as Precision couplings,

fluid accuilstors, abaft sowls, and Pr-essure actuators.

Carter, 3. C., et al., "MYA&XneMnt Of awrfaco Strains In Dnapbrasws"

Engineering, 168, Decmer 2,. 1949p p 581-583.

In the employment of metal dlspbragias in presue gas it is almot In-

variably the Practice to measure the c~entrsl deflection Of the plate. This
method bas good, sensitivity with moderate pressuresp up to, say, 100 lb per
sq in. As the inm pesmr is increased, hoever, the plate thickness
needs to be greater in relation to the dism~ter and the maxis- deflection
decroaaet unless the diameter is Increased sufficiently. The permissible Press 8e
diameter being limited,, special methods ore described for use at high measure-
to measure deflections with sufficient accuracy to obtain Pressure
writs correct to within *1 percent.

2 Casaccip S. X., and Picollir, G., "'A Stu4~ of the Bending of Constant-
Thickness Toroidal Shlls Uer Aiaiimtrical, Loading", Boille Blanche
* (17), p l1-38, Jazmmy) -Feruary, 1962 (in ftench).
Is
The anaytical study' of the axisymtricil beading of toroidal shells
* always highly complex, eve If only small elastic strains ore considered.
in
For certain loadie cases, aPproximte analytical solutions can be found

17

t -3
 the form of a linear c1bination of two Heakel ftactions of order 1/3. The
particular solutions for other loading cases are either eHised-by mens-
of a Fourier series, the coefficients of which are calculated by maans of-
continuous fractions (V.V. Novoshilov), or by cobining si~le fctions and
a Loeel function (R. A. Clark). In the method described, the basic equa-
tions are transformed into a Fredholm integral equation stem, analytical
*solutions of which can be obtained with the aid of asymtotic expressions.
The authors restrict themselves to an exmination of a noerical slution,
which is applied to the calculation of stresses and strains in a toroidal
bellows element undergoing axial loading.

60.
Cassidy, L. M., and *Grattan, R. J., "Computer Techniques for Stress Analysis
of Reactor Vessels", Nual. Congress Paper No. 57, 196.

7he authors attempt, in this paper, to demonstrate the facility with which
certain stress analysis problems associated with reactor pressure vessels can
be solved through the use of high-speed Computers. Some of the solution,
and methods outlined have been previously presented and are further dis-
cussed; others are believed unique to the authors. The analyses presented
have been programmed Cor the IM 704 digital computer.

61.
Chapman, M., and Kirk, C. R., "A Survey of Some Fast-Responding Pressure
Transducers", Royal Aircraft Establishment, Farnborough, Ralaud, I.R. 15,
ovember 1962 (AD 41O886).
Descriptions are given of various types of zransdicers, incluling some-
which use the movement of diaphrags to vary the electrical resistance be-
tveen electrodes.

62.
Chandler, W. L., "Bellows in the Refrigeration Industry", Refri@erating Eng.,
A_, Part I, pp 29T-299, November 1941; Part II, pp 381-387, Deee r 194i.
Brief design formulas are given for determining the spring rate, pressure
resistance, and effective area of a bellows. The use of two-ply bellows
and the estimate of fatigue life are discussed briefly.

Chernina, V. S., "Calculation of a Tubular Compensator for Extension and

Bending", Energom&shinostroenie, 8, 1959 (in Russian).

The asymptotic solution for toroidal shells given by V. V. Novozhilov is

used to calculate the stresses in an omega joint under axial tension and
bending of the attached pipe, and the results are compared with experimental
data.

18
Ibria V.S., " Ulit ion of TorOIG&IMA

seies ~utiostir ts
The athorobtais
~e

torus radius Is lato~r. than the distanob fra oi

um

Iv14
*tonr-evltot,
Iiidatl

center of the torus so that the shell Intatsects the- azis of revolution,;
The second problem is concerned with the other extrm for which thetou
radius is mmli mnd the distance from the axis to the center of thetou
Is very large.

Chernima, V. S.. * "( the krtem of Differential iiaxtions of -SquI1biriam of

Shells of Revolution Ume 3Snding Loads'.. Prikl. Nkt,. Nekh., ?, pp 258-265,
1959 (translation).-
When the nozxiyintric l oads of a shell of revolution. are represented in
terms of a Fourier series In -the cfrcuferntial. aog2le the differential -

equations can be red13ed to an eighth order system of 6rdinary difflerOn1"a

equations. This paper. shows how this system can be pertially Integrated to
obtain a fourth order. system If the loads deen an ;a*~ the sine cosine

6.Chernin, V. S., "The Stressed, tate of a " a,. fCnoltdBloa.

zoergowshinostroienie, IL 1961 (in Rus sian).

The asyjqtptic solution for the deformtion of a taroideL *31 is used to

calculate stresses 1i a aemitoroidal bellm~ss and the reOsults are eped
with sow experimental inse-remuts on one bellows.

Cheaia, .~S., "Ibe a0ressed State of a Toroidel ash of Msdiu Thickoess",P

isv. AN ,~MN, Nmkh. I 11ah., jo p 90-l14p 1959 (In Russian).
Tisi paper consifts the dftwation of toroidal shells -of inedimn thicknessp
by expanding the shell sqaios In pawens of the ratio, of thickness to local
canvature in the manrof Seimer. The cures we given caing the
stresses obtained by the 2100W tbeoY of thin shellIs aud atheory includin
second-order thickness corrections.

19
 - - -NEW

68.
Chernina, V,. S., "Tbe ftyessed State of a Wbular Conftusator 0,ar41ng.
Under Beading Co-tos. Inzh. obI, 2%* 1955, pp 133-14s9' (in tUsiiaa)..
This paper cons iders a toroi~s1 bellovs unde bonding loods. -Tha -boautioc
is*ctdmee using the mpte9 :c theory of !tcvobtlov. Galqulat ion are
made fcw amn troidal bellows anid coops.ed, vith -sem erperimental data.O

Cberrainca, K. F.,. "0, the Vleraxe Theory of Shells%, In Thyeatications on

Jla.Zct and Platicli, . 3, Leningrad Vniv., 2L~1A, PP 3-23 cin Hussftn)-
The author of this paper attempts to develop criteria for determiting theai
the meabrant theory of shells is applicable. S~ exa;iles are disc-ussed.

70.
Cherwkh, K. P.., %he 'Idwar Theary of SasI15b IT, eningrtA Stat.- Univ., 1962
(nRustiah).
This is the vecorId of two Yvb.imes ani the linear theory of shells. In ad1di-
tion to a discussion -of merny aspects ofl shela theory, this voue includes
a' discussion 6f the toroidal shell,, torus-shaped bell.&,s,, and senitoroidal
beflois * A solution is obtained for these shells using Noyozhilov' s asyup-
totic approach,

Cherith, K. F., and Saaina, V. A., "Calculation of Toroidal M*Us, 1'..

Invest tigns; on Ilasticit *ad Plasticit" , 2, ab, (3) of the Leingrad
1nv,9b3, pp,247-346 7in.Russian).
This is an extensive survey article on the state of the art of the theory of
torodal1 shells tbrough 1962. kNyw of the landmark theories for toroidal
shellIs are presented and discussed, incluliing Stange'sa, Dhl'a, Clwklz', an!I
Noahiloy' s. A number of the papers an fame-d bellows of variousx types are
discus sed. The references are a valuable cross section of the Russian litera-
ture In the field tbrough 1962. Some use of nluerical appoaches by Bulgakoy
In Russia is noted.

72.
Chi, IQ- 3an, "Bibliography and Tabulation of Daming Properties of Nonwtallic
Mter~als", Univ. of Minnesota, WADI-.T-60-5l40, September 1962.
This bibliography vas compiled as an aid for those people interested In
damirig research and related fields. it contains a nearly complete list of
references on the damping properties of nomaetaflic materials for the period
fron 1929 to 1959, together with an index of nownclature, units, and test
methods. Tabulations of the in-phase and out-ct-phase coqponnts of Young's
modulus and the shear modulus for the various naterials.are also shown.

20
 Chizhovy, V. Fo., "Staility of CorruaLted Cyrlindrical Shells Loade by luar
Pressuire", Izv. Vuish Uch-ebn. Zaved. hesh., ,16f p 3.i (n.s~)
This-paper considers the buckling of bellows under external pr1e he
approach used ls to consider the bellows as a cylindrical shell idth redwced
effective rigidity equal to that of the bellows. The effective rigidity :is
calculated from the corrugation shape using epergy methods. Buckling loads
for a number of be3llows were calculated and cmpared with experiments. noe,
experimental critical loads were from 60 to 90.percent of the calculated--

Clark, R. A., -Asyptotic Integration In Shell Theory With Applications to

Toroidal-Shell Expansion Joints", Contract No. DA-33-019-CRD..l996, October
15, 1955, through Cltcber 14., 1956.
A method of asymptotic integration of an hc~nous differential e~w~tion
containing a large paramter, developed in apeiu report, is applie
to the differential equations of the theory of thin elastic shells af revolu-
tion. A particular problem of a toroidal-shell expansion joint subject to
an ax-ial Icad is considered. haed on. the assuqtion that a certain parameter
is large, explicit expressions for the- nuax. bending stress, mximm direct
- - stress, and axia deflection are obtained for a shell of arbitrary radial
cross section and varying thickness.

Clark, RI. A., "Asymptotic Solutions of Toroidal Sbell Prci,1tasp', Quart. Appl.
Jath., 16, April 1958, ?p ii759.
A few years ago a moethod. of asymptotic Integration was develop~d by Z.
Reiusner and the author for a nombmgeeh differential equw,.on involved
In certain problems of the theory of thin elastic toroidal shells. Using
this method,. solutions were fod for the prcblem of bonding In a curved
tube (considered as a sector of a shell of kevobition) and for the uthe-
matically. similar problen of a toroidlal expansion joint subject to an a=i8l
force. The pupos of this paper is to Indicate how the method used before
can be refined and how the appoxite solutions previously cbtained for
these problems can be Singsalized and extended. N -eorderterms or cor-
rections of two different typs are obtained for toroidal shells of circular
cross section and uniform tbickneos. Sm of the results are fo=md by
specializing solutions derived first for toroidal shel of arbitrary
cross section and possibly varying. thickness, asuming only that the abell
has- a plane of syetry predicular to the axis of revolution.

K-6 Clark, R. A., andL Reissser, Z., "Deformtions and tresses In Bourdon
Tubes", hs, , D.ep.
cember 1950,pp33i-il

The state of stress and defontion In the tube sm be considered as; a

superposition of two states: (1) stresses and deformetions in a comlete

21
 torus miject to wll. pressure; (2) stresams and dtformtions n n~qW
plete torus acted upon by bending m nts at thea onds.- t.f -1.1 "" " atjj
as a pr"lea In the theory of axiuiineti teul &fonmations ot-thin shells- of
revolution. State 2 is treated In the seem mau'e after introduftion,.into
the theory of a nona iqwimtrical compoet of clrcufteren ial dfsiac"nt

TT..

Clark, R. A.,. "C the Theary of Thin Mastic Toroift1 ftellsou, I. Hath.
Phys., K, Otcber 1950, pp 146-1T8.

In this paper, toroidal or ring shell problems are considered from the point
of view of the small deflection theory of thin shells of revolution loaded
found by applying methods of asymptotic integration of a differential equa,-
tion involving a large parmter multiplied by
a function which vaitihes at
certain points. As examles of the general procedre develoed the particular
problems of an "(neW~ piping expansion joint sub~ject to axial load and of
a corrugated pipe subject to axial load or internal pressure are solved.
In particular, explicit formulas are given for the mxIon stresee *nd
displacements in terms of the dimensions of the shells.

Clark,' R. A.,0 and Reissner, E.. "A Problem. of Finite Bending- of Toroidal
Shells". Quart. Appl. Math., 1.0, January, 1952, pp 321-3-~U.
This paper is concerned with a nonlinear problem involving finite axi-
symetrical deflections of toroidal shells with circular cross section. Mile
problem is that of an expansion joint for two strsi~t sections of a cylin-
drical shell loaded in an ax&ia direction. Tim particuar kind of expatd5ioz
joint investigated is generally called an "Oeega" joint.

79.
Clark, R. A., Gilrcr. T. I., and Reissmwr, E., "Streas e nd Defomztions of
Toroidal Shells of Elliptical Cross Section", J. Wp. MWch., , 20 &L-c
1952, PP 3748.%I
This paper is concerned with the application 'of the theory of thin abells to
several problem for toroidal shells with ellptical cross section. These
problems are.as follovs: (a) closed shell j~c to uniflom normal wll
pressure, (b) open shell subjectdd to end bending -- Is, (C) couinstion
of the results for the first and second problem In 'mh a vw as to obtain
results for the stresses and deformtions In 3oiardaii tvbes. In all tbree
problem the distribution of stresses is axially uymtric but onl~y in the
first problem are the displacements axiaully symwtric. The mWgitu4e of
stresses and deforuatiour, for given loads depends In all thee pobles on the
mnagnitude of two parameters.

22
 Cohen, G. A., "Computer Analysis of A7mmtrical Dfction of Othato"Ic

gratimt is developed for the awe general case of artbotropic shells sb

ject to sinusoidal loads.

StandardB Cryogenic Na. Lob., July 3960.

This first phase of the Cmendium covers ten properties of 10 flids (Part
1), tbree propeties of sol Ids (Part 2), and an extensive bib.iop&#a of
references (Part 3). Density,. expansivity, therl condutivitY, specific
heat and enthelpy, transition heats, phase equillbria, dielectric cOwitSt,
adsorption, and surface tension and viscosity f=r the solid, liquid ad gee
phases of heliu, hydrogen, a:d neoun, nitrogen, ongenp airs arboa 4=oxide:
:fluorine, argon, and methene we given vbarever a&quate data could be collected.
Thermal expansion, therm]. conductivity, and specific heat and entelpyf are
given for a. number of solids of interest In crrogenic engineering. ])ta
sheets, primarily in graphic torn, are presented frcu "best values' Of data
collected. The source of the mterial used, other references and tables Of
seleted values with apprcypriate cmnta are furnished vith each daxt&abseet
to docmmt the data presented. Conversion tables and other helpful Wfrw--
tiom are also incluied.

82.
Cook, W. F., "Stresses In Seal Welds".. Knolls Atomic Power Lob. Rept. 1o.
WKA-WC-l, July 20, 1957.

Ti. soluticm for stresses Inaw-ed in sea welds by displazementa ana preszure
fores is shown. Teresidxs are in the tc~ra nf ctrief f'2 vhich strem5ex
ma be calculated.

8.Cooper, R. Me, and Mifrin, G. A., "An Experiment on Circular Platee In the
Plastic Rane, Proc. of the Second U. S. N41. caers for Applied Iuclanica,
AM..1954, pp 52T-534i.

Nine simply siqpcrted atid-steel. circular vaztes Vere loaded, well I'to
the plastic range umdsr coamentric.. =ifmif ly distributed I.a,a IWtzActQ'1
strain and slope ata vwe cbtsinsd as ftmctionis of !oai. ~ei" ~
sade vith the predicted limit loads or~ Sopkin and P~

if
_______ _ ___ ___23
 Cope, E. T., and Wert, E. As., "Load-Deflection Relations for Larile Plain,
Corrugated, and Creased Pipe Pentis", Trans. of A=,K, PP.5412, 195-.

This paper presents the results of tests made'on 10-in, and 14-in. sealess
steel pipe loaded in the same manner as the. simple -cases shown on PgA* 526
to R29, "Piping Handbook", by Walker and Crocker. C aison is madi between
calculated deflections and the results of tests in the six case for &,plin
pipe bend. The relation between the results- of tests on corrugated and creased
pipe bendi and the calculated results for plain pipe bends of-the samie Aimen-
sions are shown. Such calculations are included as are necesay to shbiv-the
method of analysis by which the load-deflection relations of a plain quarter-
bend and its tangents are determined*

~Cordero, F., et al.., "A Nonlinear Instrument Diaphraga", J. of Research,.

Natl. Bur. of Standards, jq (6), June 195T, Pp 333-33T.
Details of fabrication for the production of sensitive diaphragms having .a
cm-trcaUed nonlinear pressure-deflection characteristic are presented. The
desired characteristic was such that when the diaphragm forwd one plate of
a condenser in the frequency-controlling network of a Wien-bridge oscillator,
the. resv.ting pressure-frequency-transfer characteristic would be linear b e-
tween -30 and +30 dynes per square centimeter. Typical transfer curves are

86.
Crites D. A "For Stress Analysis - Brittle-Coating Methods", Ptod. Ing.,.
L2 (Q25, Noveber 2T7, 1961., pp, 63-72.

acnd in a series on stress-analysis techniques, this article describes howr

to select costings, prepare and analyze crack patterns, and discusses a new
micro technique.

Crites, N. A., "Your Guide to Today's Strain Gages", Prod. Ing., (I)
February 19, 19q, pp 69-81. -

Third in a series on stress-analysis techniques, this article discusses

the popular typyes and methods for installation and protection.

88.
Criteso No As., Grover, H. J., and Hunter, A. R., "eimetal Stress Anal-
ysis by Phtoelastic Technriques"., Prod. ngo., U (18), Setember 3, 1962,
pp 5T-69.
Fifth in a series on stress-anaysis techniques, this article gives informa-
tion on equipment, miaterials, model calibrations, and stress interpretationsl.

24
 89.
Cwuuing, E. P., "Thirteen Ways to Use Metallic -Bellows", Prod.
(6), June 1952, pp 162-163.
" Sketchets show typical and unique applications ot metal bellov, sud warys Of
transferrIng motion to other elements.

Cushing, F. S., Chreitzberg, A. M., "Research and Developlent on Cells With

Bellows Controlled Electrolyte Levels", Electric Storage Battery.Co. First
Quart. Rept., June 10 to September 10, 1964 (f64-33918).
Preliminary Cd/KOH/NIOOH cell tests have verified that bel,!ws action in a
sealed cell can be used to flood tbe plates under low pressure conditions
to maximize discharge capacity and to drain the plates under high pressure
overcharge conditions to aid in gas reccbination. Metallic and nonmetallic
bellows were considered. Plastic pillows, partially inflated -rith air,
and sealed within the cell, are proposed as bellows substitutes. The pillow
AVfV value is approximately twice the present metallic bellows volume
efficiency.

Dahl, N. C., "Toroidal-Shell Expansion Joints", J. Appl. Mech., 20, (4),

December 1953,. PP 49T-503.
In the problem considered in thispaper a thin toroidal shell of center-
line radius a an itrus mideurface radius b, slit at its inner edge and
welded to relatively stiff cylindrical pipe sections to oran expansion
joint. Stresses and axial deformation under the axial load P are calculated.
A solution is obtained using the principle of minimum complementary strain
energy. It is found that the results obtained in this way are valid as long
as p is not too large. This solution is plotted together with the asymptotic
solution obtained by Clark to show in what regions of p the solutions over-
lap, and both solutions are compared with experimentally determined stresses
and deformations.

Daniels', C. M. "Designing for Dtuct Flexibility With Bellows Joints", Machine

Design, a (211, October 1959, pp 14&6-155.
In duating systems, bellows offer flexibility to absorb thermal expansion
and contraction, installation misalignment, structure deflection, end vibra-
tion. This article presents a review of the factors that influence their
selection and application. It is based upon the design of aircraft and
missile fuel systems where ducts run the gamut of extreme temperatures, pres-
sure, and flow velocity.

25
 93.
Daniels, C. M., "Pressure Losses in FM.xible Mtal fting", Prod. Ing.,
April 1956, pp 223, 225, 22T.

Friction coefficients and pressure loss data are given for corrugated flexible
hoge of annular and helical types at high fl64 rates (Renolu Aider 2'x.

Daniels, C. M.., and Fenton, R. Z.., "Pressure Lose Factors for Internally
Linked Bellows Joints", ~i,3 September iii 1961, pp 18T-189.
1) .~hn

This article presents experimentally determined loss factors for fluid flow
in two types of internally linked bellows Joints: the chain-link type, and
the gimbal-ring type.

* 95.
Ianilova.. I. N., "Calculation of Toroidal Bellords Under the Effect of Axial
Tensile Strength During Creep of the Uaterial", Izv. Akcid. Nauk SSM, Mdp,

I 10, 1958 (in Russian).

This paper considersa an omga joint uzdergoing creep under axial loading.
A method of solution is sketched for an exponential creep law.

DeCrescente, M., and Janz, G. J., "Brass Bellows Gauge for Null-Reading
Manoeters", Rev. Sci. Instr., 289 (6), P 4i68, June 195T.
This note describes a mechanical gauge made from comrcially available
ca~onents and requiring a minimum of machining and glass blowing. The
gauge is capable of accuracy at room and elevated temperatures coumparable
to that of the mercury manometer-cathetcmeter'assembly.

27± Delmonte,. J., "A Versatile Miniature Flush-Diaphragm Pressure Transducer",

ISA Proc., , pp 171i-177, 1952.
A variable-resistance, unbonded strain-gage type of electrical pressure gage
is described that is basically 1/2-inch In diameter by 5/8-inch in thickness,
and weighs about 20 grams. "Lhe gage was designed to measure pressure
fluctuating at high frequencies as well as measuring steady-state pressures
with errort of less than one percent of full scale. The gage has been ten-.
perature compensated over the range of plus 165 deg. to mInus 65 deg. Y, and
its output is convenient to drive recording galvanometers directly.

26
 98.
Deneff, G. V., "Fatigue Prediction Study", Dou gAs.rratt'-Co. ,ep1ort' -
WA1D TR 61-153, January 1962.
Fatigue life prediction of complex structure is investigted frwthe stand-
point of fundamental lactors that influence the preoiction; namely, stress,
fatlgue strength, and damage method. Stresses developed by a general stress-
- analysis procedure are coined with appropriate fatigue-strength data to
estimate the fatigue strength of a structural Joint. Factors influencing
the joint fatigue strength under spectrum conditions are anayzed an& a
method of estimating this fatigue strength is presented. Dnmge.methods ,uti-
lizing several types of fatigue-strength data are considered. The influence
of geometric factors on the fatigue strength of basic material is. also studied
and presented in a normalized form.

2 1I-ickinson, T. A., "Problem: To Make These Bellows" Welding Eng., 3 (10),

Pp 17-53, October 1954.
A brief description is given of resistance-welding equipment developed for
making thin, stainless steel bellows.

100.
Dolginov, L. S., "Calculation of Steam Pipes With Convoluted Bellows Without
• Tie Rods"., Sudostroyen~ye, ,.1959 (in Russian).

This p&e considers the use of bellows as motion compensators in a piping

syztem with offset pipes. A method is given for considering bellows without
tie rods. This permits considering any part of the steam line taking into
account the flexibility of the bellows both in axial compression and in
bending. A successive-approximation technique is used to obtain the solu-
tion for a given problem. One nerical example is given.

101.
101-Donkle L. B., "Now to Avoid (Corrugated) Metal-Hose Failure",
Prod. Eng.,
.1 (30), pP43-48. July 25, 1960.
Metal hose in bih-pressure, high-temp.eature, flexing applications can
last a hundred years - or fall in minutes. The author discusses causes of
failure, and tells how to overcome them.

Donnell, L. H., "The Flexibility of Corrugated Pipes Under Longitudinal

Fo--ces and Bending", AM Trans., 5., pp 69-75, 1932.
Pipes with circumferential corrugations are used in expansion bends in steam
lines and in other applications where longitudinal or bending flexibility

27
 is requifred. In this paper the longitudinal fluczbilty i aclte o'

I several types of corrugations, and from this a 'reiidii ~lu 10244tik Y"
isderived by 4fhich corrugated pipes can be calculated as if they were simth.
A set of experiments is described vhich check these repalts.,

103.
Douglas Aircraft Co., Missile and Space Oystems Divisions Test Rept.p
"flellows, Duct, Oxidizer Tank Vent Qualification", January 214, 196A (*3 143333).
Tests of a 3-3/4-inch-dimter bellows included com~vession upring~ rate,
proof pressure (44 psig), leak (bvbble), defiection-pri!*sure (44 Pi -e" -300
F), life cycling (50 deflections), pre.jsure cycling (200 cycles 0.29 pilg)$
and burst.

1014
Douglas Aircraft Co., Test Rept., "Qualification, Bellows, FUelI Tank Vent-
External", December 9, 196149 IZP U15 56.69.00-v(-09.
Tests of a 6-inch-diameter bellows included proof presure (61i puig),
leakage (bubble), spring rate pressure cycling (200 cycles), ambient
deflection cycling (50 cyeltsS, life cycling (200 cycles), and burst (1900
psig).

105.
Douglas Aircraft Co., Test Rept., "Qualification, Bellows Tee, Puel Tank
Vent", December 9, 196&, ILZP l15.56.69.00.DT-o8.
Tests of a l4-in--h-diameter bellows included proof -pessuro (e7 P8141,
leakage (bubble), spring rate, pressure cycling (20D cycles).. SIixt
deflection cycling (50 cyclIes),9 cold deflection cyl- (150 cycles at _31&
F), life cycling (2000 cycles), and bure.t4(500 psig).-1

106.
Douglas Aircraft Co., Missile and Space *rstems Diva., Test Rept.,. "Qulifi-
cation, Bellows, Duct, Oxidizer Tank Vet"., Pecmber 9, 1964,p IMR 1556690O717-
Tests of a 5-Inch-diameter bellows included proof pressure (14 peg), leak-
age (bbl) spring rate, pressure cycling (200 cymcles), deflection cycling
(50 cycles) cold deflection cycling (150 cyclex at -300 F), life cYcling
(200 cycles 5 , and burst (925 psig).

28
 107.
kh'sAicraft Co., Missile adSopa Sysem Dimn., Test. Rept., WJL
cation, Bellows, Propulsive Fuel Vent", 8Stjj;r' 15, i*~l. $, e00..DI.05.1
* Tests, of a £-iziph-disaster befllws included proof presuw. (&6. psig), lag
£(bubble), spring ;ate,,. pressure cycling, deflect' (5 cc,.
.y3a
coldI deflection cycling (150 cycles at _350 F).. szld burst29Opi

108.
Douglas Aircraft Co., Missile and Space Watans Divn., Test Report sie
10
Development, -Positive Upuloion System Auxiliary Propulsion ~e"
August 3., 196&4, iiz l5.56.6.o0-D-C&.
Tests of 6 welded bellows included leakage (bubble), deflection cycling*(10
cycles), proof pressure (I&5 psig),. 4 1bration, life cycling*(1000 cycles),
and spring rate.

Douglas Aircraft.Co. Test Rapt., "Qualification,, Bellows, LOX Tank Fill

& Drain Line, Fuel Tanik, Deceber 9, 19&q4 I=P No. 315 .16.7906DrT-0.
Tests Of aL 1l/2-Inch-diamter bellows included proof pressure (16D psig),
leaag (bubble), spring ra~e, pressure cyclingI (200 cycles) siient ileflec-
tion cycling (50 cycles), cold defl action cycling (150 cyciesS, life 6ydling
(150 cycles), and burst (750 psig).

ll.Douglas Aircraft Co., Test Bet.,. "Qualification, Bellows, LCK Tank Fill &
Drain Line%, Decnter 9, 19611, im iNo.U.16.79.0-w-0l.
Tests of a 5-l/2-inch-damter bellmw Incldd proof pressure (215 psig)*
leakag (bubble-), srnrate, pressure cycling (200 cycles)o ambient deflec-
tion Cycling (50 cycles, cold deflection cycling (150 Icycles at -320 1P),"
life cycling, and burnt (1900 psig).

Douglas Aircraft Co.,. Test Rept., "Qualification, Bellows: LaK Tank ?ill &

Tests of a 4-/2-inch-disamter bellows Included proof pressure (11o psig),

leakage (bubble)., spring rate, pressure cyclig (200 cycles), ambient 'deflec-
-I tion cycliug (50 cycles) cold deflection cycling (150 cycles at -320 F),
life cycling (150 arlesiS, and burst (3000 psig),

29
 Douglas Aircraft Co., Test Rept., "Bellows, 4,000n V&, -
fication Test", July 22, 1964 (AD 45441).
Tests of a 6-inch belos included proof pressure (17 p, g), .1k (be),
spring rate, deflection cycling (1000 cycles), pesr yln 10
cycles at -320 F)., and burst.

-- Douglas Aircraft Co., Missile & Space Systems Diyn., Test

Rept., 1.e lows
Assembly, Relief, Fuel Tank Qualification Test", July 2, 196& (AD-45 20).

Tests of a 1/2-inch-diameter bellows included spring rate, proof pressure

(64 psig), leakage (helium), deflection cycling (50 cycles), pressure cycling
(0 - 42 psig, 200 cycles at 345 F), and burst pressure.

-- Dougas Aircraft Co., Missile & Space Srstems Divn.,

Tech. Memo., "Expansion
Joint, GH2 Engine Vent, D/E", November 26, 1962 (AD 416867).
A description is given of tests to evaluate two universal bellows expansion
joints for providing flexibility for an engine. The tests include proof pressure
(60 psig), leakage (bubble), pressure cycling (200 and 500 cycles to 15 psig),
deflection cycling (50 and 200 cycles), and surge (100 ]psi1).

"Douglas Aircraft Co., Missiles & Space Hystems Divn. Tech. Memo., "Oxidizer
Fill Line, Flexible Bellows", November 11, 1962 (AD 40l998).
The oxIdizer fill line assembly transfers liquid oxy en frca the ground
supply tank into the vehicle oxidizer tank. The system Includes a flexible
metal bellows which is located within the vehicle oxidizer tank. Tests were
conducted to evaluate the bellows under slwilated service conditions. The
tests included proof pressure (80 psig), leakage (bubble), pressure cycling
(0 to 49, 200 cycles), and deflection cycling (25 cycles at 49 psig).

3.-Douglas Aircraft Co., Missile & Space Systems Divn., Test Dpt., "Design
Evaluation/Qualification Test of Internal Metal Flexible Bellows", -Januar
8, 1,03, nff ll5.16.69.oo-DT-o

Tests of a 3-inch-diameter bellows included proof pressure (80 ps ,%),

leakage (54 psig, Gil), pressure cycling (200 cycles, 0 ).to 49 psi.L2).
and deflection cyclin (25 cycles, pressurized with LL2
l2
30

1 ________________________________- _______________________________
 Douglas Aircraft Co., Missile & Space W"atfts Dirn.2 Test-Rept., ckV
Disconnect, LH2 hgine Vent Wttm' Oatctme 29, 196e., I 115.56.69.0-;YT-Ol.
Test" of two stainless steel and one aluinuam bellows conducted, ,to evaluate
the design of a disconnect assembly =wder flight conditions in-cl1u*Ae4',sin1g
rate, leakage (-25"%F, 3f1 P28), disconnect (3 t~mes, -050 F,1 J.ig)
and vibration. pi

Draper, C. S., "A lew Nigh-Perfonme Zogine Indicator of the Strain-

Gage Type,, MeAron* Sal., 16 (10), pp 593-610,. 0-tober 19 4~9 .
This paper describes a new presume receiver using a high-flexibility cateflary-
type diaPhrao and a tempersture-coqensating wire-wound strain gage .da signid!
for satisfactory pressure-masuring characteristics with greatly reduced
response to temperae and extraneous mehanical effects.

f~119.
Dressler, R. F.., "Bending and ftretching of Corrugated Diapbragas" (J. Basic
Ig.) A,%M Trans.. 81v pp 651-659., December 1959.
Solutions of the exact linear elastic-shell equations for all1 stresses and
displacemts are presented for a typical co rugated-diaphraga shape for
three thicknesses varying over a 9 to 1 range. Results were obtained by
nuerical integration using an electronic digital computer. The effect of
thickness-diamter-ratio variation is discussed with respect to both stresses
and resultants, and peak values needed for design purposes are presented.
Circiaferential and meridional. stresses are found to be equally important
throughout the thickness range analyzed. Bending and membrane stresses are
likewise equally important throughout the range. Peak values in some cases
occur near the outer rim.

Drucker, D. C.,. and Hopkins, H. G., "Ccubined Concentrated and Distributed

Load on Ideally Plastic Circular Plates", Proc. of Second U.S. hatl. Congress
for AP.Pl. W~ch., AM, pp 517-520., 19541.
This paper has two main objectives: to extend the work on circular plates to
the experimentally realisable cases of large or small overhang, and to demon-
strate that answers may be obtained in a straightforward manner for any
yield condition and almost any radially symtric loading. As an example,
the Treaca yield criterion is assumed and the collapse condition is calcu-
lated for an oehanging circular plate subjected to both a central force
and a uniformly distributed pressure within the support circle. Solutions
for the simply supported and for the built-in plate appear as special cases.
 ThnSbes C.,
Drucker,D. and J-3ld,
of Revozi" Mah.,,Analysis o61-M
Ap. "Limit
j.. T., tcw ih 9w
--

The yield surface for & thin cylindrical shell is shown to be a ver gpod
approxiation to the yield surface for any y trially loiAed thln s ll
of revolution. Hexagonal prima approxisations to this yield suifice aS Ppp-
riate for pressure vessel analysis, are describei sd alacased in terms of
limit analysis. Procedures suitable for finding v1'r and lower bounds on the
limit pressure for the complete vessel are developed and evaluated. They
are applied for illustration to a portion of a toroidal zone or knuckle held
rig.dly at the two bounding planes. The cmbined end force and oent which
can be carried by an unflanged aylinder also is discussed.

122.
Drucker, D. C., "Plastic Design Methods - Aivantajes and Limitations".
Office of Naval Researcb Tech. Rept. No. 24, Brown Uni., July 1957.

Assebly and welding stresses add to the stress imposed by loading, and it is
inevitable that stress concentrations will produce some local plastic flo"
in the best of designs. An exact plastic design would be a formidable tack,
but neglecting work-hardening provides a simple and yet a reasonably satis-
faztory approximation. tered limit design. At the limit load the idealized
structure collapses. For the overwhelming majority of structural problems,
a design based qpn a reasonable factor of safety sainst this plastic collapse
provides a more appropriate structure than a desUMn based upon elastic action.
Although limit analysis is relevant, a mare complete analsis is needed for
plastic buckling, strengthening by secondery membrane stresses, and brittle
fracture. Illustrations are provided.

Dubinskiy, S. A., Rusanova., Ye I., Scol, B. F., "Calculation of Toroidal

Bellows of a Lo-Pressure Pipeline", Sudostr. , 5, pp iA-16, 1956 (in Russian).

The asymptotic solution is found by the method of NoTozhilov for some omega
bellows with either clamped cr simply supported edges. The results are
claqred with experlments.

"Electron Bea& Welds Pay Off for Bellows Maker", Steel, 148, pp 78-79,
*rch 6, 1961.
This article describes the use of electron-bean welding as an alternative
to tungsten-are welding in the manufacture of welded bellaws for use in pressure
sensors.

32
 125-11S, A- H.: an owrd, JH.,"btto Cnidr Wn Selecting.

Brief descriptions awe given of the design, fabrication, life, and ueo
metallic bellows. Aif-rd~iRcsr sicusd
* r

*126.
"tAn Engineering Evaluation of Methods for Prediction'of Ftigue Life in'
Airfraie Structuresa' Lockheed-California Co., Mmrch,. l06i '(AD 0,71A9)

Rrom a study of twenty proposed fatigue-life-prediction methods, ten of the

procedures were chosen for evaluation numrical , with a grou of 78 cbi1ex
spectral test results representing approxiately 266 indjidual speciens.
An experimental Program generated zonstant aailitude axial load 8-1 type
data on simple notched c-oupons of 7q76-T6 aluminum allc r sheet for use In
the analysis procedures. Orered figectral fatigae test data from these aasc
type coupons were utilized from another concurrent ABD fatigui research
program. A series of specins of a complex joint were also fatigue tested.
These data were analyzed by the selected procedures to confirm or provide a
possible means of improv.Ing the selected 'fatigue-life-prediction aiithods.

Engi, W., -How to Determine the 'Effective Area' of.a Disk DiaphraW%"
Regelungstechnisch Praxis (Ijnich), Y (5), pp 165-168,. May 1959 (in Germn).

Onie sid,- of a disc diapIbsg is loaded by a Pressure which is opposed by a

force acting upon thei other side of the disc.* As a result, the diaphragm is
deflected in the direction of the greater pressure. Thic article shows a
method of calculating, for small deflect ions, the "effective aree", that
is, the area from which, by multiplying it with the pressure, the opposing
force can be found. A device is described consisting ainly of a very stiff
pressure transducer which at the sane time produzes and measures the opposing
force.

Ericson, G. L.p Boyd, W. X.,. and Miller, P. D., "Corrosion of, Titantu and

THLYka.,Aprl 3,.1958.

preliminary croinevaluation in liquid and gaseous fluorine at tmea

tures between -320 I7and +220 7. A.1 materials exhibit pjccmlsing corrosion
resistance under the test conditions. MUre elaborate experiments would be
desirable to fully establish the utility of titanium in fluorine service.

33
 Iatabrook, L. H.,i and JahainziW E.L "Stress Analysis of Welded-
General Electric Tech. Inforibin Series (MV6- Wy
e) 2l)
8; 196.'
The general solution for welded bellovs ginun In KOL-1089 w.e applU4,to,
a bellows loaded by; radial edge forces and edge mmnts at its -ends. &jm-
ticms for deflection. rotations, forces, mcents, and stresses throughouat
the bellows were developed using mnatrix algebra. The limitations of the anal-
ysis were investigated. Nmerical results vere obtained for an 8-inch stop_
valve bellIova using an IK 650 ditgital1 comuter. . T.e analysis -$5 quite genieftal
and ma be applied to any welded bellowsa of similar- gecety. b.c ete
solution for additional loading conditions, sm-ch as axial and pressure- loads,
may be obtained by su~perposition. Matrix algebra was used, so the resulting
equations are in conver~izt form for soluttion on a igt pute.Hg
stresses were found at the pint of &a 11 applied radial displacemtent at
the lairer end of an 8-inch stop-valve bellows which is free at its uzpe
end. For this case, the solution became unstable after 3 convolution pairs
from the lower end. This instability is p;robably due :to: (a) nziial
errors arising from taking sall ditferences of large numeta an4/or to)
application of the solution to a bellows configuration which exceeds the limi-
tations of the theory. TY resolve this problem, further study was
rec end d.

130.
Estrin,. X. I., -Onae Method for the Solution of a Homgneoas Problem
for a Symmetrically Loaded Toroidal Shell", Priki. aet. Mekh., 11, pp 019-622,
Septeber/Octcber, 1953 (in Russian).
This paper gives a minor variation of Noyozhilov' a asymptotic solution for
toroidal shells.

Exline, P. G., "Pressure Responsive Elements", ASIX Trans.. 8, pp 605-

632, Noveber, 1938.

Pressure-responsive elements coinonly used in indicating and recording instru-

I ments are mnmeters, free-piston gages,. Bourdon tubes, diapbragas, and1

bellows. The mthemtics of the primry elements, manometers, a free-
piston gapes is simple and direct. Analyses of Bourdon'txvbes and diapbragms
show that predictions of their performance curves are tot amnable -to precise
caloulation even under the simplest conditions, which conditions do not
generally obtain In instruments. Generally,. the instrument maker must rely
W~on empirical knowledge, coupled vith approximate calculatiods, for design
data. Linkage adjustments must be provided to correct for variations in
dimensions and mterials normally encountered in manufacture. 'Eh~is. is
placed on the usefulness of bellows as pressure elements,. and approximte
formulas for calculating their behavior are given. The reselts of' considterable
experimental wceik on the characteristics of bellows of various sites. and
materials show that in many cases they are ideal elements for pressure-
responsive instruments. An ex*Vle of such use in a special-duty pressure
gage is given.

314
 132-
Psmili, J.., and Archer, R. RB., "Finite Asymtric Defaistion of SkwIfov'
Spherical Sbla. AIMA J., W (3, pp 5o6-5lon, nerdh 1965.
A procedure is'developed for the integ tion of the systent of nonlinear
partial differential equations governing the asymmetric deformtion of
shallow spherical shel. A suitable iteration wbams baned on a finite..
* difference approach in shown to yield the amssastic post-ucklizg sti~ii for
the spherical cap under uniforst pressure. Thus, the praetea1y~ iipbtan
asymtric "lower buckling loase is mrutuiticaL~y cosated for the first
time. igplications to more general asymtric buckfing ' obless -re i-
cateci by the exaple of the spherical cap with the load diftributed uniforml
over half of the surface.

Farrar, J. F. P., "Davelcpmnt of Metal Belows In Oerany Duiring War Years

Through 191.6", F-no Fimi Rept. 1076 (Ps 78651), March 194T7.
Ashortreport is given. of the developmnts in the mnufacture of austal
belowsinGermny during the wrthrvjgh 1916; b abelwofgoodqult
were manuftectured, but waah more slowly than is comaon In the United States;
homuver, no developments in the industry purportedly took plme during the
war ant~ production methods as comared to those in the United States is
antiquated.

I
. 1311.
Farrell, X. *T., "h Developmnt of Advanced Cryogenic Presmr 8witches
for Ballistic Missiles", Prebank Co., June 1960 (AD 2413423).
A low-press rewitch progr has resulted in the dee~tof an advancied
crycmni pressur witch which functions with two diaphrew~ essentially
the nas& the seix pressue switch described in Voume 1 of thin report.
The lw-pressure switch has Ieontrtd tham it is capable of vithatuading
the severe e -- ~ts conditions encountered In ballistic missile usage
by passing the Qualfiction and Flight Rating Testing which included coa-
bined environmental testing.

Favor, R. J.,. at al., "Lmn~stlgation of Patigu ehhavior of Certain Alloys

in the fieratawe Reapg Ro= TsMeature to -1.23 F-, Battelle0 Mwial
institute, Tmne 1961 (WD 266343).
The fatigue behavior of certain alloys has been investipmted In the tampera-
tore rag room Iteapeisture to -1.23 F. Thsllcqs evaluated were materials
curretly use for comonents in cryogenic amssile systems. The results of
an initial literature mwmrh are presented ga iclyas 8-N curves and
fatiue strength-tempweatur cross plots. In the azperimiental proarmp
equimnt vs eglasa VO test mll1 sheet specimens in fully reversed
banding; constant jamiau deflection experizients at tempratures down to
-1.23?, Detilid deuiriptioss of the equipment and specimens are presented.
Fatigue data obtained on 11. alloys at ran temperature, -11lOPp -320 F, and

35
 -423 F are presented graphically. The metallurgical histories and 4hemicaL"
analyses are described.

136.
F'Feely, F. J., Jr., and Goryl, W. M., "Stress Studies on Piping kpansion
Bellows"!, J. Appl. Mfech., 17, pp 135-141, 1950. "

As a result oZ numcrous and costly failures of the stainless steel bellows

used to take up thermal expansion in the piping of petroles= and chemical
processing equipment, a bases hae been developed for designing these bellows
to operate within reasonable stresses. A formula has been deri7ed to show
the total stress induced in the material as a result of the combined effects
of pressure and movement. The validity of the approximations used in this
formula has been verified by laboratory strain-gage measurements on an ex-
perimental bellows. A relationship between several variables in the design
of the disks has also been determined and serves as a basis for dimension-
Iug them to achieve the most economical proportions. When this relatioL-Ohip
is satisfied, the maximum permissible movement per disk can be obtained. This
is shown graphically. Suveral supplementary formulas are givev for deter-
miningbellows characteristics which may affect the design of adjacent piping.
The problem of designing an expansion joint for high-pressure service is
discussed briefly, and so: preliminary laboratory data are presented on
a commercial joint of this type.

13 7-
Filipov, A. P., and Bulgakov, V. N., "Application of Quick-Response Coam-
puters zo the Investigation of the Strength of Plates and Shells", Prikl.
(2), pp 125-134
!ekh., 'a6l (in Russian).
This paper describes the ry a computer may be used to solve problems
in plates and shells. As examples, a toroidal shell is solved by finite
differences, and a curved plate is solved by energy methods and the integral-
equation approach.

138. Finnie, I., "An XcpFrimental Study of Multiaxial Creep in Tubes",

1963 Joint
Intern. Conf. on Creep, Sect. 2, New York, pp 21-26, August, 1963.

The different methods availabli for the prediction of creep rates under
multiaxial stress are discussed and it is pointed ouat that the best experi-
mental studies in the literature lead to apprently contradictory conclusions.
Tests on tubes of aluminum and lead were made under torsion, internal pressure
ane ?nsion in an attempt to resolve this difficulty, but none of the methods
of' tcina prediction in the literature adequately explained the resulta.
This discrepancy is attributed to the effect of hydrostatic stress which has
hitherto been assumed negligible in creep strain predictions. It is sug-
gested, based on the known mechanisms of plastic deformation in metals,
that the equations used for strain prediction may have to be modified when
the temperature exceeds about half the melting point in degrees absolute.
O this basis many of the apparent contradictions in the literature may be
rasoled.
36

. . .. .. ; . ... . . . . . . . .. . . . . ... . . . . . . . . . : .. . . . . . ; . .. .. .
139.
139-Fitzgibbon, D. P., "Experimental Masurements of the StifTnesR of a Free-:
surized Bellows Systemw, Space Technology Laboratories Rept. No. U(.8402,-
October 1, 1958.
The stiffness of a pressurized convoluted bellows is~dependent not dny on
the structural stiffness but also on the pressure stikfness of-the bellows.
The pressure stiffness arises because the resuliants of the pressure forces
acting on .the deflected bellows produce a moment which is proportionel to
deflection. It was recognized that these two contributions to the stiffness
of a bellows could be balanced against each other to reduce the magnitude
of the mome t required to deflect the bellows. Subsequently, a theor tical
analysis was :ade in which-the stiffness of a bellows system was found as a
function of applied pressure and pivot point location. This analysis (Seide,
P., "The Effect of Pressure on the Bending Characteristics of an Actuator
System", ASMI Trans., 2, 1960, AD 605937) predicts that points of zero
stiffness can be obtained over a range of pressures for various pivot point
locations. The investigation reported herein is the experimental evaluation
of that theory.

14o.
Fleming, L., "Research on Transdf-Cers for Extreme Environmental Temperatures",
Bel]. & Howell Research Center, Pasadena, Calif.

The problem of utilizing a high-texperature deflection sensing element in a

pressure or acceleration transducer is chiefly one of finding and fabricating
materials for springs and diaphragms which have good elastic properties at
elevated temperatures. Studies made at Gulton Industries have concluded that
no metal is available which has adequately low elastic hysteresis and creep
above i000 F. Structural nomnetallics such as beryllia and high-alumina
ceramics show, however, considerable promise.

14i.
Flindt, C. B., "Theory of Restrained Corrugated Diaphragms", Engineer (London)
202 (5246) pp 193-195, August 1956.
The corrugated metal diaphragm has been used for many years as the basis of
measuring instruments, in which the force due to fluid pressure is balanced
by stresses set up in the metal as it deflects. Yany theoretical studies
have been published which seek to explain the behaviour of such diapbragms,
birt only those dealing Vith linear deflections have been very successful. In
this article the theory of linear diaphragms, originally due to J. A.
Haringx, is developed with the purpose of calculating the thrust which can
be exerted by a restrained diaphram under uniform pressure conditions.

142.
FlUgge, W., and Steele, C. R., "Toroidal Shells With Nonsyitric Lcading",
October 1, 1959 (AD 229108).

The problem of a general shell of revolution, without geometric discontinuities,

under slowly varying edge loading has been solved for dome-shaped shells and

37
 f'or nonshallow ahells but not for the toroid--- digtiwatiye becwAs t q--thq
interaction bet~reen *bending* avid *membrane" effecte.- Tbe .$1h-,rd&2 -Par-
tial differential equations for a shell of revolution reduqe for a msqiWO4-3
edge loading to an eighth-order ordinary differential system - three siw1-
taneous equations. These proved unwieldy for the toroid and so werqe adned
in this work to one nonhonogeneous integral-differential equation uVbidh MUn
be further modified to a form quite similar, except for the -Integ~l, ..to a
fourth-order formulation of the axiaU~y symtric prdblaa treated ?eriQU52dy.
With this formulation, the complete solution for A toroid with ede loaing
of a given harmonic requires the numerical solution of rtvo sia~ae equations,
equivalent to equations of the first order, a considerable anounmt of straight-
forward numerical integration, and, for the most general boundary c-onditions.1,
1~ the solut ion of 8 simultaneous algebraic equations. Althorzh t'his coul& be

t
tedious by hand,. particularly for several harmonics,. it wouldbe routine for
a digital comuter.

Forrst. P. G., Fatigue of Metals, Pergsmon Press, Long Island City, 1962.

About half the book is devoted to the fatigue strength of metals and the
influence of such factors as metallurgical structure, stress concentrations
sureace treatment, corrosion, and temperature. In addition, fatigue testing
techniques and the fatigue strengths of joints, components, and structures
are treated in some detail. There are also sections on-engineering
design to prevent fatigue failure.

1"
Frederick,, C.* 0., -Model Correlations for Investigating Creep Deformation
and Stress Relaxation in Structures%, J. Mech. Eng. Sci., 7 (1) PP 57-66s
1965.
This paper establishes a range of. possible creep, model correlations based
on uniaxial constant-stress-creep data and variable stress creep laws. It
is shown that the model behaves as an analogue which can be used to solve
the differential equations governing the deformation of the prototype. The
most general correlations hold for time-1&-dening materials where creep data
can be fitted by a stress index. The least general ccrrlations hold for
strsin-hardening materials whose data can only be fitted to a function of
stress. Elasic strains and variable loads are included in the analysis.

14-Galletly, G. D.,. "A Comarison of Methods for Analyzing Bending Effects in

Toroidal Shells", ASSE Trans., 80, PP 413-4114, September 1958.
An accurate estimate of the effects of bending loads acting on toroidal
shells is, at the r-esent time, a tim-consining problem for the designer.
Be is, in consequence, interested in the accuracy with which the faster
approximate methods can furnish him with the information he requires. The

38
 author had occasion to utilize two approzinate mantods In t1e30
ysis of a toroidal shell of -'evolut ion as well ais an: exaict, intho&.- iiie
purpose of this note is to present the results of these aM3 1 yses Iar ka
they pertain to the calculation of the edge diaplacmments, and rotationi.'

L Gslletly:.GD.. "Edge Influence Coefficients for Torbida3.

Shells of Positive
Gaussian Ourvature' t , AM4 Trans., 2Z (1), pp 6o-68, fe-bruary 1960.'

!Nb3es are given for the edge deformations of constant-thickness toroidal

shells subject to uniform pressure and edge bending loads. Over 100 different
sh-011 geometries were investigated and the results are Presented in dimension-
less form. Possessiou of these coefficients, which -w'ere obtained on a
digital comsputer, means that a rapid-and accurate formulation of the conxpati-
bility equations at toroidal shell-junctions is possible.

147.
Galletly, G. D.., "Edge Influence Coefficients for Toroidal Shells of Negative
Gaussian Curvature" '(J. Eng. Ind." ASIE Trans.. 82 (1) pp 69-T5, ftbruary 196.
Continuing the work presented previously, the Present paper gives additional
I .- tables foor the edge deformations of constant-thickness toroidal shells sub-
ject to edge bending loads and uniform pressure. The two papers tw~bher th'as
cover a wide variety of toroidal-shell geomtries and enable a designer to
calculate in a simple matner the edge mocrnts and shears at toroidal-shell
Jiunct ions.

148. llet., G. D, -On Particular Integrals for Toroidal Shells Subjected to

Uniform Internal Pressure", JAME Trans.. 2, pp 4i12-413, Septeber 1958.
For toroidal. shells which do not include the apex 0= 0, two comon approxi-
mations are used to evaluate a particular integral of the relevant differea-
tial, equations. These approximtions, which are knovn to vio.~xte the coupa-
tibility relations and ame abosen to satisfy only the equillius conditions,
are: (a) the trsnsve me shear Q4and the rotation V are both set equal to
zero, and the stress resultants Nand No are given by their memrne values.
(b) Sane as (a),. except that the rotation V is given by the memrane rotation
Vm. The puarpoae of the present note is to point out that these approxima-
tions are not very good even iihen O is as large as 30 degrees. To illustrate
the point, a canstant-thickness toroidal shell under uniform internal Pressure
is inveigi.ted.

I. Galletly, G. D., and Radak, J.R.K., "On the Accuracy of Som Shenl Solu-
tions",. AMD( Trans.,- 81 p 577-583.. December 1959.

R. A. Dingle's method for finding asymptotic solutions of ordinary differ-

entia. equations of a type such as occur in the bending theory of thin shels

39
 -
C

of revolution is pres~ented in ontlize. This mthod leadfs to the 0~ autpi asI

R, 3. Langer's method usadL for pi~ble&* or this kind, Wa -pemxitsi iLW.4l Z
ytical and less formal Interpretation of the wumptoti;C tr-eatment- of .,UW etioms.
A comparison is given of edge in~fluence coefficients due to-beading len -b-
tained by use of these asymptotic solutions and w.=Arcal integration of te
equilibriuma equatios, respectively. the parti--ular shells invvetigated ire ot
the open-crown, ellipsoidal,, and negativez.curvatunre toroidal types. The results
indicate that the agreement between1 these solutions --!s satisfactory. In the-
presence of uniform pressure, the use of the mmbrane solutions far tht deter-
mination of'the particular ilategrals appears to lead to accetablt rmsults.-in Itbe
case of ellipsoidal shells. Rowever, in the case of toraida3. ebtllla the 4m.
kwane and the numerical solultione disa;Wee significantly.

.5-Gartner, Do. Krexey,, Ho, t:ew Procesz for Production of Single-W&II

Meta~l Be V-4
by Chrcmizing and Their Tasting", Techniky 117, PP 72-73.. tebruary 1962 (in IGermfin)
Farmed bellows are used as packIng elements. The special advantagRes of this
type of sealing are complete tightness, absence of maintenance aod of gland
friction. The disadvantages of single-wefl. formed -*elownmuseIl as otealing
elements are their low life and the limits of permiuisible work pressure. Thei
usual materials for metal-formed bellows are tcmibac.. yellow braza., st;Ufss
steel (18/8-CrNi), etc... In the new method of the VEB M4easuing Inatru1.-nt
and Armature Plant "Karl Marx" of M~agdeburg.. the formed bellows are of ferrite
steel developed by chromium p'.&ting steel 5 .Ti 5.

General Inamics/Astronautics Test Rept., "Design, Proof Teat Report for Bel~lows
Assembly - Life Chill-down Manifold", June 20, 1963. IM~P :Lc5..69.c0-D-l.
Tests of two bellows assemblies included proof pressure (15 psig for 5 min. with
Wt), vibration (at -300 F) and life cycling (10 rsig ~- 500 cycles, displaced 20).

152-
General lymamics/Astronautics, Test Rept.,, "Rose Aesesaly--ie-ruier., Liquid
(Ocygen SUPPlyv, MA-3",. January 18, 1962 (AD 2883"0.
Tests of 15 flexible-metal hose specimns Incuded init*Il performance test5
(75 50
ad pig 20), vibration,. acceleration, 3ife cycling (1 2,0:D3 cycles),,
and buret (30 psig).

General Dyrnamic s/Astronaitic s Test Rept., "Repart of Preproduction TEets of

Bellows Assembly", November 25,, 3959. nU~ 115.l0.6o.o0-D5_0l.
Tests of two concentric bellows included leakage (at 18 paIA), proof pressure (412
psig),, deflection cycling (1000 cycles at -300 P),' and burst pressure (70 psig).

404
 Aals/A~steugt~cs Test Rept., o o. ~a~u31W~~
B~id-Zagne Coutrdl", Oetbe 10, 19062 (AD 29660).
Tests of a braid-coerad~ 3-inch-diumeter bellows included visual ekiai~n
presaiwc deflectim2 (250 psig_; 100 in-Ib), life cycling (200 yls,'of
pressure (000 psig L12) ad biurst (1410 psig).

Genera I Dsmicg/Astronautice Tat Rept.,. "Acceptanace Test of Fittings, 0

100 p)siZ1 .9Aue 22, !961 (AD 276304).
A .ceptance teatc are described of a flexible elbows flexibli line, special line,
transfer line and elbow. The tests ihcluded visual exasmtion, proof pres-
se (150 psi , an6i two tnypes of leak teets.

!keberich, W. W., Yartini, 0. F-,, and Zsckeqy, V. P., "Serrated Stress-Curves ot

Katas-tsble Austmiite In Alloy Steels", As. Soc. for Met~als Trans... 58 () p

The macroscopic yleld&behbaviar of mtastable sustenite In nine aloy steels Vag

stued, It uts found that the degree of austenite decoiupootioa was diretly
propor±'ial to the~ ac~mt of strain.

G.esceeln, 3., "Ho Mtal Be~lovs are Fabricated by the Cook Electric Co."
Automotive and Aviation Industries, Z, pp 36-38, merwary 15, 19 4&
5.

Mnuaturing facilities are extremely flexible, providing with few exceptions

a jcb-shop type of operation cvable of kan&Ling a vast variety of parts and as-
m!eChlies in any quantity frua am to hundreds of thousands. Assembly processes
emody ;ractically every known mettod of welding and brazing or soldering.

128.
Gibbs, D. F., "Dpring X~±pbas"O 3. Sci. Instr., , p34-35,, January 195T.

A brief description- is Siven of &ipras convoluted in variouis ways.

Msewl, A., "?laptkc Deforton ,of a Circular Djapbraga under Pressure',

3. A01l. Wech.. lip pp 28W-296, 1948.
In this reprxt a numrical solution It given Of a set of ejuations consisting
essenltiFrlly of tbree plasticity laws, two strain-diSPlacMent laws, Wd two equi-
libriia lairs 'which describe the action of a clumed.. thin, circular diaPhr&W
 as it yields plastically when pressure is applied to one Side. The stres3es, ..
strains, thickness rariatlorn, and deflections for ary thin, circular diaPhrama_.
of a given material may be computed.by the nwmeical ifitegration Of thb et.tlib-
riuiu conditions, the geometric conditions ielating dgsplieuednts in&' trs~nk,
and the stress-strain laws. The solution maybe reduced to th- solution of a
second-order differential equation with the radial distanae r as iidepenhiixt
variable. The solution depends upon an expermentally deteraiied "fuIctioh T(y),
which describes the stress-strain properties of the mate rial, and upon three
parameters, the presure p, the original thickness ho, and radius a of the clamp-
ing ring. It is found that for a given material, a faml]y of curves with pa/ho
as a parameter serves to predict the solution for any thin circular diaphragm of
'he same material. This analysis has been carried out for a particular function
i(y) based upon results of tensile tests made on a specimeni of medium steel.
Graphs of theoretically and experimentally determined values of deflection,
radial and circumferential strains, radial and circumferential stresses, and
thickness corresponding to various pressures are presented which apply to all
diaphragms made of the same steel as this specimen.

16o.
Gohn, G. R., "Fatigue of Metals: Part 1, The Mechanism of Fatigue", Mater.
I Res. and Stands., (2), pp 106-115, February 1963.
This is the first of three companion articles on all facets Of fatigue. Part 1
reviews all of the theories of fatigue crack initiation accompanied by 44 ref-
erences. Early fatigue failures are examined. This vast background of effort
points to the fatigue model of Wood as the current, most plausible discuo3sion of
the mechanist of fatigue. This model recognizes the microstructure-and assuw1s
that the dislocation theory is valid. Attention is confined to small plastic
strains ±e alternating about a zero mean stress. This type of fatigue need not
result in strain hardening below a certain strain amplitude, called "reversible
plastic set". The deformation process changes (under cycling) from coarse to
fine slip in the crystal as the strain amplitude becomes less than the reversible
plastic set; anl fatigue deformation accompanies fine slip bands or course slip
bands consisting of an avalanche of fine slip bands. This explains how fatigue
develops at some intensified slip band; it accounts for the formation of notch-
like contours at certain bands; it accounts for the tendency of the slip to
concentrate in the bands first formed, and the numerous observations that show
fatigue usually deteriorates the surface.

161. Golaerg, J. E., Bogdanoff, J. L., "Static and Dynamic Analysis of Nonuniform

Conical Shells Under Sym:P.trical and Unsymmetrical Conditions", Ballistic Missile

and Aeros,ace Technolo&. Academic Press, New York, pp 19-238, August 1961.

The equations for determining stresses and displacements in synitrically and

unsymetrically loaded thin conical shells are presented in a farm which is es-
pecially convenient for numerical integration on an electronic digital compter.
The usual assumptions of classical shell theory are emloyed. Thickness and
mechanical properties of the shells may vary along the generators and temperature
distributions which do not produce significant circumferential variation of
mechanical properties may be banaled. Results dbtained for several Problems
are dispayed.
X of symmetrical and unsymtrAital vibratin

I ______ ____ ____ ___ ___ __42

162.
Goldberg, M. A., and Pifko, A. B., "Large Deflection Analysis of Unifonldy
Loaded Annular Yeuiranes", AM J., 1 (9), py 2111-2115, Septemer 1963.
An Iterative technique is employed to obtain 'aroxi te solutions-to Fool's
nonlinear membrane equations. Four 3iiftriy loaded rotationally syimtric
membranes are examined: (a) an annulus fixed at both edges, (b) an annulus
fixed at the outer edge with a rigid plug in the interior, (c) an annulus
fixed at the outer edge and free of tractions at the inner edge., and (4) a
solid cirbular membrane fixed at the outer edge. Jtxerical results are pre-
sented for each problem. The results of case (d) are cmared with the
existing exact power-series solution presented by Hencky. Stresses, and de-
flections computed by the iterative technique for case (d) are within 1.14
percent of those predicted by Hencky.

Grafton, P. E., and Strw, D. R., "Analysis of Axisymetrical Shells by

the Direct Stiffness Method", AIAA J., 1 (10), p 2342, October 1963.
A method for .the structural analysis of shells of revolution, ccaowsed of
materials with orthtropic properties, is discussed. The development is
based on the direct-stiffness method. A truncated cone element is introduced
to take advantage of symetry. Derivations of the stiffness and stress
matrices for the truncated cone element are given. Several examples are
solved on thi digital comuting machine using a program that is based on the
truncated cone element. The results are cmp&red to other theoretical re-
sults, and the correlation is excellent. Extension of the technique to
handle linear unsymatric deformation and nonlinear symmetric deformation
is discussed.

164.
Greenbau, G. A., "Coments on 'Numerical Analysis of Unsymtrical Bending
of Shells of Revolution'", AIM J., 2 (3). PP 590-592, iarWh 196&.
This note cmnts on the paper by PAdkms&y, et al., on the finite differ-
ence solution of shells of revolution. Te note gives conditions appropriate
for shells closed at the apex. Tbs permitt the solution to be obtained at
the axis without the necessity of leaving a Mll bole on the axis.

.Griner, J. C., and Miller, H. C., "The Cogtibility

of Naterlials With
Chlo ine Trifluoride, Perchloryi-Flucrids, ad Mixtures of These', Pennsalt
Chemicals Corp., April 1961.
Com&ittbility and corrosion rates of allcfs of aluminm, copper, magnesium,
nickel, titaniun, steel and stainless steel. and columbium: molybdernia carbon,

43
 graphite and tluorocarbon plastics in cblorine trifluoride,. perebloryl fluoide
and mixtures of these at 30 C were investigated. Titaniun, columiinm, so3yb-
denum, carbon and graphite were rapidly attacked in C113 . Corrosion rates
of others were extremely low in ali liquids. In the vapors, instances of
higher corrosion rates were noted. Teflon and Kel-F adsorbed moderate aowitu
of 01F3 and C10 3 F. Passivationby C1F 3 was unnecessary for reduacing corrouion
of properly cleaned metals. Corrosion in wet 01031 was characterized by
localized attack, but somse stainless steels were resistant. Miantua exhibited
increasing impact ignition in liquid C1.03 F beginning at 19 ft-lb,- b,-ut~even
at i140 ft-lb burning was not sustained. No other metals showed impact iLg-
nition in 0113 or 01031. In explosive shock tests, C1031 gave a stroe in-
teraction with the metals tested than did 0113 and alusin~m shoved a greater
interaction with the fluorine chemicals than-low carbon or stinless steel.
Greatest enhancement of explosive shock occurred with titanium and ClO3 .n
explosive denting and perforation of steel and aluminum cylinders containing
0113,* C10 3 F and their mixtures, no enhancement occurred. A high order ex-
plosive interaction occurred between C10O F and titanium cylinders perforated
by a shaped explosive charge.

166.
Grigoryev,. S. V.,, -Some Problems in Flanging and Beading Jembranes", NASA
Tech. Transl. F-32, 13 pp, June 1960.
The hydraulic method of beading may be used equally well for memraeas of
cold-worked -materials and for mterials which have uniergone therl treat-
ment, to give them high yield strength. The method of flanging and beading
mesibranes which is under consideration gives considerably more stable
properties and less residual deformations, elastic fatigue,, and hysteresis

Grinstad, C. E., Frawley,, R. N., et a!... IArA IMprovd I-ndieator fmr 3eastwing
Static and Dynamic Prese. SAM J. (Trms.),. L2 (11) pp 534-~555, 944.1
The princ iple of operation and the mec'.ianicia design of an Improved indicator
for measuring static and dynmic pressures are discuseed im thir. pape. The
condenser type of indicator was seleeted by tlhe- autbors for engine work be-
cause it is compact and 2trdy, it is easily serviced, it has a hihntua
frequency, and it is relativelyr Insensitive toi shock and vibratlon. This
type of indicator also does not reqre~a zwecanical linksce betweeni the
pressure diaphragm and the electrically sensitive element.

168.
Groirer, H. J., Gordimn, S. A., and lacksomi, L. R., ",fatigue of W~tale aid
Structures", U.S. Govt. Printing Office,. 1954 (currently beinig revised).

This bck is directed particular~r to the dtesigner or engineer with scme

knowledge of, but limited practical e nperience, in fatigue problems. It
is intended to provide a sunry of present-day inforaation and to provide

44~
, references to published literature for more complete information on the-
most important items in such design. It is recognized that in a nuber of
instances present-day knovledge is incomplete, hence the information
presented is considered as a guide and a starting point toward a&Ieutte
solution of specific problema.

Grover, H. J., and Bell, J. C., "Some Evaluations of Stresses in Aneroid

Capsules", Proc. of Soc. Exptl. Stress Anal., 5, pp 125-131, 1 9 48.

Two lines of attack were followed: first, experimental evidence of sur-

face stress distribution was obtained by the use of brittle-lacquer coatings
on several capsules subjected to various loading conditions; second. a smthe-
matical computation of stresses in a model, quite similar to comon aneroid
capsules was carried out to give nuerical results. The primry purpose of
the investigation was to study drift an(! mechanical hysteresis in aneroid
2apsules and to develop methods of producing capaules with less drift and
hysteresis. This involved the consideration of many factors, including the
choice of material for diaphragms; optimum heat treatent, aging, and cold
working of diaphragm materials; fabrication of diaphragms; methods of joining
diaphragms to produce capsules; and effects of variation in design of capsules.
Previously, attention had been mainly directed to the conventiol corrugated
diaphragm type of capsule and to commonly used materials su~h as beryllium
copper and phosphor bronze.

170.
Gutterman, E. P., "Diaphragms and Linkages for Pressur,.-Operated Flight
Instruments", Office of Tech. Services, P 14858, 225 pp, April 1956.

Pressure-sensitive, circular convoluted diaphragms experience severe bi-

axial stress concentrations during deflection. Resulting overstrains near
the periphery and in the convolutions contribute to lysteresis and drift and
limit the usable sensitive range. Scae of the conclusions reached during
the investigation of these configurations were: (1) a noncircular, welded,
radially-convoluted, free-edge diaphragn shape can be asde to reduce to zero
all strains lying in the capsule surface, within the first order of approxi-
mations - only relatively evenly distributed bending strains remain; (2)
free-edge diaphragm-forming dies may be shaped to satisfy the design equations
by a simple, automatic coeputing linkage attached to a standard shaper, (3)
age-hardening alloys are superior for diaphragm manufacture - raw material
should be annealed dead soft before farming - since large grain-aize is unde-
sirable, a high-temperature, short-time hardening cycle should be employed;
and (4) hysteresis and drift of assembled capsules may be decreased and
usable range increased by artificial aging incluling the successive steps
of deflection cycling, temperature cycling at maximm stress and moderately
extended heating slightly above maximus service temperature, und maxirau
stress.

45
171.
Balford, Go R., and Morrow, JT., "Lov Cycle Fatigue in Torsion", Mniv. of
f3linoie, October 1961.
Completel.y reversed torsional-fatigue failures are roted for two alunam
alloys.. 60-40 brass and SAE 1434.0 steel in the life range of 1/2 to 20'00
cycles. Cyclic hysteresis loops were measured and the total plastic-strain
energy to cause fracture is reported. The results are interpreted usin
hysteresis energy as a criterion for fatigue dazaege and also in terns of .the
Coffin equation for low cycle fatigue. Both methods of interpretation give
good agreemnent with the test results. For practical purposes, the data are -
adequately described by substituting shear strains for nornal strains In the
Coffin equation.

172.
Hamilton. P., "The Development of Electroforuing Techniques",, MoDonnel1
Aircraft Co., march 10, 1964 (AD 43518)-
A test program was initiated to further develop electrofoming techniques
investigated under Test Request 513-277. Mandrel mterials selected for
evaluatign were polystyrene, plexiglass (lucite), cerrotru, polyurethane,
and redwood. Nickel was electrodeposited froat a sulfamte electroplating
bath on the prepared mandrels to the desired thickness. Cmleted electro-
formed parts were separated frcm their mendrels utiliing appropriate tech-
niques. Parts successfully eleotroformed Included a wave guide, a bellows; -

and a pitot -tube. A sandwich structure attempted was not completed. A studyr
form of redwood was designed and electroforned to evaluate the deposition
of nickel on inside and outside corners as well as in blind holes. The in-
ternal stress produced by the nickel sulfamte bath was deterzined with
the aid of the Brenner-Senderoft Contractoneter.

M-aringx, J. A., "Design of Corrugated Diaybragms", A&M Trans., 7Z, PP 55-61,

61-62, 62-&14, January 1957.
Three previous papers by the author aet forth methods of calculating the
rigidity of corrugated diapbr-ua, the streases in the sheet mterial, and
the nonl3.nearity of-thbe relatioa be-tween lead and deflection. As a further
step,. the introduction of a few simplifying restrictions having no fundinntal
effect on the problem leads to the concept of a chart giving at once the
dimensions a diapnrsga must have so as to conforn to specific requirmets.
An example is included by way of illustration.

1T4 .
Baringx: J. A., -Instability of Bellows Stabected to Internal Pressure"pPbflips
Research Repta, (3, PP 189-196;, June 1952.
W

Like thin-walled cylinders, dealt with in a previous poaper, bellows also my

becm unstable whan loaded by interial pressure. The critical value of
this pressure,, which in accordance with Euler' a well-known formula, is governed

46
 by the rigidity of the bellows with rtaPeCt to bending, has been cbeutid
only for rectangularly shiped corrugatimas, and has been checked experimntally.

IWig,
*
J. A., "Nonlinearity of Corrugated Diaphragms", AppI. &fl. Rea~
sect. (a), 6, pp 45-52, 1956.

be expected for lredfrain.D e&so nexpk tI h-.ta

the introduction of the corrugations into the flat plate, though unavoidably
increasing the italrigidity, involves an important gain in maxlnu de-
flection for the same degree of nonlinearity.

17. aringx, J. A., -The Rigidity of Corrugated Diaphragms", Appi. Sci. Rea.,
seat. (a), 2, pp 299-325, 1950.
When the corrugated diaphragm is replaced by a fictitious flat plate ofI
similar properties it is possible to derive a linear differentia equation1 for
the deflect ion. The coefficients of this equation, hmeverar in aco-
plicated way and its solution for the pressure-loaded diaphragm is only
given for thick end for thin sheets separately. For thick sheets the profile
of the corrugation appears to be !nessential, whereas for thin sheets it is
necessary to distinguish between trapezoidal, triangular and arc-shaped
corrugations. Ply an obvious device the results for thick and for thin sheets
are fitted together, so that the deflection can also be determined for the
intermediate range of mediiu sheet thickness. The final results of the pres-
ent calculation are compared with aessients carried out by others and are
found to be in satisfactory asreentent with the experiments. It is to be
remarked that, ccoqared on the basis of smal deflections, the introdufltion
of corrugations Into the sheet leads to a considerable increase of rigidity
of the diaphragm. Teprevailing assertion thst the flat plate is more
rigid then the acrrugated diaphragm holds only for large deflection., because
of the nonlinearity between the, load and the deflection of the flat plate.

2Lliwthorie, R. C., "Flowr in Corrugated Hose", Prod. Bag., pp 98-100, June 10,
1963.
An analytical method for calculating pressure losses is given which assuses
thiat the corrugations behave as a series of uniforenly spaced orifices and that
pressure drop is caused by a succession of individual flow expansions. Teet
results are given.

4
_____________ ___47
 Ew~thornwaitep R. M.,. -gDefaeatioo -of Platen iLn the 3~astica!4te3je .
Proc. of the Second .U.S. Nati.. Congress -for Appl- YAeWrij A8UjppW.52l,526j
1955.
The deflections of plates with circular symtry are computed for an elatic-
-plastic material. that obeys the yield condition of Tresca and the asst4&d
flor rule. At any Point- on the -plate the entfre.-thickriess_ is a2a=OdtoJ
be either fully elastic or fully plastic. Azi .4uaz'~ plat351!aptd
at the outer radius and clamped to a centrally loaded rigid disc is ahalyzed
in detail, the simply supported- circular -plate with a.-central coddentritbd
load being included as- a limiting -case. The testing of a: vtsel.pI*te in-
descrIbed and the results obtained, are comared with -tbetbewy.-

2"
~High Precision Barometric Bellows", Engineering, p709, June 6, 1958.
Precision barometric bellovs can be produced in any of the conventional
materials used for instru:2ent and control system, i.e., phosphor bronze, -80.
20 brass, or aluminii= brass alloyrs; and-also in Moneli- cijwo-niceW Andl&-

.1 stainless steel. A tolerance in spring rate as low as *10 percent. cazi be

Provided, on a production basis.

Hill, -Re , Theory of -the Plastic Bulging of aMetal Dispbra.Wby, rateral

Pressure", Phil. Hag.,l, p 1133, 1950.

Explicit formulas are obtained for the stresses I-,metal diapbrao which
is :bulged Plastically by lateral Pressure.* The predicted f4nfluemc 'of work-
hardening on the shiape of the profile, and -on the relation -between~polar
strain and curvature, agrees veil with experiaentaL data., -A slIAp~ exp ression
is developed for .the instabiity -strain.

Rise, I- C., -Design, Developient, and .0erat ion of.Mst&l~Iaphra

-- Reactor-
Service Pumps", Oak Ridge hktl Lob., aKI 28111, Ny 10, 960
Aqueous homogeneous reactors require pumps capable of injecting relatively
small quantities of highl.y radioactive, corrosive fuel solution into the
high-presstire qystmv These pia~s uit hae.s a long ite ne-free lfe' be
absolutely leak-proofbe resistant to the radioactive and corrosive environ-
ment, and be easily replaced by remote methods uhen necessary. 91daulically
driven metl-diaphraWpW s -capabjle 'q- pactIng these reguraens ve
developed in tb*-.oenoa Roaetor..PrQjeact .at, OME.. ftis report. PUr-
ize s-the develapn-4 l<program, the.-oCper9Lting experience with the. pumps -in
tests and In reactor service, and the design informtion derive, from the
develomnt program.

J48_
 182.
Hoffman, J. H.., et al., "Diffusion Bonding Beryllium Copper for Ultrahigh-.
Strength Joints", Welding J., Welding Res. Suppl., 4i. pp 160s-166s, il;
1962.
Bond strengths averaging 108,400 psi were developed in joining precipitation-
hardened beryllium-copper by diffusion techniques. This is in contrast with
an average of 40,000 psi obtained by ordinary brazing techniquesi :The unique
" method of diffusion bonding studied in this progam involved the.use of
molten alloys of silver and gold to fill discontinuities at the.bond inter-
faces. These molten filler alloys were in turn, completely. or in soe
instances partially diffused into the base metals' resulting in sound and
continuous base-metal-to-base-metal bonds. Joints of this type produed
under high vacuum conditions, using a silver-copper-indium filler alloy,
resulted in the highest bond strengths developed on this program; however,
those effected in dry hydrogen atmospheres (-80 F dewpoint or better) using
the same filler alloy, average. 96,000 psi. Data were also developed on
the effects of heating beryllium-copper -at 1550 F and at 1475 F on grain
growth, elongation, reduction in area and elastic modulus to determine its
usefulness as a high-strength fabricated material for unusual applicatioiis.

183. /
Hopkins, H. G., and Prager, W., "The Load Carrying Capacities of Circular
Plates", J. Mech. Phys. Solids, 2, pp 1-13, 1953.
This paper is concerned with the load carrying capacities of circular
plates made of a perfectly plastic material that obeys the yield :onditions
of Tresca and the associated flow rule. Various conditions of rotationally
symmetric loading and support are discussed.

184.
18--.-Howard J. H. "Designing With Metal Bellows", Machine Design, 26 (1),

pp 3T-46, 1954.
This i3rticle summarizes principles of different applications of bellows,
the proper selection of bellows materials, determination of bellows propor-
tions, basic types of bellows assemblies, and assembly methods as they
influence bellows specification.

185.
Hu, W. C-L, "A Linearized Membrane Theory for Prestressed Shells of Revo-
lution", Thesis, Stanford University, 1964.

" The ma new applications of expandable or inflatable structures demand a

theoretical study to elucidate the mechanical behavior of an elastic shell
membrane beyond the limits of the linear shell theory. The present work is
directed toward the development of a theory to determine the effects of
Ssall, axisymmetric loads on an inflated shell of revolution, or, more
generally, on an elastic membrane, in the form of a surface of revolution,
which has been "prestressed" by a uniform inflating pressure and edge tractions.

49
 Hudson, G. E., "Theory of ve rl,-ot.. Def~mtion of a Thin Dia~ "
T. Appi. Phys., 22, pp 1-1 -IvtoA~ .

The theory presented in tt 3 5a1t :.aLe was devl pd in an attekt

the obeerved motion and. detormtio, of c1-m- e a iprsue
in certain underwater explosion experiments and "in cktain bechic 3pgs
The approach enables one to set up certaii eiusi of motin, vh~;ybe
solved in finite form under certain conditions. The' s-olu-ttons eale9
to specify, for instance; the f iral de:Fm-ied da_#pbra-8' trdii, thet. 4i6-
bution of thickness after def6rdation,. the isaiiitie Awhii ti* t'601
tize for deformation to take place, and min otlber qusntities.

Hunt, L. B.,. "TheRistory or Pressure-Responsive Z.z-ents", 3. Sci. Inst.,

21(3, pp 3T-42, March 1944.
Over one hundrel years ago Lucien Vidie, a laiyer,and aiteur scientist or
Nantes, carried out his experimental work on the measurement of atmospheric
pressure by means of a closed and evwcuated mtallic vessel; his first
patent was applied for in 1844. Much depends upon the indications of the three
types of pressure-responcive elements which have been developed frcu tbis
work - diaphragms, flexible bellows, and Bourdon tubes - and it is of come
interest,. therefore, to look back into the early history of this form of
pressure measurement. For reasons connected ;with the-elastic propeirties
and behavior of these elements, with special reference to the materials
best suited to their construction, the writer bad occasion to undertake a
careful search cf the literature. In the course of this search 4;here
emerged a number of points of historical interest which have not apparently
been recorded in sAy~ connected manner. The article outlines the general
sequence of events.

188.
Hurlich, A., "Properties of Materials at Liquid-7Mcy aind Likuid-*drogen
Tempeatures", General W~namics/Aitronautics, February 28-Narch 15, 195T.
A survey was made of the properties of materials at liquid oagen (-293 V)
and liquid hydrogen (-4#23 F) temperatures to assist in the-selection of
suitable materials for propellant tanks, bellova, and propellapt lines In
a setellite missile.

Rusen, C. F., and MaClel1an, R. G.., "Dvel6pent, and Fabrication of Omni-

Directional Accel~ercmeters".. Frd Motor -*.o;, Aeronutronic Division, July

A A description is given of a pressure transducer which Is mounted so that

the diapbrao Is at the center of the mrcury cavity. Solid-state #train

50
 S- WO P C.

elements are bt-.,ded to the back of the pressure-sensing diaiphragm and, are
wired in a til bridge circuit to convert tbe prealife-_induqed stresrs to
higbh.voltage output. R-I diaphragm has a.-,err bi- , hich
waual~
eliminates problems ass £ted1 with its use as th-6 iingsj -elemsiit ini t
accelercumter. A heavy integral flange@- rit intla iwth i~A
case distortion.

11 rdraulically Formed Seamless Metal bellows", ~Ashinery (London), :L4 (1893)

PP 131-136, February 3, 19!49.
In the process of mnufacturing seamless metal bellows, a thin.walled closed-
end tube is caused to flow coatiinumauly under an inteinal fluid pressure of
several hundred pounds per sq in. in a collapsible die. The latter consists
of a series of plates equal to the rnuber of convolutions required, spaced*
equidistantly surrounding the tube. The internal pressure causies-the mtal
to flow transversely between the plates as the tube collapses endwise to form
the bellows in one continuaous operation. No annealing is performed on thxe
* bellows after forming. A pert icul~r advantage of the process is that the
amount of col~d working applied to the material Is unjer close control. this
results in a very uniform grain structure and wall thickness, which is of
particular importance at the curved roots and crests of the convolutions,
where most of the load during flexing in service is carried, and where fail-
ure of a bellows invariably occurs.

191.
Irvine,. C. N., and Batrnett,, J. H., "State-of-the-Art Literature Siavey on
Fabrication Techniques of Advanced Ducting Comonents", NASA Tech. Memo.
X-531T3, December 2., 1961&.
The purpose of this state-of-the-art literaftu'e survey was to determine
and define the mnfacturing techniques required and the process difficulties
likely to be encountered in the fabrication of elbows, bellows, and ducting
assemblies for use in advanced dacting systems ranging from 2 Inches to 50
inches in diameter. T~he repc~rt briefly covers the methods currently used in
elbow, bellows, and dacting fabrication as well as acn alternate methods
available and new methods werdevelopmnt. Sme Information conctrning
the properties and fabricability of several 4linm, Iron-base,. ninkel-base,
and cobalt-base aflcqs suitable for ume In cryogenic ducting systems -is
also presented.. h survey indicates that advancements in the state of the
art of fabricating large-dimter elbows, bellows, and ducting will be nec-
essary for new, larger duating system applications.

192.
Jackson, Je D., Miller, P. D., Bqyd, W* K., and Fink, Fe Wo, "A Study of
the Titanixm-Liquid Oxygen Pryophoric Reaction". VAM TR 60-258, Meich, 1960.
A review of the literature indicates that titanimu is iqpat sensitive
under liquid c7sn below the acceptable limit for other metals. An experi-
mental program was begun to deterstine the mechanism of the titaniumLC

51
 1i actrolUsn;
reaction. Several factors were investigated. sing" iy
t
using unalloyed titanium (75A) and an alloy (6A1-V) thabti re cif
cleaned. The factors were: (I) ,Xposure of a t.esh suface by frac ".
and tearing; (2)deformation by impact using steel bans; (3) ac of oth
specially cleaned flat surfaces; (4)LaX pressuri. and velocity; ( ) calling.
The results from this program indiate that no one of the above, per sey 1s
a primary cause of this reaction. A proposed mechanLim in that heat gen-
erated by impact produces a gaseous omWyen, which is compressed at local sites.
A fresh surface exposed by the impact reacts with the high-pre&-Ruft P"95
oxygen. Propagation is dependent on the asunt of heat giierat e&
rate of heat loss from the affected area.

193.
Jackson J. D., "Corrosion in Cryogenic Liquids", Che2. Eng. lzog., (),
pp 61 . i. 1961.
The large use of liquid propellants in missile rystems has broghtm Orse
serious problems to the missile designer and the materials ener.
important problem is-the corrosion behavior of materials of construction
under the various exposure conditions of the missile and the auxiliary
equipment. In the missile, short-term .exposure occurs; however, the "aterials
may be stressed almost to their yield strength. In the auxillay etuiit
(such as storage tanks, pipelines, and pumps), long-term exposure, under Lw%.h
less severe strength requirements, occurs. This article discusses corrosion
behavior and mechanical properties of metals used in handling liquid oxygen
and liquid fluorine.

Jackson, J. D., and Boyd, W. K., "Compatbility of Pro- *-ants 113 end
114B2 With Aerospace Structural Nateriels", Battelle eorlal Institute
31C Memo 151, April 27, 1962.

Considerable interest has been generated recently in the use of two fluxri-ated
hydrocarbons (propellants 113 and 1l102) because of their moderate boiling
point ard low specific beat and heat of vaporimtion. These compounds are
being considered as propellants for vernier rockets which are used for
stab3.liz tion of missiles and spacecraft. For this reason, the corrosion
properties of these fluorinated hydrocarbons have been investigated with
metals comionly used in missile applications. Thi muorandm presents tbe
corrosion dats and sumarizes certain physical and chemical pro~rties of
propellants 113 and 11-4132.

Jackson, J. D., and B!yd, W. K., "Reactivity of Metals With Liquid and Gaseous
oxygen", Battelle Memorial Institute 1ac Memo 163, January 15, 1963.
Since the first cbservation of a violent reaction in early 1959, the c€psti-
bility of titanium and ils alloys with liquid oxygen (LO)has received

F 52
 considerable attent ion. Initially, lab oratory Investigations were prli~rlj
limited to impant studies utilizing the ARM Impact tester or modifications
thereof. Later the Air Force initiated a propme- to determine the usebsuise
of the reaction, The results of thesef early studies were- previojp~~k
marized in ThUG xemorandi= 89, dated March 6., 61.. YDae recently, 'the. gactors
necessary to promote reactions between titAniua izd, liquid- or -gsippi!-tqQ'gen
(GOX) beve been studied under conditionssiiallr to those 'which would.b* en-
inmsie
countere n pc ser-vice. Itis the puros- oft18O
to summarize the present state of the art in the light of both past and
.Present. developments.

Jackson, J. D., and Dyd., W. K., "The Compatibility of Iftterials In LEK--

Vehicle Tank With Nitrogen Tetroxide.During Vibrational Impact", Battelle
Memorial Institute,. NAS 9-3-100, January 1, 1964,. to present.
The LIM oxidizer tank is to be constructed of Ti-6A1-Iv, which my-be injAct
sensitive to X2 %! under vibrational impact associated with lunar landing.
liuact damage of the thin 4-m-11 sheet (stressed to 100,000 psi) was observed
at about 20 in.*-lb of force. No,reactivity between Ti-6A-&v and k>%~ was
found when impacted with either 20114-T6 aluninua or Teflon for 39,000 cycles
at about 6 in.-lb. No reactivity was observed in fatigue experiments whee
]104 arsh surface or chips were produaced. Fatigue failure did occujr after only
cycles at an impact level of about 11 in.-lb. Beat in the amo'.uf of 2
to B~ tu/ sin was generated by the vibrational impat.

197'.
Jaffee, R. I., et al., "Forming: and Beat Treatment of Corrugated Diaphragms,
ASK Mrans., 11, pp W6047T, 4i78-479, 194&9.

The mount of cold deformstion resulting froa foriming corrugated diaphragMs

of beryllium-copper,. 6D:20:20 corper-nickel-wmiganese, titnium-EMvar
(Ni-Span C), and Grade A phosphor bronze uvas evaluated ad found to consist
of a small over-all elongation which was nade up of larger localized elOngaL-
tions of 5 tc 10 percemt maxiuwm reduction at the tops of Corrugations and
mailer localized contractions at theL bottoms of corrugations. The specific
effect of cold work os the age-hardenirg response was determined on sheet
material. Tensile tests on material age-hardened in the as-received condi-
tion and with an additional cold reduction of the order of 10 percent redu-
tion of thickness indicated that the effect of this mount of cold work was
negligible insofAr as the tim dependence of the aging curves was concerned.
The conclusion was &-a~~ that as far as mechmnical properties are concerned,
a diaphragm can be heat treated according to the ag-hardening curve of t'be
unformed sheet stock without danger of averaging the most deformed sections
of the diaphragm.

-anssen, 0.,. "Asymtotic Integration of the Differential Equation for

Special case of 4mestrically Loaded Toroidal Shells%, J. 3ktb. Ft7s.,
(1) PP 1-17, April 1960.
This paper deals with a special case of toroidal shells where the meridian

- - - ---- 53
 circles bave the axis of rotation as co"O tonMput at Aube ori-in, The
saell deflection theory of thin she3ls leads td a diffternti*l eq'_
WWI~
iahich is given. The purpose ofThspper to to _vresent a now and
acciim.te -5thod to cbtaln the r~articulir inie-T* o-f qtinoft
sametype.

.Tenz~r. C. J., Vapbrag2 Device-$ U. Be Patent 2,*162,308, June 13, 1939.

The invention relates to expansible and contractible diaphrap devices sm3b
as diaphream devices comrising either a single flexible mtal disc-v
vil1 deflect when seijeate to differ~mtis1 pressures on the'da
or a pair of flxIble metal dim-a joined tofsther to f ~jnIble
chamber wrhich will. expand and contract due-to asL o~w~ of pressue on
the inside or out side of said chamber. Ykme p4u1,the Invention
relates to the el.,iinstion of temeratulijfi~ 5~ flexible mtal diapbragm.

KeeL, 3 evie r anXAmysis or uiuuAftive-sigu-imae "Mines5,

Rand Cmop. 90 Wp, August 1963 (B163-21106).
An inreatigation of the basic concepts of cuulative dmW and a compaison
of several current ctm3.atIve-fatigue-dange theories are presented. It
is simn that emmative-dsmage, theories can be categorized by determinling
tbe basic ssiotious they contain regarding (1) hov to prcoerly add toether
the damage produced by eachi stress cycle, iihen iwdifferent cycles are
mix-ed togcetr,, and 2 bo- dam prgressres atiwSh ave.
Several current euxulative-dmge& theories are examined old c ard in
terms of prediction, of fatigue life =zder spi-tri o"Idng. The i~ortant
coseqenc of differe!wes In the theories considered 14 not different Pre-
dieted lives, btiA rather differences In the a t of interimil that Vill
provi~e a qpeclz"ied lifetie, as foind by application of the tbeoriei In
desigz1.

201.
faecbele: L., -Dsgning to Prevent Fatigue Faibras",. RsAd Corp.,
25 PP, Febr'4ary 196% (AD 6L126T).
There are three win "I=~1 area" in preventing fatigue failures: predict.o
of the flietwting etresues that cause fatiffin, tehavior of the StrUftu~al
mterial usdargoing-fthese stresses, and scatter In stresses endbunteiid III
service and in fatigne behavior. This pape coneentrates on tbese major
yrcbleu =*eas. The data moot iMportant for 6esign, and Som techlniqyes cr
using these ftta, are presented and disoussed. The esigner's role in select-
ing the fatigue problest that should be, studied is =ohmasis

34-
i
j---l
202.
nns, A., '"Anlsis of hells of Revolution 4*jected to SBy-etrical
and Smsymetrical ,Loads"i ASIM IMMAns. N9p 46744"42 SOPtember l96I.
The boundary-value problem of defomutioa of a rotatinally ey~trc-shel3l
•"i stated in terms of a new system of firft-order ordinary differential
equations which can be derived for any consistent linear-bending theory of
Shells. The dependent variables contained in this system of equations are
those quantities which appear in the natural boundry conditions on a roi
tionAlly symetric edge of a shell of revolution. A nvierieal -methd.df- -
solution which cobines the advantages of both the direct inteatithad the
finite-difference approach is developed for the analysis of rotatiomaly
syietric shells. This *etho4 eliminates the 'loss of accuracy- ewcowmtdkd
in the usual application of the direct-integration approach to the--aai*sis
of shells. For the purpose of illustration, stresses and displacemnts of
a pressurized torus are calculated and detailed nmerical results are
presented.
I

.+ 203. 2. an, Yu. I., *'CalcJAtions for Spatial (or Three-Dimensional) Construction",
Rasbet Prostranstyentrzk Knstrukt., V 1963 (in Russian).

This paper considers a number of problais involving toroidal shells includ-

ing toroids filled -eth liquid, noncirculwr toroids, an-i toroidsl bellows.
Solution is cbtained in trigonometric series using eergy methods.

204.
- gley, T. M-.. Jr., and Bumcnd, J. P., "Bellovs Failure in Solid Separation
Loop of the ERT Mockup", OM Rept CF-57-2-24e, Febuary 5, 1957.

Me failure of a valve bellows appeared to be due to a ccub:1ation of stress

corrosion and crevice corrosion. Stress corrosion occuwred as evidenced
by thbi tranagranular branched cracking found In the bellazs and in the base
which was joined to the bellows. It seemed probable that chlorides were
present, which, along with the residual stresses present in the bellows
ausemby, created the necessary conditions for stress corrosion to occur.
Crevice corrosion occurred pb ably due to heavy deposits of solids at the
bose of the bellows, which created a condition of orqen Impowrishent.

epp., S, RL, "Hg Pesur Epni J i8t s", ASM POWe No.

" The need for a h~gh-presswe expansion j oint In a _vrocess unit

was etcountered by the Isso Research and Engineering Ccuzay in the 6esgn of
a vapm heat exchanger of a fluid hyroformer. Past failures due to stress.
ScorTosion cracking of thin-valled, low-pressure expansion joints yromte4
rjvcto of the best c= 1cil high-pressure joint; nsy, the-teru type.
-ft4s was because this shmpe joint could only be fabricated to a x1Wm
thickness of O.0T8 In. For this installation a thick-valled switorus

-~ 1 55

-- Y -
 expansion joint was employfed, for which an experintal sttess abilysls vat
as& to establish movement and pressure design limits. These data vere.
then correlated vith theory and a general design fomuula was establisheA.

206.
Mlngbeil, E., "Zur Theorie Der Rotationschalen vog ftandpunkt ftwiscber
Recbnigenn", Ing. Arch.. 27, pp 24~2-2J49, 1959 (in Ormn).
This paper discusses the interrelation of the shell equations of iNeissner,,
eolke, and M~iz. The kMnz equ~ations are four ffrst-order coupled equations
which are the appropriate form for the direct nusrical integration mthods.
No m rical examples are given.

207.
floaner, J. X... and Levine, H. S.,, "Further Ccparisans of Elasticity and
Shell
Theory", ~~~Polytechnic Inst. of Brooklyn, PMRLRp.N.6~ uyl6~
The problem of an infinite circular cylindrical shell subjected to periodi-
cally spaced axisymetr ic band loads is ive stigted using a Reiss e-Nqghdi
I higher-order shell theory. The generalized equilbriu equations, vh!'-Ch
include the effects of transverse normal stress and transverse shear aeform-
tican, were successfuflly uncoupled, and an expression describing the transverse
displaement of the mdian surface vas obtatned. Tesolution of this ex-
pression vas then used to caleulate the stress resultants, stresses, and
displacements. A comprisoni of the resulting stresses and displacements
with the exact elasticity solution and soma loiaer-teder shell-theory solu-
tions of the sme prcblea vas carria out for ratios of Ine to outer sbell
radius equal to 0.7, 0.8, 0.8T, 0.9., 0.93., and for a ratio of distance be-
tween band loads to outer shell diameter equal to 0.2.

208.
Wboojtra, L. F., and Lewoe, X- X., "Lov Cycle Fatigme Research on lull-Size
Pressure Vessels", Welding 3.., Welding Res. Suppl., kl, pp 29T5-30(3,,
July 19w~.
This paper discusses the results of an experimental study to determine the
lw-cycle fatigue characteristics of 'full-size pressure vessels incorporatin
a variety of nozzle configurations of interest to the reactor desitner and
pressure-vessel industry at large. Nmrical. data on stress and strain
concetration factors, redistribution of strains upon cycling, and a dSescrip-
tion of cyclic test facilities are included.

2 cita .FLaeLAadPikt,.G,'ul-iePeme
Vessel Tesuting and its Application to Design", AM Paper No. 63-WA-29d, 1963.

ft resets of fuU4:sixe pressure-vessel tests are sumsewized aund used to

demostrate the va11ldity of design procedures devolcoW by otbars. In

56

V__
 particular, design for low-~cycl.e 6piastic) fatigue strength anI fracture
safty
reinvstiatd. teresults justify the design stress recommda-
tiocis made by PVRC to AM~ and conffru the M~ fracture -analysis diagram-
proce~dure hypothesis. It is emphasized that full advantage can be taken of
high-strength materials only by Improvement of fabrication and Inspection
procedures.

210.
Korneki, A., -A Thin-Walled Toroidal Shell Under Uniform Pressure Lose,
Rozpravy Inzyierski, 14 (1), pp U19-M72. 1956 (in Polish).

This piper presents an approximate, solution of the toroidal shell bounmdedF

by tvo perallels and loaded with uniform pressure. Both the homogene=s and4
particular solution2 are obtained, by asymptotic theory in teras of tabulated
functions-. Several exam.ples are presented.[

211.
i Kornecki, A., "Symmetrical DeforstIon of a Thin Toroidal Shell of IMP-
tical Cross Section", Bull. R~es. Council of Israel, Sec-t. C, I (1), 1959.

t This paper presents an a;Woximte computation method of stress and strain

compoents in a thin-walled toroidal elastic shell of elliptical cross saction,

I
I *smUl
limited by two parallels and loaded arbitrairily but symmetrically with respect
to the axis of revolution. Making use of R. Neissner's variables and neglecting
terms, the vrcb3im is reduced to the deterintion of a camlez
function X(0) satisfying a differential equation with suitable boundatry con-
ditions. Applying asymptotic methods of integration, the solution In
closed form Is found,, the desired function X being tabulated or expressed
W means of simple asymptotic formulas.

212.
Lxrnecki, A., -Asymptotic Solution of a Toroidal Shell Subjected to Eoo-
syrntric Load&", AZA JZdral, a (11), Novmer 19&1.

This paper considers an aqqiotic solution for~ a shallow tcrodal-shell seg-

mnt, for arbitrary nonzsposqtric loads varying slowly in the circumfer-
ential, direction. ?be solution is obtained as a pwar series in sin 6 vhere
o is the angle between tba, nonl to the sbell3 and the center line.-Th
solution converges only for the codition vhere the absolute value of 9 is
less then %/2.

5T
 213.
Kraus, H., "AReview and Evaluation of Cmuter Progams: for-teAuyi
of Stress in Pressure Vessels!', Pratt-&'Witney Alroraft tPWA-2$16, OI;&
1963 - September 19 0i1 .
It has long been the aim of the Pressure Vessel Rese h Comittee to-MV.
the analysis of pressure vessls on a firmer basis,. and as a result,-a't
review and evaluation of existing cooquter solutions wase underaken. After
conducting a literature review of the subject of compi~er programs for
pressure-vessel analysis,. the authors of the most promising programs were
contacted in order to determine their willingness to tarnish copies of thei.
programs and instructions for their use.- The field of possibilities. was
narrowed to four programs whose athors gladly furnished cowxtert decks .AW~
i!nstructions for their use.* A fifth program was dttafted but could not be
studied because it was written in a languae incompatible with the oom-
puter inst-alled at the author'Is company. This report sumaiss the findings.
j It isarranged Into tbwee =aor parts as follows. The introdwction-presents
a simmary of the eqi~at ions of shell theory, a review of the literature od.
computer solutions, ard a discussion of the main methods of solution the&t
are employed in the computer program. The section, Comprison of Comuter
Programs.. discusses the most interesting program .tich have been found and
presents information on the scope of each, ease of use, running tim, numeri-
cal results for samiple caucs and so on. Finally the DIscussion and Conclusion
sections give a summary of experiences wilh an evaluation isnd: recendaions.

214.
hadrin, N. 8.., "The Strained State of a Folded Sinusoidal Shel" Iry. AN SSSR,
OrN, Mekh. i gash., ~,pp 1419-150, 1959 (in Russian)..
This note considers a bellows vith shallow sinuoidal coiplutioas. The
author considers axial and pressure loads as well as thermal, st ?eases.
Formulas are developed on the assuimption that the corr'ugation dep**h is ff&&3
compared3Eo the cylinder diameter. Some calculations are made zhowing the
thermal stresses for different corrugation depths.

215.
Kuxasaka, T. T., and Barnet,. J. W., "Predicting Metalworking Properties From
Tensile Test Dsta",j metal Prog., pp 84~-86, March 1965.
The fabricating characteristics of metals and nomeiAllics can be rated 'by
numerical factors derived fromE tensile-test data.- Once the 'basic equations
are derived,. the method-is simple. Experience in the sbop confirms the
re.tings.

216.
Lekhin, A. A., "Calculation of Certain Elements Which Include Torus-Shaped
Parts",, Inzh. ob... M, pp 66-77, 1958 (in Hussian).

I The asymptotic thaory of Xovohiov is used to develop tables of influence

coefficients for semitoroidil shell segments.

58 - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
 217..
Lemont, E. A., "The Dveownt of Advanced Cy ui rssr iioe
For Ballistic Missiles". Pfreebank -Co., March 196C)D 2105) -

This report describes a research- and developmedt progrst for t-w- ---
ment of advanced cryogenic pressure sitches. Two nbve1-prehsure. iit
concepts using diaphragms were discovered and provin-eieri4isn .-,
vanced techniques of omienvironmental testing were ezployed. 6eiWiitch
was developed that was adjustable between 100 and 1000,psia. Another4s0itch
developed was adjustable for actuation between 10 and 100 p-g. Both
switches had a unique deadband adjustability.

i 218.
218oLanger, B. F., "Design of Pressure Vessels for Low-Cycle
Fatigue", J. of
Basic Eng., pp 389-402, September 1962.

Methods are described for constructing a fatigue curve based on strain-

fatigue data for use in pressure-vessel design. When this curve is usedP
the same fatigue strength-reduction factor should be used fcr low-cycie as
for high-cycle conditions. When evaluating the effects of combined mean and
alternating stress, the fatigue strength-reduction factor should be applied
in both the mean and the alternating component, but then account mist be
taken of the reduction in mean stress which can be produced by yielding.
The complete fatigue evaluation of a pressure vessel can be a %Sjor task for
the designer, but it can be omitted, or at least drastically reducedp if
certain requirements can be met regarding design details. inspection, and
magnitude of transients. Although the emphasis in this paper is on prestsre-
vessel design, the same principles could be applied to any structure made of
ductile metal and subjected to limited numbers of load cycles.

Laupa, A., and Weil, N. A., "Analysis of U-Shaped Expansion Joints"J. Appl.
Mech., pp i15-123, March 1962.

An elastic analysis of U-shaped expansion joints under axial loads and in-
ternal or external pressure is presented. The analysis employs the energy
method for the toroidal sections, and the theory of symmetrical bending of
circular plates augmented by thick-walled cylinder analysis for the annular
plate connecting the two toroidal sections. The general solution permits
the investigation of any U-shaped expansion joint falling in the range of
thin shells, for any arbitrary combination of axial force and pressure
loading. Different forming radii may be assigned to the inner and outer
toroidal sections, and it is permissible to vary the average thicknesses of
the two toroidal sections and the interconnecting annular plate indoendently.
Expressions are given for the load-extension and load-stress characteristics
of U-shaped expansion joints, and a numerical example is presented conering
the present solution with results of existing approximate analyses. The
method presented here lends itself readily to proraining on n electronic

computer,.

59

J,
 220. $-

Laxan, BO is*, pCel-Pa1tigue MnerAq*tiu *er Cyclic L J-ngC_:it;im5

AFEJ, Aeron. Systems;-Div., A&.;-63.6440, Iby1963.
The limitations imposed by synergistic ~or. interactionenea -the vidity
of linear superposition and sition approaches -are disciise. pm&<
ticular, single-variable data are shwn- t bei generally lbadie~ute 1e~
dieting mchanical ]properties under conbined loads and envfrP6 tse ~
Immensity of a testing progra required to. investigate all i- ratci -
tions. or load and environment is illustrated and the sigdificant-ridoctions
possible if synergisms are known is demmstrated. The operative slicrean-
isms and synergisms under ccmined static and fatigue loading are discussed,
with particular reference to microstructural changes of mechanical and chemical
origin. Analytical methods for predicting creep and rupture under variAble
stress from static creep data are reviewed and compared with experimental
results.

221.
Lkee, G. H., and Van der Pyl, L. X.., "Bibliograpby on Diaphragms and Aneroids",
AME Paper No- 55-A-180.- T11 PP, 1955.
tThis is' a ccagrehensive annotated bniiograp1y of the available publisheI d
data on diaphragm and aneroids up to the end of 19541,

222.
Lepik, Tu. R., -Equilibrium of Blasic-Plastic and Rigid Plastic hlates
and Shells%, Inzh. Zhur.,I (3),v.pp 601-616o 19&l (in Russian).
This is a survey article of the developmnts of the elasto-plastic theory
of plates and shells for about 20 years prior to l9(*. The bibliography
includes 160 Russian papers and 91 non-Russian papers.

223.
Li, Y. T., "High-Frequency Pressure Indicators for Airodyzamic Prcblems' t ,
NACA Te-ch. Note 304i2. 52 PP, 1953.
Tbree different types of pressure Indicators developed at the ibtssachusetts
Institute of Technoloy we discussed In this paper.. Nach of these imdi-
cators has a unique feature, but £11 'ire desind In an attempt to ccoine
both high-frequency response and high resolving power into one Iflstrmnti,
Of the mchanicalsselectrical-transducer type of presure indicator, the wire
strain gage leads in siapicity. Tecapa Itance typ is more versatile
because it permits the use of very high frequency carrier system and thereby
cuts down the effective interference In the electronic system. The Gy~M
utilizing the stretching of a barim-titanate disk produces large' sigeuls
and- results in comact design, but it can only be used for dynomc msure-
muet when temperature variations are sligbt. Five different types of
pressure receivers, the cylinder, flat-daphrg, spheric*l-diqrw catenary-
diaphr., wan stretched-diapzraga or memrane typs,, va tested. fte
flat-diaphra type leads the others in simplicity, the spbaz'ical-diapbraws
typ exceeds In dynamic performance, and the catenory-dapbreft type-is the
one least affected by temperature change.

60

y'S
 2-24.

Listrova, Yu. P., and Rudis, M. A."iitEulbimo waoiot&&B- a!

Isy.: Aked. kauk SM.., Otd. Tekh. Nak Nk.i3ii)~ 199 - 33f1%3

This pape considers the limt- load of-- aautoroidalhl c dt~-0

and%0 - . For this case the shell I*sneiijy iniAjjrade titte
It is assumed that the shell Is made of rigid-plastic-I-at~jiil i~i iadd
by pressure.

O5.
Little, R. E., "A Simplified Method for Determining Fatigue Stress Using._
Mohr' s Circle",. IWecine Design, pp 1i3-148, March 4.. 1965.
Fatigue cracks are initiated by alternating shear stress and propagated by
alternating normal stress. Consequently, Maimuz amplitude 'of -these stress
components must be calculated in fatigue analysis. Meaii treus caiopne!"fts
acting on the planes of aximum amplitude of alternating stress also ififluence
fatigue strength and must be determined. This article sho,0 iOhow dir Iifcircle
* can be used to determine and visualize easily and quickly these components
of cyclic stress. This approach also providds a convenient means of inter-
preting visually fatigue-test data for various states of combined stress
with superimposed mean stress.

226.
Losco, W. F., et al., "Pressure-Deflection and Fatigue Properties of Zfrcalay
Diaphragms"' Bettis Atomic Power Lob.,. WAPD-T-2228,p 63 pp, July, 19W.
Plate type fuel elements containing comprtmented and unbonded cersoic fuel
are of considerable Interest because of their high thermal performce and
long life capabilities. The urhonded Zircaloy cladding over individual
fuel compartiments in these plates are essentially thin diaphragms, fixed at
the edges, and are subject to deformation under the action of bydrostatic*
pressure. The pressure-deflection and fatigue characteristics of sixh
diaphragas,0 under a variety of ouit-of-pile tmperature and pressure conditions,
bave been determined and are described in this paper.

22Lo.. E. D.,. "The Use of Pneuatic Diaphragm Units",.

Instr. Eng., (5),.
pp 85-1 April 1958.
The principles of operaition of pneuratic diaphragm units are desceribed.
Their use in rerforming simple operations with air--pressure signals Is
* indicated, and it is shown how they wW be over-connected to provide awe
complex air-pressure relationships. Examples are given of their ap~licition
to automatic control.

61
228. -

yon, F. H., and Grover., H. "ome

S. -%&.stions for Research iniDta'
Technology", AsoE Paper No. 56-A-221, 9 pp.. Noveber 1956.
A number of suggestions for research had appeared3 in tbi 11Ut~tuie ana:-
others had beean advanced to the AMh Diapbragn Muconite. These *5
discussed in this paper frcm the design viewpoift. The present.ation .i-
tended to provoke discussion and suggest pocuibly illuminating j *_is
to certain problemus.

McClaren, S. W., and Best,, J. H.2 "Plastic Strain Accuuilation'- New Fatigue
Life Forecaster?",. SAE Journal, 73 (9), pp 68-T5, September 1965.
The behavior of a group of aerospace materials unde repeated loads at--
high uniaxial and blaxial stress levels to 3prodvice low-cycle-fatigue d@Me
has been evaluated. The results suggest that plastic-strain acciulastion
mVs be a better predictor of fatigue life than Miner's theory.

McKinney', G. E., -Metal Bellows", Mech. 249g., ~,PP 5T3-574.. June 1957.-
As discussed in this article, the formtion is one of the awe difficult
steps in bellows minufacture, since final wall thickness is rarely more
than 0.001' in. to 0.010 in. Extremely careftl control of tube dimensions
is inndatory, since the characteristics, of the finished -bellows depend Ii a
very, large degree on the condition of 'the metal, the fim*1 dimensions, and
uniformity *of the walls in the bellows tube.

231
"MchiedBellws. are Ligbt, Resist a&igue", Space/Aeroti,. pp 60b63,

Makarov, V. X.., Lakhtin, A. A., and Lo'vtukiy, E. V., "The Possibility of

Using Convoluted Bellows Under High Pressures", Ddam. ikshinostr., .,, PP
26-29, 1959 (in'Russian).
This paper gives som experimental neasureints of stress for a semi--
toroidal and U-shaped bellows under internal. pressure and axial loads- and
comnpares them with a theory developed by Lskhtin (not described).

62
 _4

233.
Mrca, P. V., a"d Turner, C. It, "Elastic Solution in the Limit Msiig
of S fls of Revolution WihSeilaeetc 'oas ~l~~; ~ -

ftg. Sc~l, , (3), pp 252-257, Septembei 1961.

The elastic Solution is uedI6c calculate lower and uer b~und."
call 1O d ihalls of-re,o1uticu.I r~ ootii h e' i&
nunler lof~_Oa3 6 o-6
nwber f cWizirie~yd s'urface s arej used t6;_C-: eito2c
&lo
fT ecVii -
ismi. 'hresults btja-recoi~ie with, expierimekil a.Ohriat

Marcal1, P. V., ar4,Turner, C. E., "Iiinricsl Anayisoth ltc-ati

Behaviour of Axi*y6etrical.y Loaded Shells of Revolution", T,
Sd.,P 2 (3), pp 232-23T, 1963.

Equations have beeni derived which enable a numrical 'analysis of axisy*%ntri-

e13 &Ladd shells of ievolution to be made. The'" e&uitions apijo:y 'eqall
to small or large deflection shell theory with linear or nonlinear -itress-
etrain laws. NumericaLl results are Presented for the case of an elastic-
i01aic axially loaded --arrugated expansion be3llcrrs.

'H- 235.
Marin,. J., Paper, from -Progess in Applied mterials Rsac.Vp =9
19, teidon & Breach, New York.
This Paper surveys the fields of plastic and creep properties of ehogneering~
materials as related primarily to the macroscopic stress-strain or inchanical
bcehavior of materials.

236airtin, D. E., "An Energy' Criterion for Low-Cycle fti:gue", ASU Trans.
83D.- i, PP 565-571, Deceuber 1961.

A lair-cycle fatigue criterion is propo med which eiplcys as an idex of daange

the portion of the Dlaftic-strain bysteresis which is associated with' the
work-hardening sagent, of the stress-strain relation. This damge worl-energy
rfation Provides a theoretical basis for the plastic strain-cycl equation:
N/&c= C, which was previously advanced emirically. The constan C i
this equation had Pre'viously been approximated by Coffin (AME 11(1961), Rev.
3562) on the basis that a tensile -test is equivalent to a 1li-cycl fatigue,
test. Author evaluates C on the basis of energy absorbed ding work-
hMrde~g alone in static tensile tests. If ef is the static fracture
ductilty he finds C equal. to ef divided by the sar roo of2w eCffin
haks found C equal to ef divided by 2. Author comares his value with pub-

f 6tests
lished low-cycle fatigue tests to find better agreemint fore room-teneiaure
than Coffin but Poorer agreent for tilevated-tesperature dati.

63
 ~~37

The Mrtin Company, aet Rert. so. 11006, hkvaliistion teg of id ~

Flexline",L u 1Y 31,. 1961 (AD 2837141).
This repcet Presents the test pruceduwes used and the-test results, obtalnmd
during the evaluation of an insulated 2-in, flexible tibg-c~c
designed for use- with nitrogen tetroxide propillants. .The _i
nitrogen flow (18-21 lb per ida., 15 psi, 4oo aiirgen and nirogeu.
tetroxide lcr (114-1 lb Per Yrin., 13 Psi, 100 F), and sttcMsur 7
psi).

The Martin-Denver Co., "Qualification Test Report of Alinm I-dd lw"

August 23, 1961. IMP U1570.69-00-7P3-0l.
The tests of a 13-inch-diameter aluinum bellows Included visual examiac-
tion, proof pressure (35 psig), spring rate determination, ultimate presrkare
(90 psig), and life cycling (4e50 ccmpreusion cycles).

23-The )krtin-Denver Co., "Flight Certification Test or,Oxidizer Pressure

Line",. Dece~er 26, 1961 (AD 295351).
Tests of two flexible metal hose ass'mb1ies Included visual examination,,
pressure and deflection cycling, leakage (bubble), pressure (165 psig),
Pressure drop (1.143 lb/sec G52), vibration, operational cycling,9 and
burst (1650 psig).

jh rinDne Co. "Flieht Certification 'Test on Pressurization Line",

Tests of two flexible metal hose assemblies included visual examnination,

Life tests (115 psig pressure, deflection amisalignment cycling),pro
pressure (205 psig), *ibration, flow, operational cycling, andi burst (82 -

1010 puig).

2411.
The Martin CoopszW, "Qualification Test Reprt of Flexible Metal IBellows",
T*I. 2111, June 22, 1961. IXP 315*706900-73-.. pesrzto

compression (ii50 cycles', and ultimate Pressure (120 psi).

'I4
 241 The Martin Ccamany, ")ftaflurgical Investigatiom or Iaakage. in Xissilel
3-66 Stage 17 Oxidizer Feed Line Beilos", June 15, 19a1. "W'45 .T 90o -F3-ol.
I A
nitrogen tetroxide leek was caused by streus. corrosion between- the inr
I and outer plys because of entra-ent of chloride-ionti!ing -coaainanta.'!
~ ~Because residual stresses caused by foming an ntllt &0,0oq,o qir-
ments can always be, expected, cleanliness b'~tieen the two plies is of. the
utmost Importance.

2143.
I~tbexny, J. D., "Bellows Spring Rate for Seven Typic-al ConvolutioA ibpes;
Data Sheet", "Machine Design, (1, pp 13T-139. Aa~ry 4, 196Q.
W
A bellows is analogous to a spring.' Several theories have been developed
to give load-deflecticn relationships for different bellads configurations.
Formulas based on these theories are presented for determining the spring
rates of T typical convolution shapes.

2414.
Nathe son, H., and Eden,, N., "A Highly Sensitive Differeptiel Wianometer",
Rev. Sci. Inst., 12, PP 502-596, 19148.
This paper describes the construction and performae of a differential
manometer with a saudtivity of 0.001 mm. of mecury over the rena. -of ab-
* solute pressures from 1 ma of macury .to I atmosphere. The sensitive element
eaplcrfed is a pair-o:: nesting diapres, displacement of which is mesured
by the resistive mabalance of an ur*boaded electrical displacemenkt Sge to
wehich the diabaas are attached. Calibrations are presented for the aensi-
tivity at I- and 10-um gage current.* 1ta are prese*nted which indicate the
operating characteristics under varying conditions of absolute press;e
t"persture, gage current and geometric orientation.

thews,. C. C., -Picking Metal Tubing for Flexing Uses,

Ma1. M.in Design

Constant metal movement makes tubing used In flexing parts prone tb--ftigue
failure. ftis diqcuvsion of flexing applications is not limited to tholit
j end uses specifitfly destgned to provide flexing action, i.e., DOUrdon
tubes, bellows, flexible mtal hose, etc.* It includes those applications
where metal uoamt is umwelcom and usnelly caused by pressure surges or
eqiWnt vibration, i.e., pressure tubing, bydramlic lines, diesel engine
fuel jies., aircraft parts, etc.

_ _ __ _ _65
_ _ _
 2146. -

. J. "g ~
Rat rature Metal BelM 8eAle for Af~*tk~~sl
Accessories, J. of for Ind., pp1 28l-283 kAaut 196.,
In this article the- following conclusions are Me&: (I el rfn over
a wide temperature range require greater precision and. q &W l ti
the standard cartridge seals usedlinmoat rota'ry? equd-i+, ()th
of flatness on sealability is of prhme imortiiie ii~ thiisfatssut
be maintained over the entire envfrormental. range of the'iseul asm~,~3
seal-,ring flastness in the operating position is mawe imortant than flEk-
ness in the unloaded state because sealing is required In a loaded posi1tion,
not in the free position, (4) as Wped increases, the "r~leaof eft bcO
and fatigue also inczruwene, (5) the runout and surface quality of thet rntingi
ring is of paramount impot~nce, and (6) the selection of bollovs matiihls
and heat-treatment hare a~ definite effect on operating life arid coitinUous
quality control must be maintained on the assemblies to guarantee the high
reliability that is required for aircraft and missile copzents.

MIeber, R. L., -Component Gimbal Joint Bellows Liquid (Ocgen Ingine Feedt
Lines", Solar Rept. Rm 1434. June 26, 19&x1 MZP 115.16.75.00-71-02.

Thae development tests of the gimbal joint bellowd (7.9 inches diamter, -

2 pyjIno~e V.8), included determination of axial spring rate (coilression)

* the angulisr spring rate (dflection through 6 degrees), angar fatigue life
(with Intcernial pressure of 103 psig), and pressure stabilty (22 deg. Aigu
lated position, pressurized to failure). -

2148.
NIlor,. P. 3., "The Ultimate Strength of Thin-Walled Shells and Circular
Diapragms Subjected to ]Vdrostatic Pressue" Intern. J. lMech c.
(2/3), p 216-228, April 1960.
The ultimte strength of thin-willed cylinders, spherical shells and cir-
cular diaphrgm subjected to hydrostatic pressur is Investigated for miaterials
where strain-hardening characteristics can be fitted by the empirical equa-
tion 3 - A(i) + ;)n. Experimental values of nxiwum pressure obtained In
-the deforming of circular diapras; of initially cold-worked materialsae
shown to be In good agreement with a theoretical treatment by Hill.

"Meta&L Bellowhs Metal Ind., LO Wi), PP 69-70, January 24, 194T7.

A brief description is given of the hydraulic method of forming metal bellows.
This involves unrestricted flowing of a thin-walled metal tube under consid-
erable hydraulic pressure in a collapsible die, the letter omprising a
number of plates equal to the number of convolutions and spaced
equidiptant, on and surrounding the tube. The metal flows tranavezoely
aftwe the plates under this internal pressure, the tube being collapsed
e*dwse as presmare is applied, the ballows thus being formed in one con-
tinuous operation.

66
 II ~ .LtIE~~i ellw",Process ControlL,2 5,pp 173-47-T, Iby 1955.
Bellows are used in a great nmber of int~ntsand in. other equimet
but inny users ire handicapped by a l:c=oIsfiint i' oto~o~h
subject. This article is designed to help by disc-Usn scm. of -the--w
inortsnt Points. The Particular types described are k*-koulically -famed
* s~etal bellows, but apart frost the method of mnfcue-adthe speclal
qualities these possess, the article applies eqtualy well. to -other -tyjes.

-Metal Bellows-., Nech. 109g. pp M5A1, June 1957.

brief description is given of the design an&inufactuwe of formied

"Method for Selecting a Metallic Bellows", Design News, 8(16), pp 57,. 59,
61, August 15, 1953.
7:
After selecting the proper bellows system, the designer is faced with the
* -problem of deteraining the exact bellows specification. The problems enter-
Ing into this calculation are those of determining the operatn charac-
teristics of the system. The factors entering Into these calculations -are
first, the natural iping-resistance of the bellows; seond, the resistance
* of the opposing &QAdjsbz spring strength; and third, the fmactioml resistance
A of the valve or switch mectmnism excluding the adjucting spring. Informtion
is given In the article to-assist the designer.

253

'*yer, R. R., and Burac,p N. N.,p 'Conical Segment Method for Analyzing Open
Crown Sbells of Bevolutien for Edge Loading", ADA J., . , pp 86l
April 1963.
A solutiou, accurate, rapid, simple enough for design use, and valid for all
regions, ba been obtained for the stress distribution and influence coeffi-
-I cients for' a vsriable thickness shell of revolution fomd by a generator
of arbitrary shape. fte shell is s tivided Into a series of equivalent
conical *aents whose Individual thicknesses are the local semnt average.

I Conditions of continuity then are applied at thek bondaries of each conical

segmnt to evaluate the IndeternmUt edge sheows and somnts using digital
equipment. Influence coefficient omparisons for a wide range of sbell
geomtries are a&between the cowe solution and solutions by other mertbods
from the literature and show agreemut withla ii percent. MW cone solution
reciprocity relations ore shown to be valid to five significant figures*
Limiting conditions indicate that good spprcimtions of the influence coef-
ficients and the stresses can be obtained by using 10 comas In most casee.

6T 77

T.

4 --- ---- ~-~L

 ~MleD. 0.,. "Direct Mechanical Dtermination of the Dnauc 3epasf
Diaphragms". J. of the Acoust. Soo. of Am., X (8), pp'l 147l-IT8, Autust 10&1.
The freefield resonant frequency and the deflection under given staic Pres-
sure of a thin, circular, clamed-edge diaphiaga -ay be calculated tbO**tic-
ally by well-known equations. Also,. experimental seas exist fori a -det-
inination of these quantities. No satisfactory theoretical WiexprmhnW a
method has appeared, however, by which it is possible to obtainS fr a-com-
plicated physical system, the diaphragm deflection as a function of the
frequbency of an applied sinusoidal pressure, over a frequency range that
includes the natural diaphragm reso~ance * An experimental method is described
by which a piezoelectric driver is employed to generate a slnusoidal pressure
of variable frequency, in a confined gas, to which owe side of a test dia;-
phraga msy be exposed. Equations are derived describing the gas-coupling
medium and the piezoelectric driver. PV use of these equations, It is possible
to predict the characteristics of the apparatus. Modifications of the ap-
paratus are reported that extend the dynmaic-;res sure amplitude and the use-
ful frequency range, and the possibility of utilizing the apparatus for
dynamic calibration or pressure gages is noted.

V1iller, D. R., "Bibliography on Tbermal Stresses and Low Cycle Fatigue"..

Knolls Atomic Power Lab., Report KAPL-2048, August 20, 1959,
This bibliography consists principally of references to domestic publica-
tions on subjects that pertain to therstal stresses and lowr-cycle, fatigue
in nuclear reac~ors and power apparatus. Both nsjor subjects are Included
in the sam bibliography, because cyclic thermal stresses or ther-ml-differ-
ential strains are frequently sufficiently hiab to causie fatigue crack.
ing onthe order of 100,000 cycles or less. The problem of fatigue cracking
in this range of cycles-to-failure is cons.1det-ed low-cycle, fatigue. The

I references are placed, according to the cbareter of the principal content.,

in the following groupings: Material Test Xtas Semmice Experience and
Component Tests, Stress Analsis, Design, awl Wieceflaneous.

W-1r.D. R., "Therml-Stress Ratchet Macbmnim in Pressure Vessels"t

AmU PaperNo. 58-A-129, 1958.
The comintion of cyclic therl streases sand sustained Internal pressure
in a vessel is shown to be a source of progressive expansion of the vessel
if the stresses are sufficiently high. CrIteeta Aeseaxted allm det-ernina-
tion of limits to be imposed on stresses In order to prevent progressive
expansion or to allow estimation of the expauttin per cycle where stresses
are sifficient to produce growth. The effect af strain-uzdening of the
metal on progessive reduction of the growth rate Is discussed.

68

......
_
 Ludortod. The ditogramlecs fo tiforces., deleetlods., an lstre sap-
ptrtis it thinshll interetion. re coseed.s maye roral urices. ofe
reltion shel evails uingse anerzd ad axmetin Te
smodultonMet
lrat'g onseed aludte shrbitr aitribtion~st of mliantalsiznfo
Eaxinu sreae withing. ths laos ofdte resetn dectiob~nds. Mhe efftsa
to tmeratre oriatds, hich te altredo fthicares
and ireton asvfwell
ascesrfugalsigaiint fiue roaticnrabou. the is ased pibraistion
prtbes wcthe sh esecionus he aonisedof reogtio. nuital Inte-
srtnproaesnlodngs atdhe shell ofite
utomtidjstmAlthof inotra fror

is written in 7CRM3iA, a nuser of sub~routines are 'written In Philco 2000

asebly language, v~ich li-1ts use of the program to this machine.

258-
Morrow, Jo~ean, and Johnson, T. A., -Correlation Between Cyclic Strain Range
and Low-Cycle Fatigue Life of Metals:, Mter. Rles. and Sd., Ip pp 30-32,
Jammay 1965 IA65-14624l).

An evaluation is described of available dsta on the fatigue lives of 118 metals

(ine Iudux1 21 steels, 2 cast nickel-basedi alloys, andi a hot-pressed beryflim)
to assess which of the three sugstions (&,-- to Coffin, to Menson, and to
Peterson) for a universal value of cyclic strain that would provide apiroxi-
mately the sam fatigue life for .11 metals, gives the best corelatica with
the data. A tabular coqprison of the three rules of tb~ ]proposed zlesrlY
shows that Peterson'sa suggestion that ±1 percent strain zamses faillme in about
1000 cycles gives the best agreemnt with the central tendency of the expe-ri-
mental result a, andi bas the least scatter. None of the three auggeated rules,
hovdeverx, gives particularly safe estimates.

259 . z, H., "Zin IntegratIonaverfabren fur die 3erechnun dier Niegeapannungen

Acbensymtrischer Schalen tinter hAcbenzymtripaher jelasturigm, Ing-Arch.,
PP~ 103-11.1, 1951 (in GerMan).
itfnz treats the anlysis of axiqsmmtrcslly loaded shells of revolution by
use of the calculus of 'variations. With this spfrosch it is possible with the
use of know& Integras, to redwue the sixth-order system to a qystes of ft
first-order differential equations for which namerical-integration methods are
applicable. No wimrica exmles are included.

Ii 69
260.
Murp1.r, G., "Analysis of Stresses and Displacements in Reat-Exchanger Expansion
Joints", AHR4 Trans., 74, pp 397-402, April 1952.

This paper presents a procedure for the analysis of stresses and displacements
in an exyansion joint, when the latter is assumed to be a surface of rXvd1u-
tion. The analysis as presented employs a method of euccessive 'approximtions,
but in general, the solution will be rapidly convergent. Possible thta-al
gradients in the joint are not included in the analysis as giVen, but may be
introduced in the solution.

261.
Nakamura, K., "The Tables of the Functions for the Stress Analysis of the
Toroidal Shell of Circular Cross Section", Bull. Faco Eng., Yokohama Nat. Univ.,
10, pp 25-49, March 1961.

In spite of many investigations, the analysis ofstresses in the toroidal shell

of circular cross section is somewhat difficult for practical use. This my
be partly due to the fact that such analysis re~uires a considerable amount
of numerical work. The author has developed a general method of solving the
toroidal-shell problem, which permits this difficulty to be overcome.

262.
,Nakamura, K., "A Contribution to the Analysis of Stresses in the Toroidal
Shell of Circular Cross Section", Bull. Fac. Eng., Yokohama Nat. Univ., 6,
pp 103-109, March 1957.

Most practicable pressure-vessel heads have profiles made up of two circular

arcs. To analyze the stresses in such vessel heads rigorously, calculations are
based on the Love-Meissner theory, but the shell-theoretic calculations usually
involve a great deal of numerical work. In this investigation, tables of solu-
tions of the simplified basic equations for the toroidal shell were calculated.
These tables which are valid within the limited range of interest are presented.

Neal, Molly, "Survey of Expansion Joints for Pipework Systewm, I", Eng. Mater.
and Design, 8, pp 168-175, March 1965.
This article contrasts the respective advantages of rubber expansion Joints,
slip Joints, and various forms of bellows Joints, and surveys the units cur-
rently available.

264.
Newell, F. B., "Diapbragi Characteristics, Design, and Terminology", A E
New York, 74 pp: 1958.
in a logical.
An beentomade
effort hasand
arrangement, showtohow they arediaphragm
describe affectedcharacteristics
by changes in material, dimen-
sions, and treatment The manual is divided into two principal pats: The

70
 -t -t, Y -A.v

first defines a disphragm and its perfcrmance characteristics, describes samle

methoda of measuring and representing them.. and shows howthey are related-&-nd
used; the second part deals with ways in which diaphragm -cap be ccorntru ted anl
howf vearious design details and treatments affect performance characteristics.

Newland,, D. E.., "Ruckling of Doub~le Bellows Expansion Joints Under Itternal

Pressure", T. Mech. Eng. sai., f (3) 196&.
Coritkgated bellows expansion joints may buc.kle under internal pressure in the
same way as an elastic strut may buckle under an axcial load. This paper Is
concerned with the aralysis of this phenomenon for the "universal expansion
joint" which incorporates two bellows joined ,y a length of rigid pipe. T-he
principal conclusion is that, by providing a correctly designed supporting
structure, the critical buckling pressure car.be inzreased to up to -four times
in value for the same system with no supports.

266.
Nippas, E. F., and Fishman, Hi. B., "Optimum Spot and Seal Welding Condtions
for Inconal V, Welding J., Welding Res. Sup. ,pp Is-4s, January 1954.
This paper describes the determination of the optimum spot and seam welding
conditions for 0.010, 0.015, 0.021, 0.031, and&'0.062-in.-h Inconel X.sheet.
Inconel X is a precipitation-hardening alloy for high-temperature use.* The
unique properties of Inc-onel X and their effect upon spot and seam -welding
are discussed.

267.
Nolte, C. B., "History of Bellows-Actuated Flow Meters", Instrumients, ~()
pp, 79-84, January 1950.
Trhis article describes in some detail different types of flow meters and the
development of bellows-actuated flow meters. Advantages of the various con-
figurations are explained.

268.
Nolte, C. B.,, "Rupture-Proof 304oa-s-Type Flow Meters",. ISA Proc., ~,pp 44-
47, i930.
A brief history is given of the development of bellows-type flow meters which
would not be ruptured by overpressure.

* 269.,
North American Aviation, Space and Informtion Systems Divn. Test Rept., LU
Feed Line Components, Saturn s-Il", march 6,. 1964.
Prior to the tests and results described, problems were encounitered with the

71
 )
AM Z~

orignalbellws Acominaton esig* f smll onvlutin rdii ndq6*;

sbceo
origna bdeigA ccleifatione ofssmall radi
cycvles)oth an te-

perature and at L212 temperature while pressurized at 88 pa1g. The cycle life
exceeded the endurance requirements. In crier to verify and supplement theoret-
ical calculations of the gimbal joint analysis,. strain gagps were applied in
critical areas on the gimbal joint during the pressure tests. In additiop, a
stress coating was applied to verify the accuracy of strain-Sage locations.

10.Norwood, D. L., -Sheet Metal Formability at Ambient Temperatue% Metals Ba.

Quart., W(1, pp 41-51, F'ebruary 1965.

A method was developed for predicting formability from the geometrical parameters
and the mechanical properties of the material. This article describes how it
can be applied to predict the sheet-metal formability for any material used
in the 12 most important conventional fowming processes.

21.Nothdurft, H.. "'The Characteristics of Metal Bellows".. Regeluzngstechlik, 'I (10O),

pp 334-338,'195T (in German).
In the use of metal bellows the following characteriales are of interest:
(1) elasticity, (2) permissible stroke length (3) pressure resistance,(i
critical capacity, (5) effective area, and (65 durability. Attespts to find.
theory of plates
swe of these parameters with the aid of equations used in theMethods
gve results that disagreed with values found by experiment. for the
mneasuring of these parameters are therefore described and the results of these
experiments are stated in the frm of approximation equations.

212-, "Survey of Componet Requirements and Availability for Gas-Cooled

ctlear Reactor Power Plants", The Franklin Inatitute, Octob~er 1962.

wae we-ttentoto assess the state of the art in design requirements and avalla-

and lists, of manufacturers are included.

T2
 2T--

Novoshilov., V. V., Thin-Shell Theary Translation of 2nI Edition, P. loorcdhoffy

shapes. A considerable portion of the book is devoted to the complex formu-

lation of the equations for shells of revolution. Tae asymptotic solution of
a number of shells of constant curvature is discussed in considerable detail.

2741.
M=.n, H. I)., -A Guide to Static-Pressure Transducers TMat Have Diaphragm,
Bellows8, or Bourdon Pressure Cells", Prod. Eng., Lo (1),. pp 48-4&9, January 1959.
This article describes pressure transducers which sense pressure directly and
coatinuoasly, as simple-Bcurdon pressure gages do, and which convert the result-
ing movement or p3sition of the pressure element into electrical units. "Static"
here means low-frequency responses--less than 14 cps. Another article, on piezo-
electric transducers, covers higher frea-tencies.

Onat, E. T., and Haythcsrnthwaite, R. M., "The Load-Carrying Capacity of

Circular Plates at Large Deflection",, J. Appl. M~ech., ?1 (1), pp 4,9-55, 1956.
This paper presents an appoximate analysis for the load-carrying capacities of
initially flat circular plates under various loading and edge conditions nnd
subjected to slowly increasing load. Mde load capacity after finite deflection
is estimated by as~uaing a velocity field based on the boundary :,nlitlons and
on the incipient velocity field of the flat plate, the analysis bex-jg made for
a rigid plastic, nonstraft-hardening material that yields according to the ma..,ci-
mum shear-stress criterion. In several cases the results obtained ecapared
favorably with test data for mild-steel plates; however, for very thin plates,
better agreement was cbtalned by means of a purely membrane-type analysis,
which is else, presented.

Op het Veld, A.J.G., uCr-Ni Steel Grades at Low Temperatures (Below -180 C"
Metalen, 20 (4), pp .0-115, April 1965 (in Dutch).
A critical survey of the nehsnical properties of austenitic Cr-Ni steels at
and below -180 C is presented. Attention is given to the influence of differ-
ent structural factors, such as cold deformation, and the presence of matenl-
site, ferrite, sim phase., carbides, and nitrides on mechanical behavior.

T3
 Osilova, L. N., and )hwarkin, S. A., Tables for Aakls-of 'ToroidAl hl"i
Moco, ad. NIkkBSSR, 91 pp, 1963(n uia.

the asymptotic solution of toroidal sells by 16vzilov

2xeniv le, based on hand calculation of toroidal shelli. Punctibds are
ted to permit the
t;are pr ted in inter~yals of one degree in meridional angle end 0.01 in the pai..-
-ter: R
a=-2, o !r ot!5 1, where r 0is the radius of the center of the torus .from
the axis of symmetry and Ro is the radius of the torus.

Osterman, J. A., "Study and Preliminary Design of a Hermtically Sealed Eydraulic

System", Lockheed Georgia Company, August 24,. 1962 (AD 285c81i7L).
Investigations of techniques for hernetically sealing hydraulic components and
systems were conducted, and the preliminary evaluation and selection of internal
seals,. hermetic sealing devices, and new system concepts were completed. Tenta-
tive requirements and design criteria for hertically sealed components and
systems are described in the first quarterly progress report.

4 2T9.
Ota, T., and Ramada, M.., "Oni the Strength of Toroidal Shells", Part 1, A Propo-
sition on the Solutions Part 2, Examples of Solutions, Bull. Japan Soo. Webh.
rung., 6 (20;, PP 638-654I, 655-665, November 1963.
* A method far solving the fundamental differential equations for the symmetrical
problems of toroidal shells is proposed In this paper. Teperturbation mthod
is used, and the solutions are cbtained In such a form. that the numerical
values of the solutions my be easily found,, if the values of the parameters
are given. The ranges of the values of the parameters In which the solutious
are available with sat isfactory accur~acy are checked..

280.
Palmer, P. J.., "An Approximte Analysis Giving Design Date, for Corrugated
Pipes" Proc. Inst. Mech. &ug.,, M (20), pp 635-G*il, 1960.
This paper gives an approximte method of analysis for corrugated pipes and
ducts, of the type in which the corrugations have a constant radius of curva-

strain energy, in which the smllesUt possible numer of taerm are employed.
The advantage of the method is that results giving extension, benVing stress,
and circumferential stress can be evaluated for corrugations with different
included angles, and simply presented in graphical form. These results are
then readily available for design consideations covering corrugated pipes with

714
 -~~~ -11-~ X -i- ~ A-

internal pressure, axial load, and bending moment. The approximate result skar
compared, 'where possible, with some results frcm more precise methods-of.~~a~
and also with some experimental results, and the comparison S9hae -the prwe-f
method to be sufficiently adcurate for design consideraitions

* 281.
Palmer, P. J.., "A Method of Analysis for Axially Symetrical ShelJls With
Constant Meridional Curvature", Q rt. Mech Apl . i Ph,
p iI
1959.
This paper gives a theoretical method of analysis for shells of constant thick-
ness forming a surface of revolution with axially symmetrical loadiing and with
constant meridional curvature. The method of solution is to solve directly the
basic differential equation governing the problem. The method is ccmpreensive
and applies to a wide range of design problems, although in the presint ifistemoe
it is used to determine the stresses in pressurized corrugated duating.

282.
Patterson, J. L., "A Miniature Electrical Pressure Gage Utilizing a Stretched
Flat Diaphragm", NACA Tech. Note 2659,, l 7 pp, 1952.
A variable-air-gap inductance type of electrical pressure gage is described
that is basically T/16 inch In diameter and 1/4f inch in thickness. The Sage
was designed to masure pressures fluctuating a. high frequencles. It is Sclso
capable of measuring steady-state pressures with errors of less than 1 percent
of full scale and has proved to be of value as a general-purpose electricl gg
for aeronautical work where small size and minimua response to acceleration
forces are imp:,tant; factors. Design equations and curves are presented which
can be used to predict the deflections and fundamental natural frequencies of
stretched flat diaphragms.

28.Payne, D. J., "NiericalAnslysis of the Axisynmtric Bending of a Toroidal

I Shell",. 3. M~ech. Rag. Sci., i&Mi, PP 356-3&s4 1962.

Finite difference approximations are used to analyze a synetrically leaded

toroidal shell. It is shown that the equations derived by R. ]Reioszw. in
terms of II~ro (the product of the radial force per unit circumference normal
to the axis of revolution and the rqKiux measured noml to the axis) and V
(rotation of the meridional tangent) avoid the difficUties which occur., &Wing
to a singularity at the pole circle of a toroidal shell, when iNeissnar' s equa-A
tions in terms of U = Q,2 (the produat of the lateral shear force per unit
circumference and the second principal radius of curvature) andi V are used. A
numerical example is solved with both forms of the equations. The values ob-
tained for Hfq, Qq, and V 41.fer by -less than 1% near the edge of the shell but
in the region of the pole circle there is no correlation between the two sets
of results. Co~armson with Turner'sa analytie solution shows that the Values
and positions of the mxlua, and minima differ by UP to 6% and 3 deg. respectively.

T5
_ __ _ _ _ _ _ _ _
 284&.
Fenny, R. K., "Axiwyzmtrie Banding of the General M.fl of- Is vout M_.Miag
Creep, J. Ilecb, Ing. Sci., § (1), pp 44-46, 1961.
There are many applications in various fields of engineering where 'thinuhell'
theoz7 can be used to give adeqjuate solutions to practical problems. Although
a lot of attention has beia dtvoted toward solving the elastic nhells, rela,-
tively few investigations' Z shell bthavior In which jilsticity or creep occur
have appeared. In dealing with the creep of cylindrical shells, Poritaky fol-
loved the procedure outlined by,Mendelsohn and others for soaving plate and
disc problems. A different approach used by Onat and !iikel or by Calladinle
dealing with edge-leaded cylirders woulAI Meufre further development before it
could be used in practical preblems. Tbe present p.,pose is to extend the
method of Mendelsohn to the general shell of revolution which is loaded and
heated wtisymmetricaly.

2.85. PeesI. "The Role of Stress Distribution in Fatigue", Exptl. Mech.,

i(4), pip 105-115, April. 1961.

This paper constituting the 1960 William X. Murray lecture to the Society for
Experimetital, Stress Analysis, is a review of an area in which the author has
distinguished himself. The main thesis of the paper emphasizes the importance
of the stress distribution in the region of the mximua stress In understanding
the strength behavior of structural numbers under service conditions. A dis-
cussion of the efLfective use of surface hardening illustrates this point in
sinple, yet effective terms. An illuminating presentation of conditions for

I crack formainad rpgtion in notches numbers should pro~e of value to

design engineers.

286.
Pfeiffer, A., "Theory of Corrugated Diaphragms for Pressure-Heasuring
Instruments%, Rev. Sci. Inst.,p 8, pp 660-6&, Septenber 194T.

A1pothetieal model for the distribution of stress in diapbragms is discussed.

It is ccestructed of mebers in series, with each consisting of one flexuiral
and one tension spring acting in parallel. This hypthetical model leads to
a correl.ation between pressure P acting on the diaphragm, corresponding deflec-
tion of the diaphragm h, effective diameter 2r. sheet thickness d, and "plate
modulus" F, as a masure, of elasticity. This formula checks well with the dsta
fcx the author's test material as well as with other experimental results on
diaphragms with pressure-proportional. deflection. The probable relation of the
separate terms of the formula, to definite annula zones of the diaphragm as
well as its applicability to diaphragms with other than pressure-proportional
:1 deflection are discussed.

76
 Pike, B. We. and Gibbs, N. Be. "Study on Aneroid Capsules%, Jo Appio Phys.p, -
~,pp 106-108, January 194#8.
The analysis of aneroid capsule performance by the expansion of the observed
Pressure-deflection curve in Gram suwtion-orthogonal polynomilals is illusitrated.
on a series of 15 runs on 5 typical capsules. All of the deflection curves
* examined contained significant cubic and quartic terms while a few shoved-fifth-
order terms. These high order terms are very unstable in ,vagnitude and sign for
successive deflections of the same capsule, and this accounts for the very com-
plex hysteresis curves which have been observed. The relationship of the6se high
order terms in the deflection characteristic to the accuracy and cost of precise
aneroid barometers is brought out, and it is suggested that barometric systems free
of elastic redundancies must be devised if accurate barometers for radiosondes
and similar instruments are to be mass produced.?

2883.
Pohl, S. We., "Compatibility of Bellows Material With N2C0 and NWH", Douglas
Aircraft MSSD Final Rept. M.64-151, 1964.
N2Cz exposure tests were run on AM 350-SCT bellows samples at T0 F and 160 F with
o and 1 percent water add-ad. Three samples each and three stress levels (above
yield, below yield, and no stress) were included at each condition. Thne exposure
tests for 2-week., 4-week, 8..veek, 12-week, and 24 -week tests are reported. Com-
parison tests have been included on AM 350-H and SS 311T-A.

289.
Pollard, F. He., "~Research Investigation of %rdraulic Pulsation Concepts".. Republic
Aviation corp.,. February 29, 196A (AD 431304).
This is the Fourth Quarterly Progress Report under Contract AP 33(656)-10622.
During this period, system efficiency and transais sian-line-loss metho&3 were
derived. Experimental system efficiency and line-loss data were secuvtl from the
miniaturized system. Designs of a diaphram- and nechanical-type transformer were
studied. Syrstem-failure and functional-effect anielyses were made. Also, a
preliminary glossary of terms peculiar to p-.lsating tvdraulics was comp:-Eled.

290.
Poney, Go, and lkrmach, Xo, "Sose Relationships Between Coefficients of Anisotropy,
Work Hardening, and Deep-Drawing Tests", Rev. de *--t., 61, October 196k~ (in French).
Tests were made on extra mild steels, light alloys, and other interialr 'ihich
*can be found in the form ofasheet or strip.

77
 Porter, R. N., and Stanford, R. D., "Propellant Zxqplsir4 i1nined- BWp~caft*-
SAX-ASEB Paper 8688, April 27-30, 1964.

3l3add rs, diaphragms, and pistons used for the poit1v'e expal!ion 9f e..$rsoble
liquid rocket propellantsa are discussed In gineral terms., The histotr of,JK.L' a
work on these devices is reviewd as a background to the current, pro"r'em A
detailed account of the deve1qpent and ufe of bladders in.ae 'and iriner
spacearaft is presented. The final aectikq describes an advanced d~ieAi
program aimed at providing tecbnoloa for fture spacecraft.

"Pressure Measuring Techniques Feasibility Ivalustion", Instrumentation Div. of

Gulton Industries, Inc., 132pp, ul 1966-Noveber 1961.
Based upon the results obtained &.ring an Investigation into the fabrication of a
ling pickoff approach appears thaean vpper practical eau lmtO

I approxiately 1000 7. Although t mnsducers were constructed vihoeae o10

F, the data obtained above 1000 F were erratic: and showed insufficient promise to
extend the usable range above this teqieratuwe at this time. Should the state of
the mterials art advance In the near future, particularly with respect to ceramic
cements, film insulations, strone ductile mtals and temperature-coensation
coonents, then the feasibility of constructing a transducer using the variiible
coupling pickoff -anddiaphragm presare-elemnt technique will he considerably
enhanced.

W "Pressure Transducing and Instrumntation Techniques", , Pk1,

& Giannini Controls
Corp., Duarte,. Calif., August 1, 1960 (AD 251111).

An engineering study and evalution of pressure transducing techniques and re-

lated transmitting and Indicating system techniques is presented In two volumes
of four books. Vol. I, consisting of Am 1, 2, andi 3, gives a general Introduction
and coves the current state of the art. Vol. II includes Improvements, miniaturi-
zat ion and new techniques, In addition to a reference section. Covered are pressure
ranges from 50 to 5000 psi and -eq-eratu- ranges from -300 to 1000 P for rocket
fuels and oxidizers, aircraft fuels,, exhaust products, engine lubricants, and lby-
&raalic fluids. Transducers, convertn pressure into displacement, and transmitters,
converting transducer displacement into a suitable electrical voltage, are covered
individiully and in comintion as to materials, design considerations, manufac-

Iacceleration
turing, performance, and limitations. Information on size, veight, input vs outpzit
characteristics, accuracy, repea 'tllity, reliability, and life is give. Dis-
cussed are effects of and cosen .on for corrosive media, tqweature,. vibration,
shock,, andi nuclear r. _iation. Applications, recamnded usage,. and
performance Imrovements, advantages and disadvantages, are cited. Muh of the
Information is presented In the form of illustrations and tabulated data.
294-

~"Production of RIO-Strength Bellows", Metal Ind... U (21), pp 4.23-4I24, Nov. 18, 1960.
Brief descriptions ars given of techniques of smking bellows of Moral and stain-
less steel.
78
 ~~takoiski P. Pp et a.. Rmr3 Analysin of Zrquationo of Thin Mells or
Revolution", IA. Rocket S0c. ZI., ~,PP 3644l,1962.
A numerical analysis is given for the solution of the general equatioa.* ai Oa
shells of revolution sub~ected to rotationalily s~ametrio preise an - M
distributions.* The basic differential equations are in a very general ftCm ~~
Permits the geomtry of the shells considered to be specified by dicet a
points. The analysis dateraizie elastic stresses, strains, Man dispaeni~
aultilayer and-multisectiomi. shells of revolution. &uraee loads, te -istlw
thicknesses, and material properties my vary arbitrarily in the wsz±4 onaldi
tion. Temeratures and material properties can also vary throngh Aie thikniu.
The solution is obtained by direct computation using a. nwerital metho& thnat
eMplors two by two boefficient ustrices and hemes avolds th - tzpLv or 4,01k
convergence.* The solution bas been programmed tn a semi-al~ebrztla laug tbit.
can be used on most high-speed caipit~rkt. Comprisons of umrical ada ion !to
known exact wAd approximate solutions -of the thin-shell equtions arc iiA6 to
demonstrate the accuracy of this method,

fta!lph X. Parsons CoQ, Teat Report,. *Results.. Prequlificatin, TLest. for FJexilb:e
*Mtal, Wire-Reinforced Rose%, ftbruary 23., 11TD28371.2).

Tests n~f 1/2, 1-1/a, and 4-ineh flexibli-mtal hose assablies Imluded visti
exaination, proof presve (150 psig), mim=w deflection. cycling (10 Cycles),
and flow (M2 ga).

2fRaymond, R. H., Prope Pre-Weld Practice Produces better Aliininu

CrYogeic
hllo-is", Nod. NMetals, 12,p 4i6, Nvveber 1963.
The never weldable allos enhance aluminua' a competitive position for crygea
bello-,s and ductingv but rigorous control of fohrication - especially welding -
is needed.

f flenerS, ad anmn, F,
pp 6T-72, NOVy 1965.
"ftperimntal Stress Analysis%, Ind. Re;, , (3),

With the availability of easily worked photoelagtic: plastics arA ecoatings RMAM
efficient and laree boom of poleiwsed-l1Aght, in addition to the ft"eloM1t Of
Thrious types of strafn-masrig devices and JbIre-effect, telaheiq use erperi-
mental analysis can be used to obtain a qjuick and thor*Oui w drAMUMin Of the
straits behavior of a structure.

79
 Reissner, I., "On~ Stresses and Deformtions in TVtoi~1 Sbefs of Circular-
Cross Section Which are Acted Upon byr Mhiform Norml1re**ure*w, Q4A, t
et., (3), PP ry717 Otber 19634.
The present objeet is a mome general approach to the-problei thropo a sytem of
d!~fferential equations wihich contain both the equaticims of the linear-ber&tag
*th-oory amd of thbe nonlnear- membrane theorya Unliiting came., ind which rewilns
appllcable In the transition region when both linear -banding am& nonlinear m-
braze attioni have to be considered simltovsly. &§eeitiealUj, the obJective
*is to determine the ranges of values of suitable nondImenuional parameters for
which linear benftng tteory and non3inear-emraa theory r" appropriate, and
also thei ranges of these psrareters for which the prcb3.ei belcs to the- tranai-
tiorn region between- the two- liniting forms of the theory. Derivation of the. differ-
* ential equations which govern the problem 5n all three rangev s seagai~lshed
through appropriate speialization of a general vyste2 of differential equations
for finite syinetrical deflee ions of shells of revolution which has previously
been given by the author.

~-ic
Reissner, E., "Oa the Theo-ry of Thin Elastic Shells' R. Reissumr Amiversarr
Vo1uae, pp 231-247, 1949.

This paper is conc~erned with the subject of rotationally symmetric deformation& of

thin elastic shells of revolution. First a aelf-contained fawiUltion of the
rcb*1em of finite symetrical dea~ections of shells of revolution is given. From
this the equations of the wwel1-deflection (linearized) theory are obtained by
spec ialization. From the gereral equations of the smll-eflectina theory in
a eysteatic manner a aimplified systemt of equati"n is obtained which applies
for shalloj. shells. It is shown that the solution. of this systen of equaticas
can be expressed in terms of Poessel funtims for the ettire eaw of paraboloidal
shells of eonstant thickness. This generslizes kncon results for
the case of a sbaillow zpherical &bell for which the neridian curve Is equivalent
to a second-degree parabola. It in also shiaun that the solution can be given
in terms of elementary functions for a class of sballw shells with vsrying
thickness, such that the prceblem of conical shells with linearly varying thick-
ness is includ-rd 9L,, a spacial case.

Reismer, RL, "Etstioally S6,ntric ?roblem in the Theory of Thin Elastic

Sbelle"3 Proc. of the Third U.S. ht2L. Congress of Apl. YAech.., ASS,, 1958.
- I General problems in the theory of thin elastic MWell reqlire the determination of
stresaes and deformationis as a fiMition of two space toar&inkties. Rotationaly
aqw.ietrie probleas have the' property that stresses and d-farnations deped on one
epece coom-dinate only. This mans that, for tive-Independnt problems, one is
concerned with ordinarv differential eqcations rather vton vwIth partial differential

I8
 equatiz.. Thais paper reviews a numer of "Ims and solutions in the fild~w
of rotationally symntric deformations of thin shells, all of them deeling
vith Shells of revolution vhich are the aost natuxal souce of such procblems.
* Consideration is given to linear and nonlinear prolem of the statics-of
shells of revolution, vith particular empasis on asymptotic solutions and
edge effects. In addition, the prolem of banding of pressurized curred tubes
is formulated in consideftble generality.

10Rekate, H. L.,. and Schvartz,, J. I., *Behavior Stud of a Curtiss-Wright Sorkel

Flexible ERchauzst Cooection for Submarinme Application", U.S. ]kuy Marine Ing.
Lab, Noeber 13, 1963 (AD 4~23003.
The vibration endurance and shock characteristics of a flexible comxxetioa for
a Curtiss-Wright engine exhom system were Investigated and fammd to be sat is-
factory. Frca previous investigations, it had been found that Stainless teel
is a mterisi for this flexible exhaust; Inconel should be considered. This
report presents a proposed aethod for evaluation of ft~ture dasigm of flexible
Connectionls.

303- 9
Rothfuss, 1. V., et al., *Design ad Application of Metaic flezurtS for
Bejuipsient With Specific Life Requiraeets", SA-ASW Paper BT1.A. April 27-30,.
* 1964.

This pape treats the design, application, variations, and combinations of fcur
basic types of metallic flexures whbich do not have Sliding or rolling load cwrY-
iog contacting sufaces - flat springs, round bars, curved beas, and diaPliragas,
Flat springs develop into flexural pivots, used to replace conventional OSCinatin3
bearings in equimnt sixeh as gyros, gimbal rirgs, Wn linkages of all typs.
Round bars Aeveloq Into toirsion bars to replace conventional bearints lth verY
limited torsional moenwt sAnt quill shafts used as flexible shgaftr- with l±2-Ited
angular movement.* Curved beasts de'velop into the &imW ring and the curved
beest lattice. Diaphragm develop into specially shaped Zontoured disphres
for torsional Load carrying capability at high angglar aisallgmnts. ftb
j ~basic kinematics, the eqymtloas for dtriigtheir
in' aem aplicitims mrs also discussed.
aplroximte Size# and

Salmarmm, F.,' Wcosopane of Canrugated Eansion 301*5w, PP In-,

*I1l0,Morc-h 1 9i16 (Translated frois Schweielsebe BusetunC)
In this paper the comlianice of corrugated pipes v IIr the influenc Of foraes
in the pipe axis directioai Is studied, with full aCCO~t taken of hoop stress
influence. Corrugated shapes copoed of straight lines and arcs are studied;
the most favorable conditions vith regard to coqliance arm found. ftbe-
qutntly the behavior of bending loading is shown for individual Cas-

81
 "Seamless Metal Bellows Behave as Srin-Loaed, Frictionless isoshe
-

Subjected to Pressure", Design News, 14 (16) pp 18-19, August 1959.

The characteristics and applications of seamless metallic bellows are described
briefly.

Selbel, M.P.L., "Differential Pressure Gage", Electroechanical Design, (9),

pp 64-65, September 1963.
The precsure sensing element is a sprirg-loaded bellows located in the pressure
housing. High pressure admitted through the high-pressure port acts upon the
outside of the bellows, while lower pressure admitted through other ports
acts upon the inside of the bellows. The bellows is compressed by the effect
of the net pressure difference but the absolute-pressure level has no effect
upon the bellows. Compression of the bellows moves a magnet along a bore in-
side the pressure housing and this movement is picked up by two follower ag-
nets, outside the pressure housing attached to a light pivoted yoke carrying
the pointer.

30T.
Seide-, P., "The Effect of Pressure on the Bending Characteristics of an Actuator
System", AM Trans. 2, pp 42 9 -43 T, 1960.
The effect of internal or external pressure on the be of a cantilevered
bellis, the movable end of which is permitted only to r te about a fixed
point on the longitudinal axis of the bea, is investigated. is found that
the bending characteristics of the system vary considerably with 5 in
ssesure and pivot-point location and that instability of the system my
The experimental results given tend to confirm the behavior predicted by the
theory.

Sapotoski, W. K., k al., "A Digital Computer Progrea for the General Axially
Symet-ic Ti-&Sbell Problem", AM Trans., 84, pp 655-661, 196Q.
This p e describes the development of a general comuter program to handle
arbitrary thin shells of revolution subject to radially symetric loading or
temperature variation. An elimination method is used to solve the set of
difference equations obtained from the basic differential equations; a feature
of the mthod is that "edge effect" difficulties that can arise with conventional
diffe'ntial-equation routinee are avoided. The program is quite flexible
and permits discontinuities in shell geometry or loading. The results of
app4ing the program to several classical problems of known solution are given.
These results permit the examination of coptational accuracy for varying
bouadary conditions and mesh sizes. Finally, se program solutions of umcon-
ventional problems are presented.

,,82
 30.Serensen, S. V., and Snneiderovich, R. M., "on the Investigation
of the Stress-
State and Strength for Blattic-Plastic Cyclic Deformations",. Ian ONl SSSI,
kMekh. i Mash.: 4., PP 136-l140, 1961 (in Russian).
This paper considers the dependence of the stress-strain behavior ex. a miter-
* lal on the number of cycles through which the specimen is deformed.

310.
Shavki, G.S.A., "Assesa~.ng Deep-Dlrawing Qualities of Sheet, I..Stretch-Forming
and Wedge-flraving Tests", Sheet AEetal Ind.., & pp 363-368,9 Nay 1965.
This paper presents a criticel review of the methods of assessing deep-
drawing qualities &&~sheet mtal developed during the last 50 years, and
discusses stretch-forming, wedge-drftw1Wg, deep-drawing, and combined simuila-
tive tests.

Shield, R. T., and IL-ucker, D. C.,, "~Limit Strmngth of Thin Walled Pressure
Vessels With an AM~ Standard Torlapherical Smed, Piroc. of the Third U.S.
Natl. Congress of Appl. Mech., AM, pp 665-667, i958.

Results are presented for the mximm preawxre which an unfired vessel with
an ASlE stand~ard torispherical head can withasad before appreciable Plastic
deformation will occur. The effect of~ flanges on strength is considered.
Although the procedures outli.:ed permit variation of pressure in the axiat
direction to be taken into account fully, for convenience the calculations
are based upon the asmmiption of uniform. interior pressure. Work-hardtning
Is not of much significaace for thin-walled vessels of the usual steels and
is ignored. A comparison is made with the AMR Code for unfired pressure
vessels. A real danger it pointed out which is especially important in con-
nection. with brittle fracture at low temperature.

312.
Smith.. A. H.., and )acDonald, W. R., "Aliasing Errors in A. C. Bridge Trans-
ducer Measurements",. Royal Aircraft Establishment, Farmborough, 9 pp,
December 1963 (AD 438804).
A pressure-sensing element (usually a diaphragm, bellows or capsule) can,
under mechanical excitation, resonate at a frequency considrab3y higher than
the cut-off frequency of the acoustical sVstes controlling the response to
* pressure changes. Trouables of thin nature can be avoided by supporting the
transducer on suitable anti-vibration mounts.

83
 Smith, 1. X.., "Analysis of Creep in Cylindirs, Spheres, wAn Tiin -blocs
J. M~ech. Rag. sci., y (1), pp 82-92, 1965.
*Amethod of analyzing the creep behavior of cylinders, apheres, and thin discs
is described in which the effects of time variation of the enyfronntal con-
dit ions of temperature, pressure, rotational speed and internal heat generation
* may be accommodated. Deformation is considered as a series of steps each
consisting of a short period of steady state creep followed by instantaneous
stress readjustment to resatisfy the equations of equillbriu2 and ccati#blit-_
The relationships for stress readjustment can be written down once for all,
since they are independent of the type of creep law used and of the length of
time interval taken. The simultaneous solution of these relationships is
presented in a form suitable for incorPration in computer progams.

314.
Standards of the Expansion Joint Ifnufacturers Association. 2nd Edition,
1962, New York.
7he previous edition, published in 1958, vas the fiest attempt to set forth
recomnded, standards, consistent with safety and service conditions, for the
mechanical design and application of all the various types of PAokless Ex-
pansion Joints, regardless of bellows design and construction. The major
objectives of this first edition were to clarify the conflicting and ambiguous
terminology then in use in the industry and to establish and maintain high
standards of quality that would ensure Expansion-Joint users of long-lived,
* trouble-free equipment. Since this second edition was greatly expanded in
* scope, particularly vith respect to complex piping configurations involving
lateral deflection and angular rotation, it was hoped that its use asa
specification mdiu would increase.

Stange L., "Strains In an Annularly Corrugated Diapbrsgm", 1g.-Archly,

2.2 pp 47-9l, 1931 (in German).
This paper gives an approximate solution of a toroidal aegant In term of
a power series in the angle 0 between the axis of symtry and the Dozmal to
the shell. It is assumed that sin 0 < 1 so that the solution is only appli-
cable to torcidal segments of the type found in diapbreaps.

IStedman, C. K., "The Characteristics of Flat Annular D1MPhr8Ms", Statimn

Laboratories Instrument Notets, I pp, January 1957.
A good deal of inforation, both theoretical and experimental, concernlng
flat-plate diapbrama has been published. But it is somwhat scattered, and
mostly pertains to diaphragm without central reinforcement. For this reason
the equations for the rigid annulus,, which is a considerably mme versatile
form, are presented In this issue of Instrument Notes with the thought that
its advantages might be more widely exploited it design Information were
auk readily available.

85-- -
 ~''itdw~,G. E. 'Netal. Deliws Involve Unusl Metaworking Operatiosi
!ihefr Mnufactur~ing%, Steel, l 1 i, pp il6-118, April 3, l941I.
Ibis a~rticle describes processes for making bydraulic and spin-rolled-attl,

318.
Steele.. C. B., "Shells or Revolution Withi Edge Loads of Rapid Circ-unferential
Variation' (J. Appl. Webh.) A.%2K Trans., 84r PP 701-70T7, 19We.
After an investigation of membrane theory, solutions to the general equations
for * thir shell of revolution of arbitrary meridian are obtained for rapidly
vary-Ing sinusoidal edge loading. The solutions from memrane theory are
shown to be valid for shells near spherical sbWp but can give quite misleading
results, particularly for shells of negative curvature. The "edge-effect"
solutions also are showrn to become significantly modified for the high harumics.

Steele, C. R., and Mrtung, R. F., "Symtric Loading of Otbotropic Obells

ol Revc Lution", Lockheed Missiles Tech. Rept., March 19&1.

:n this report, the prcblen of determining the stresses and defornstions In

I
tihooeeu,
Om orthotropic shell of revolution under the action of azi-
symetric loads is reduced to the solution of a single inhomagaeous second-
order linear differential equation with a complex dependent variable. AM-
totic solutions are obtained which are uniformly valid In both the steep and
g-hallonw regions of the dome-shaped shell. The comlementary, or medge-
effect" solutions are expressed in terms of 02bossjtg functions of a non-
integer order. The order depends both on the shape of the meridian curve
at the apex of the shell and on the ratio of the elastic moduli. The parti-
calar solution in found In terms of an appropriate linear comiation of
Lomml' a function and Thomson'sa functions. This particular solution is equiva-
lent to the well-known "memrane" solution in the steep portion of the shell,
but in the shallov portion gives significant bending stresses. The particular
and zooplementary solutions are used to investigate the behavior of ortbotropic
pressure vessels with rigid rings clmed to the edges.

Steele, C. R.., "Toroidal Shells With Nonsyetric Loading", Thesis, tanford

University, 96 pp, 196D.
The eighth-order partial differential equations for a sbell of revolution
reduce for sinusoidal edge loading to an eighth-order ordinary differential
my'stes - three sla neu equations. These proved unield4 for the told
and so vee reduced In this work to one nonhomogeneous integral-differentia
equation which can be further modified to a form quite similar, exeept for

j the integral, to the fourth-order formulation of the axially symtric Prcblam

~treated previously. Asyzqptotic form of the four solutions of the hIoeeu

86
 equation for all. harnonics vere found to coincide with the bag oscill,&ar'
"bending" or "edge-effect" s~lutions of the 4mrc problem beause the
harmonic Index n did not appear in the term thtpiroyd to be lioitaat.
Solutions of the nonhcaogeneous equation coincide with ze~rane and SIwi.-
tensional deformation solutions in the nonsW Ilow regimes of the shell, so
these have been investigated and their copttion W1qrn
bt re to
the nmerical, solution of two simple equations and sowe~ia ntgIin
Then using a geea Iton of the technique of A. A. Clark, a~ipt~tie solut-
tions of the nodoageneous equations were found In term of tbAl anae aad
inaxtentiocal deformation solutions and the f~nton used by Clark. e
results of this iziveitigation are limited to thin shela. and ulowly waryl
loads. Criteria are given for both limitations.

I Stein, X., and WZLmnp Jo A., "Bakling of Segeents of 2broidal

AIM 2nd Aerospace Sci. Meting, January 25-2T, 1965 (AIM Pape N06.
Nonlinear differential equations of equilibriu and b=W In equations we
pel~
-5TT).4

derived for segments of toroidal shells nar the equator and-for semnts'near
the crown. The equations are deived for shallow-shell saeats by Including
appropriate prescribed Initial displacemPntvin the nonlinear, flt plate,
strain-displacement equations and by varying the total potential enery of
the system. Closed form solutions to the buckling equations are obtained
for sizply supported segmnts near the equator havift eite positive or:
negative Gaussian curvature une prese load-Ing with 'various Irelons ffujfort
conditions. Results are presented. In the 'arm of charts uawing buckling co-
efficient, as a function of a curvatuwe peamter associated with the girth of
the shell and a parmter associated with the ratio of 1rindipal, cumvatmivs.
In mM Instances the results Indicate significant deviations In buckling
stress for the toroidal shells over the bucklng stress for the correspoedin
circular cylindrical shell simila loading and siport- conditions.

Stevenson,. F. D., and Wicks, Co lo, "AMetal Disphaga Atpratus for Measuring
Vapor Pressiwes Vapor Presure of Arenic (MI) Oiddeop U.S. D~ept. of
Interior, Bueau of Nines, Dept. of Invstigations 6M2, 1963 (163-14n25).
A mtal diaphraga presue relay system was constructed to asae vapor
pressures of the metal halides, o3hmldesp and other m~stances at tempera-
twoes up to T00 or 800 C.

Stippes, N, and Beckett, R. lop -Otrically Loaded Circular Plates",

. Franklin Inst., 2[, PP WA5-T9, Jun 19541.
A solution is derived for the nonlinear deflection of a circular plate which
the authors feel is somewhat more general than Way's solution.A

8T
 Stricklin, J. A., et al., 'LUrge Xlasti*n, Plastic, and Creep Defle-ations of
Beams and Axisynaetric Shells". AXAA Journal, 2 (9), PP 1613-1620.. Se-tembe
196I4.
A numerical aethod is presented for ansiyzizig large deflettons of curved
beans and large axisymmetric deflections of shells of revolution. The govorn-
ing equations that are in finite-differrence form are solved by a Newton-
Raphacci iteration procedure. The plastic etress-strain relations are deter-
mined by ass1uing three independent slip planes that are the planes of iiaximuiA
shear stresses. The stress-strain relation along each slip plane is assumed
to be linearly strain bardexvIng. Thin plasticity mod-al gives pieecewise-
linear plasticity .relations with determinate coefficients. For snap buckling
problex3, the so-called upper critical load is determined by introucing an
artificial spring opposite to the applied load. The true load is then the
difference between the applied load and the reaction of the spring. Tne
Important mathematical consequence of introducing the spring is to yield -

single-valued load-deflection curve, thus allowing the critical loads to be

determined. The method is applied to several problems inclujing low arches,
circular ringg, shallow and deep spherical. shells under elastic-plastic
deformations, and a shmLl1ow arch under creep deformation.

Strcuer, P. R., "Structural Stability of Toroidal Sbells", Lockheed Missiles

and Spae Ccmpav, June 1963 (AD 423823).
Toe toroidal shape is of particular Interest in space applications as an
optimumi design for space vehicle liquid storage containers and for inflatable
or collapsible structures. Torus dasigns have also been uned by nuclear

I
physicists to confine atomic fusion reactions in particle acceleratorM Both
linear and nonlinear methods of At:esa snalyzis hmve been developed to stud~f
the structural stability of these shells; however, no reference to buckling
behavior of toroids was located in this review of the literature covering
the period January 1948 to June 1963. A aubject in&dex Is included.

2-zOtton, G. P., -Charts for Circular Diaphragm Design; Referenaze Book She-et",
Prod. ]tog*, 18, pp 16T, 169,9 January 194T9
Charts which are presented provide design information fcc circular thin
diapbragas, of uniform thickness and with fitced edges, subjected to a lani-
foraly distributedl load. Chart A is used to determine stresses and Chart
B is used to sterzq.ine deflections -for a givea xAterial and given dimensions.

88
1 ~rindammn, R. W.,j and Douglas,* D. A.., "The Failure of Struatural Metals
"iJected to Strain-Cycling Conditions".. AME Pmpe No. 5&-A-198, .1958a
Data showing the isothermal strain-cycling capacity of tbrime utals, Incocelp
HastellcW B, and beryllium are presented. It Is noted that at frequenciqs -of
0.5 cycle per min. the data satisfied an equation of the form A.- J4 M4"~
N is the number of cycles to failure, cp is the plastic strain per cycle,, and
a and K wre cobatantc. whose values dp-do h aeiladts odtos
DtonImconel are given to establish teeffect of grain size, specimen

coarse-grained material. Long time cycles shorten the number of cycles to

failure when the strain per cycle is low. Therma-strain-cycliag data for
Incozasl are caqared to strain-cycling data at the um mean tqeaue
Good correlation is found to exist between the two types of tests.

328.
--- Tao, L~. N., "On Toroidal Shells", J. Math. Phys., , pp 13 0-1 3 ii, JTuly 1959.
Since the classical work of H. Beissner, the determination of the elastic
deformtion or shells of revolution under aiqm tric loading has been the
subject of nmrous Investigations. For toroidal shells the problem. has been
studied by Wissler and Chong using power-series solutions,, and by Chang and
Clark by mans of asyntatic Integrations. It in the pcrse of this note to
establish the solution In closed form, expressed in terms of a coqipratily
unknomn function, the Heum function. Though the present solution is still In
the form of a power series, it has sm advantages over the previous solutions
obtained by Wissler and Chong.

~la-aVenvwli, J. F., and Coffin, L. F., Jr., "hpermetal Siqpport for Go~salized
Eq~uation Predicting Low Cycle Fati ue", ran3. AME 8&D, 4p pp 533-51p
Demimr 1962.
A siimle equation proposed by Langer and based on relationships found by Coffin
ha's been comared with measured values from 12 published fatigue tests with
dIfferent steels, copper, nickel, titanium,, and alueim alloyrs. Equation
gives stress amplitude of low-cycle fatigue 8 /2 r/2+ Be whesre Be in
rzn&zanae limit, I modulus of elasticity, N cycles to failure and c = 3,124f
-1/2 In (A/A 3),. where ef is fracture dutility (from. static tensile test),
A0 Initial area,, A final area. In most cases equation gives conservative
values for low cycles, due to basing constant c on static fracture dutility.

"Test Sjysten Anaalysis Study for Propulsion Research Enironmntal Chmermp

Re search Managemnt Assoc., Noveber 23, 1963 (AD iI2lo~o).
The feasibility of developing a facility for the testing of rocket propulsion
systems eqplcing high energy, toxic, and kWpegolic propellants under simulated

89
"wae environmental conditions has been establiabed through mnlytical StU11011.
with a cocmbination of double-valled vw,= cb6Wer and an ekhauster myotst
separated by an imerueable flexible metallic interface that allows full
giabolling of the rocket motor.

Thoxpaca, L. M.., 'Velded Metal Bellows, A Reliable Positive 2xpulsi00 Device

ffor Liquid Propellants", /AM Paper No. &)-26k, Bell Acrosystems co.
The potential of the welded metal bellow device was explored during a
design study prograw by Bell1 Aerosysteas Cam In March 1963 for NASA, under
conntract NAS-3)I9. This progan was conducted to evaluate metallic positive
eipulsion doyices and to select the device having the greatest potential for
inamed application. In the initial phase of this program, an industry-vide
survey was conducted to ascertain the state of the art in develotnt of Wll
types of expulsion devices. The survey also included a reviev of the present
status of our rtechnology, where it applied, and the monit~ing of prora
currenrtly in progress at Bell Aerosystems Company. Phase I of NA-l19 was
completed with a definitive design study and parametric analysis of the three
most prumising metllia devices for a cylindrical envelope: the piston,
metal bladder, and velded metal bellows. The results of this parametric
aalsysis deaostrated the superior potential of the metal bellas and this
device "as selected for a compieensive design study in Phase n.

Thomson, A. 1. , "Rupture-Proof Bellows-Type Orifice Meters", Instruments,

K2 (I), pp -1593-1596, J16yesber 1952.
Iqxortat factors associated with rupture-k-cof duel-bellovs flomeeters incld
bellows construction, temperature comiensation. range springs, easie of ser-
vicing, freedou from~ over-range damage, high sensitivity,, simple installation,

and applicabilityr over wide ranges of differential and static pressufes.

ThcuUsn 1. G., Yang, C. T., and Kbayashi, S., Nechans of Plastic 1efmft-
tion In Metal ProcessiN, Nmcaifllan Co.. Neu' York, l._

The authors discuss the presently known setlaods; of solution to some of the
problems of metal forming obtained &cm principles of plasticity. Cmpri-
son of solutions with experimental data has beon mae where possible. Sub-
jects covered include buckling, necking, fracture mechanics, and slip-
line solutilons as applied to the processes of forging, extruding of solid and
hollow fors, coining, rolling, spinning, 90cIning, and drawing of vime and
tubing. Mutch recent information is presented that is not otherwise eas811Y
accessible.

90
I-Thoason, J. H., "Torsion Bar Pressure Transdner."
8(6), pp 46-5o, June .t6i&.
Electromechanical Deasia,

The torsion-ba Pressure transducer is quite si pe. Pressure applied at

the pressure tap applies a force on the arm which twists the torsion bar
slightly. The evacuated bellows provide an absolute reference and a mcehahi-
Cal balance. Bellows of the best available welded type madt of AiS 350 steel
were selected for inherent uniformity of the effective area, good high-tes-
perature characteristics, and good spring characteristics.

Timms, R. J., "Analysis of Toroiual Shells", Thesis, Northwestern University,

1953.
7his dissertation gives a generalization of Clark's a&!*ptotic solution of
toroidal shells to make the results to a wider variety of cases. An attept
was made to develop siaple design formulas to be used under specific
conditions.

336.
Tales, G. E., "-Water-FormedBig Exnnsion Joint for ASCL Reattor", Design
Eng., 2 (8), p 4 August 1963.
An expansion Joint was made of 304L stainless steel, 1/4 inch thick. The
* diameter of each end was 104 inches. It was decided to have the expansion
Joint bdraulical y formed rather than roll-formed (which would have been
far simpler) because the former method eliminated surface blemishes likely
to result from roll-forming. It also was considered iesirable in order to
achieve greater uniformity of metal.

337.
Tumarkin, S. A., "Analysis of Symaetrically
Loaded Toroidal Shells With
the Aid of Trigonom-etrical Series", Priki. Mat. i Yekh., 16 (M), pp 569-574,
1952 (inRussian).
Tn-h torbidAL-bellcs pa blem Is solved in terms of trigonometric series in
the angle f between the noral and the axis of sjetry. A toroidal be11ois
with axial loads is solved.

4 -MT'arkin, S. A., "Asymptotic Solution of a Linear Nonbhowmgeneous Se.ond

Order Differential Equation With a Transition Point and its Application to
* the Computations of Toroidal Shells and Propeller Blades, Appl. Math. W.eh.,
* ~ &9 pp 1509-1565, 1959.
This is a theoretical derivation of the aiymptotic solution for the toroidal-
shell equations. Estimates are obtained for the errors in the solutions.
One solution is obtained for a toroidal shell and compared with Clark's

91
 first roms; Tumrkin did not see to be aware of smof Clark's Uat~r
papers vtdich Contlneai te swn reselts that were given in this paper

0re.
. I.. "Stress and Defletioa Studies of Flat-Plate and Toroidal
~neoii3e~lws Sa~4cte toAxilEccentric, or Interma Porem Loag", 3

J. Wth. &rAm. Sai., I t~2).p pp, 1o-1A3,. 3pepbe 1959.

Tets !a axial cacpression eccentri* (hinge) loading and internal presvae
are dev-ribed for corrugateti-pipe ant" flat-plate bellows. Strain gage swV~r
and de-OfILection readings were nade for each teat. A me=Uv 3eveloped theory
for the mmll elas3tic deflectioni of toroidal. eleints is briefly doscribed
and its arplication to a wide ranp of bellovs shape& an& loadings; is dis-
cussed. Pw some flat-pgate belicars, 'vaere the depth of ccrtolutice So
auch smller than the over-all raius of the bellows, an -3pproximte thecry
neglecting circ-9*rtul1al effects is possible, and comarison is ad beweea
this approximth and the- more exact tUnary. Both thtoy and erperimt
show I.hat the J=&t ion of the maxium strisses depends markedly upon the
proportions of the beflm. General baflovic design and construction prob-
leas are exsmIned in the ligbt of the stresses likely to be caused by 6actumi
service conditions of operation.

?urner, C. I., -Study of the Bpmtric Elastic Loading of Som Shlls of

Revolution With Special Reference to Toroidal Eleowtsft, J. Mach. Ing. scl.,
(22113-329, etme 1959.
Tb overning equations for the elastic behavior of aximtically loded
thn 4atic shells of revolution are derived and applied particularly to
toroidal shells were the radius of revo: rtion is mch larger than the radius

I of ring cross-section. It is shown that ime these conditions the apjtozi-

inato equations are a special ase of the Mathieu type. Som solutions have
recently been tabulated enabling the theory to be applied to structutwes

IComarison
such as toroidal pressme-vesael heads, expansion bellows and s&mil1 *hoes*
is made with an appinte series solution of the sieequations.

Turner,, C. I., and Ford, N., "Stress and Deflection Studies, of Pipeline
Iexpnsion Dellovsw 2 Proc. inst. of mact. mm., My (15), pp 526-552, 195T.

In the design of pipelines for high-temperati service, allowane has to be

maefor the therm expansion of the heated pipe. On particular aspect of
this probla is the mse of bellOws expansio Joint$ which, particularly on
board ship, an transmit the pipe movemunts throqai bUlkheads while providing
a vatertight flexible anchorage. An. Moxinte theory has bma~ developed
for the compression of a bellows, the convolutions of which hew -- cross
section forimed by circular ares sUbtending &a sMai-angle a- Airima1 results

92
 Il
fALS?been e.alcultcd for the two cames a W/2 and of 3044. es
have been eazried ouit on six bellovs, four earuted-pipe tyZ' j* 2
oatre
Of (y ,r/j *Z- on fltplt wa
WM nVIOUL7
4dich

0414. Delton and resistance straln.g g readings V.? tjjak n

bellows,. anti reaxovaabie sgreemstit was f *qwp between the theoretical a a--eiperi-
Ae3tal resulte for the five beflotm In Alicz these couzld be cosprtd, ;%has
been touiuw that for tertaft design conditions optimau re atisonhl" exist
between bofte, wall, thiekness., and radimis of con*6btioa for the mxim
flexibgity, Stress intezaificationaend flexibility factors bave be" cale-
culated to cover a rangs of bellows prcxWbions.

Zxmr, H. M*, "D~slga Parameters for Lliptleal Tooidal Pftesure Vessls,

AeaeZn. 21 (n), PP 33-38, Noveber 1962.
A preentation Is ade of a desigm parameter curm ti ollipt-Ital toroXi
press-,re ve-suels in r-ondinmnicn1al fo-.a, relatIng the radius Cf rotation
aboeft an axis to the ratio of the ueiias of the ellipse to mintainz tfn-
aimn tbrouviouwt the press mwe vessel.

ZTVeeddsle, J. 0., The Ikoenizcal ftgpeties of Mertals, George 1114m sn4 Wriin,
1 .
Ltd&~ Loxdon, 19W

This book cowers the field of the assessment of sechanical propeties at

engineering mterials relative to desig and mrnufteture, theoretical and
practical limitations on properties, ditflcultit* in analyzing wanwecor4-
iunproperties, avid problems asoiated vith th testing of mterials and
t4pes of testing machines.

M.-Un~jD, B., "Determination of the Jtural Frequenaies of MXratio in the

Breeze Bellows", Moolls Atomic Poser Lab. Rept. K[APL4IXS-i Ilwah ks l55k
?einvestigation of the resonant frequencies of the Sme belloWs v~iCh
weeused In the 83 Mark A 8-in. stop valves vas continued. Tn effects of
the excitin displacmuts and Wese differentials upon the u ta
and haronc resonant frequencies of the extended and casiesmed beoow
were observed. Also, the effects of transverse restraint awl of fluid in
the beflovs u~pon the resonant frequtacies vere Observed.

93
~Vnder IPylp L. X., Bibliograp!y oa Diaphragms and Averoids 3 , AM Pawe
6o-wA- 196o. ",
A Literative Srvey. Caittee was created by the Research Diap*ra 3b-
cwittee of the Research Cowittee on Weanical Pressure 1lements to accn-
late a list of the outstanding papers and literature on "MDiapraa and
Aneroids%. This list vas t, include sany of the papers listed in the biblio-
grapW on *Diapbragms and Aneroidsw published previouay (AS Paper No. 55-
A-_ by G. f. Lee and L. N. Van der Pyl) with the addition of sw missing
papers and some later pazers. The 171 abstracts offer an account of the
available literature in this field. The abstracts are arranged alphabeti-
calUy under the name of the author or senior author. A chromological ind-x
is also given.

Volkov, A. N., "Contact Problea of a Cabined Toroidal &ell With Ring

Plates, Inh. ZA., I (2), pp 331-336, 1963 (3%-28k83) (translation).

This work perfarns an analysis of a system of equations expressing conditions

for contact of an annular plate with a toroidal shell and describes boundary
conditions in connection with the design of corrugated shells for bellows.

Volkov, A. N., "NDternination of the Axial Rigidity of CQctuated Mell-

With Befereme to Calculation of Bellow, lath. ZhL., 2 (2), pp 38-37 ,
1963 (in Russian).
This note presents a simple method for predicting t.w ia a rigdit
of a bellow in terms of a strip of unit w htbby the wetbod of
Castigliano (ber theory).

Volkova A. N., *Investigations of the 9tressed State of Toroidal Saell s ,

lah. Zb., 2 (M), PP 312-30, l9 (in Russian).
This paper investigates the stress-strain -tate of t-:-rl - Alt.ls using the
asytotie solutions of Novotilov. T- e-fect of the silr --t- 5t s 0 is
discussed.

Voloshin, A. A., "Flexibility and gtz*ngth of Convoluted btllows of -tpe 8e,

vestn. xshinostr., 4, pp i2-16, 1958 (in asian).
In pipelines with an internal prsmwe of a to 4 kF,ftM ed below
are idely used. The typimal design is "Ama of a .!-',3ows with an interval hinge

94
 coupler eliminating the relative axial displacewant of the- flange. iPU its
plane of~ motion the coupler permits a linear and angular displacement ofjfaj
flanges of the bellows. Spring rates for these types of notio'n arej dte~ind
experimentall3y for three bellows.

* Wabi, A. M., "t Recent Research on Plat Diaphragms and Circuzlar Plates With
Particular Reference to Instrument Applications%~ AMK trans., J (1),
PP 83-87, January 195T.
A discussion and literature survey of recent theoretical and experimental
develoWuents relating to flat plates and diaphragns is given, with particular
reference to applications in pressure-measuring Instruments. Developments
discussed include: Effects of large deflections; initially buckled lia-
phrsgms; plates subject to plastic flow; analysis of temperature and iccelera-
tion effects in diaphragms for pressuxre measuremuent. Some discussion of
instruments utilizing flat or nearly flat diaphragns is given and an attempt
is made to indicate possible fruitful avenues of future research in the
diaphragm fileld.

'.j1ter, L.,, "Metallic Flexible Seamless Bellows", Mass Prod., 31 (11),

pp 50-5.0, Novemiber 19 419 .

Metallic seamless bellows are design elements used in rmy types of apparatus,
instruments, and plant equipment. Here are a fewo well-known applications
using their elasticity in longitudinal axis: power elements for industrial
thermostats,, cooling water or radiator thermostats for motor cars, pressure
reducing devices, packless glands, packless shaf't seals, expansion joints for
pipes, flexible couplings, noise dampers, petakless valves,, special piston-
less pumps, etc.* These applications may take the form of either a piston
and cylinder,. frictionless and hermeticr.Ily sealed, or of a flexible closure.

Walter, L., "'The 'Use of Flexible Metal Dellovs as Design Elements%,

Part 1, La (2) Wirary 1953,, Part 2, 62 (6), June 1951, Can. Mach, and
Manuf. News.
A general discussion is given in both a-tieles of the design and application
of formed bellows.

Walter.,,,iv o eali
mls Belw In ain lo.Xtrad
Design, 1(opp68a. cbe190

istics of the helical spring and the seniflexible cylinder. Brief 4Iescrip.
tions are gvno t s sasml pig rsuesnigdvco
as a means of accomuodating variations in pipework.
 W0II, NI.A., and Nvrk, N,~ N., %RLa Mastic I1omios rCf*~~
)Iersnea",. App].. Mach., 22$ PP 533-538P 1955.
An lestigtion is presented fur.r tbe Plastic bebavior of CLad. circular
smbrones Subjected to by&ostutic prpevze, busea on the bhroty-tises
theory. New strain-displacement re1.-t~t.4s, valid for finite -deformations -

IA the plastic dcmin, are introduced. The solu3tion is Xpresentd lntrnp

or two simultaneous UQUliit integral equktions, which tan be solvd by
nuaerical methods. The solution also permits the fo1ulaio of the ltabilty
condition. Experinental work consisted of bulge tests caried out on n*sled
cogper plates of two thicknesses, 'hose T-y curve 'was previously datermined
from uniaxial tests. A good correlation vas obtained betiei theedretieii-
predictions =Ad experimental results.

w'~einsarten, V. I., MOrgan, 1. J.., and Seide, P., "MIastic tability of ThIbn-
Walled Cylindrical and Comical Shells Uth~~r Axial Conressin", ADAA J., -1(3),
pp 500-505,9 arch 196.
Results of an extensive experimental progran on the~stabiliiy of cylindrical,
and conical shells under axial compression are presented and discussed. Mwe
experimental data indicate that the buckling coefficient varies with radius-
thickness ratio. A study, of other data in the literature shaved that moft
of the experimenal results fell. within or newr the scatter-bead cbtained in
the present evaluation. A lover bound design curve is also contained In
the paper.

36 eiin ntel, Be 3.,. Cole, C. 7*, and Nitler, J9A,, Ielding Practices for
Usryliua Copper Alos", Welding J., &2 pp 20T-2, march 1963.
This paper provides- an Introduction to the use of beryllivs-couafting
coper allors In welded structures. The factors most affecting weldability
and velwnt properties of alloys containing 0.25 to 2.0 percent beryllima
are: the refractory oxide formed during welding, weld fluidity, base metal
conductivity, specific beat and post-volding heat treannt. bmause of the
rtefactry nmature of berylliu oxide, the quality of welds resulting from,
multipase qpwations is influenced partially by the Interpss joint prepare.-.
tion and cleaning procedue~c. I~n tbis aepeat, the welding of these allos
requires pmactices, as"ii to those fohllamd In the welding of uteris con-
tufting aluminua and titaniin as hardetnig additions. beause the therna].
coaftactivity &an the melting point decrease with Increases In beryllimu
content, the alloys containing z= bftylliw appear the mawe readily weld-
able. Vels~mnt quality and mchaical lwpropts betas. and after various
therm). treatments, for several teryflum copper all,(Wa In sheet and plate
form are'discus med. Where good protective saospheft Is provided, an
allay sinilar to the base metal c&A be used as a filler mtal. Using the
gas tungetten- and setal-arc welding procesees and the electron-beam rces
little allcying occurs. After postweld therna]. treatnunt, weld
mtal properties approach those of the bqmae mtal.

96
 Walls, X_ D.., "Development Teste ~umlmw
U 7%Ultrs%~ 8-.nlessj kot.l?
Inc. Rept. 1.3. 211%, Novmer 7., 1963,
The purpose of this report~ Is to describe the develapmnt testg porftta'~
three alwalmxum bellows designs for tp, I~xtin CompWz. The as~&atse
include a 1.5-in-ah diawfter bellows, a 6-5inch di ter -bellws, and- dut-.-
* ~asseablyp and a 10-Inch diamseter bellowsm. A twtbher purpose of thist pr
is to oastine the difficulties encountered in vibratioa -testing and -to show
that vibration dampers will be required on these designs,

Werner,, F. D., "The DesWg of DI".hragas for Pressure Gages Which use-the
Bonded Wire Resistance-A 0strait OWg", Pmo. Smc. No. Stress Ansl.,,3 ()
pp 137-1)46, 1-953.---
The problex of cuaorcaisinS anong high sensitivity, high natural frequency,
small size, adequate strength, 'wide pressmr range, linear responisep etc*, i
the desig of diaphragms fo~r diepbreas type pressure @Wgs whiceh use the
bonded wire resistance strat4 ga~ge is examied ft detail. A oystewktic pro-
cedure for finding the best compromise is worked out.* A good basis for coat-
parison among various materials is eftablished, and a coop isoz is given in
I- tabular fora. The design procedure is given in 5 equations.. and for steel,
dural, and magnesium allqi, the deaWg equations are graphed in a mnnesr
which nakes dasign especially convenient and rapid. Som suggestions for
mking the diaphragm are incluled.

Westerheide,, D. I., Clifford, J. C., and Burnet, G.., "A Diaphraga Pm

for Liquid Metal Service", Ames Lab., Iova State Univ. of Sci. & Tech.,
R&D Rept. No. N-63-11812.
The use of centrifuiml and electrcingnetic puaps for liquid metals is briefly
reviewed. Details are provided on the construction and operation of a two-
stage diaphragm pump successfully used for the first tim in liquid metal
service. From the results of a 5376-hour test of the pump it was Concluded
that it is well sui~ed to the pumping of liquid 3etals at low flow rates
where pulsating flow can be tolerated. Operatinig temperatures and pressures
are limited only by the availability of suitable materials of construction.

"When Selecting Bellows", Prod. Ing., pp 13)4-136, Aril 1936.

Descriptions are given of steps In the manufacture of gas-tight flexible
metal bellows of corrosion-resistant mterials formed from drwn tubse$ bY
*byrwalic pressure. Methods of designing for iinu cost require considora-
I * tion of bellows mufacturIng Processes.

I9I
 Wite, 1. L.,. and Pink, P. W., "Materials of Constraction for ikndl-Ing
Fluorine%, Proc. of the Propellant 1thermodyn. and Hadnling Coif., pp 161-181,
1959.
There is considerable information on the corrosim behavior of materials -ex-
posed to fluorine. So aaterial is completely Inert to attack by this-eztremly
active eeent. However, =Wn of the cmmo o truction -materials, vln
used ia the etablished tewperature ranges, tan be expected to give -acceptable
service. Nickel al1o~a can be used V to 2200 F; aluelnum and =460osim to
00 F; copper, and possibly thcrtu Ito 700 F; pure Iran to 600 F; mild
ateels Tihlow siliem content to 60~ I; and titaniu and zirconium to 300
P. Cbrooui plzte is useful to ec. least 4W F and nickel plate to hiaber
tesperatires. The use oZ tantalum. ,tin, silver$ lead, platinu is posalble
at ros teprtrs. Siliccn, va -diu, rhenium, and uraiu awe attacked
even below roon temperaftre. Liquid fluorine at liquid-nitrogan tespera-
tures is awe tarrosive than the gas at room temprature. 2he metals vir~ch
can be used with liquid fluorine are nickel and nickel alloys, stainless
steels, aluminum, manesim, mild steel, titaium, tantalum,. and zirconium.
Except for very special applications, the only organic mterials wbich am
be used with either liquid or gaseous fluorine wre the flucx'ocarbrs. -PO4-
tetrafluretlene Is the best. Of the nomet-711c Inorganic materials,
alumina Is very resistant. Mlass,, abestos,, and sam cermrts may have lim-
ited uses, and carbon materials are not recmedd for elevated temperatures.

362.
Whitten,, D. C., -Ten Ways to Use Metal Dispbrages and Capsules", Prod. log...
& M, pp 9e-93, 716brury 17, 1958.
Brief descriptions are given of various applications for metal diapbrasms.

Wigots y, V. W.., *Volume Comensating Velded Bellows*,, Deal Sems, I

De.vices suh as gyros, acelermters, and other floated Instruments are her-
seticsl~y aemld, must comret for fluid-volim chwwxe vith temperature and
often require optlum dming characteristics. This article presents typical
p~~yicsl ~'opeti~sand design data relative to use o oue
applictions
coqpansating velded bellows to achieve these objectives.

Wigotalw, V. VWs"Welded 1iprma Bellows - Fart I%, Dsimp a jw,1

(W) PP 8-9, JTanuarY 1961.
fte versatiity of velded diaphrao metal bellos has created &Wr nei uses
ard impoved perfoiace In oz~y Conventional app~clctons. Welded beflows
as a torque tranamitting device on a gjfzwiue periscope and as a replace-
mt for a spriag in a fluid fIlled control mecmanis we only two exmles.
Thay illuxtrte how the perfarme of these bellows can be varied to suit
the noed.

*1 98
 Wigotsky, V. W.,, "Welded TDiaphraga Bellows -Part Jr', Design gs 6 _(5) --PP 8-9a
February 1961.
* Design curves and charts were prepared fromt test data and mathematcal analysis
and are based upon AM 350 (AlE 554.8), a precipitation-haideiing stinliies~s
steel, beat-treated to Re 42-45 hardness. Curves an spring rate are approxi-
* mate and serve to orient the designer rather than pinpoint exact values.
Curves on the life expectancy are conservative. Teflat plate and nesting
ripple contours generally are used in the zero to 300 psi pressure range
and require a simlar design approach. The single sweep and torus designs
generally are used above 300 psi ad require considerably different treat-
ment. It appears desirable, therefore.. to present comparable engineering
data on these two contours at a later date.

Wildhack, W. A. , and Goerke,, V. H., -Corrugated Metal Diaphragms for Air-

craft Pressure-Measuring Inruments", NPCA Tech. No6te No. 738, 1939.
A large nuaber of corrugated diaphragms of berylium ccper, phosphor bronze,
and Z-nickel,, having geometricaflly similar outlines but ot various diameters
and thicknesses, were formed by hydraulic pressing. The apparatus and the
technique used in the manufacture, the testing, and the heat treatment are
described. The shape of the diaphragm was such that the central deflections
were nearly proportional to the differential pressures up to deflections or
2 percent of the diameter. The pressure-deflection characteristics of the
various diaphrasm were correlated with the thickness, the diameter, and the
* elastic properties by dimiensional analysis to obtain toualas and oharts
applicable to the design of similar diaphragms. For conarism,, some data
are presented for flat diaphragms and for corrugated diapbreas differing
slightly from the standard design. The use of the epetal results in
the selection or the design of corrugated diaphragas is briefly discussed.

Wildhsckp W. A., et al.., -Investigations of the Properties of Corrugated

Diaphragms", AMK Trans., Y2 (1),p pp 6582 January 1957.
The prqasure-deflection characteristics of corrugated diaphragms are correlated
by methods of dimensional analyuis. Erparimental results for various sizes,
materials,, thicknems, and shopes of diaphragms indicate that the performan-ae
for diaphragms of azq' given aspe my be computed fromt a dimensionless for-
xla derived from experimental data an ote diaphragm of that shape.
Linear-shell equations ame derivied for comined beading and stretching effects
with lateral loading terms for rotationall7 ymmetrical shells in appopriate
independent and dependent varlatione suitable for complicated mridial
shapes, and with boundary conditions associated with practical diapbragm,
application;. The method used for solviLng this awtes of equations on an
electronic digital comiter is described and nierical solutions are presented
for a specific diaphragm subjected to vaifocu pre sswe load'ng. aastions
are presented for future research, both thearetical1 and experimental, on
diaphragm properties and perforane.

99
 10-ilok W. A.,, and Oowek. V. N., u"hs Wlmting Useful Deflections of Cotru-
gated Metal Diqhagss NACA !~ch. Note No. 876, 191.2.
The limiting useful deflection of a dispreM my be defired as that Geflec- '

tion which is followd by an arbitrarily chosen allowable limit of bysteresis,

aftereffect, drift, cc zero shift. Preliminary results reported previosy
Indicated that the limiting deflection is minly dependent an the dimitoir,
and the aterial and only slightly dependent on the thitckzess of the SA-
pbraft. In order to Investite further the useful Limits of dispbreap per-
for~aese, -I -,-- -- nt were mad on a larp vostr of corrugated dispbagms
of similar doapes but of various sizes and variouo metals. 2he materiass
studied inalude phosphor bronze, beryllium copper, A-nickel, 3-lionl,, r-*xkel
and Inconel. The apparatus used In uWia the masuwie mmts is described.
Data were obtained on the relations between pressure and deflection, defe-
tics and kbysteresis, time and drift under constant loads time and recovery
after release of load,, and zero ukift &a deflection for the various diatragrm,
Nazy of the results are presented prazhicsllYe The performnces of the aiff-
erent maternals are comared and the charcteristic constants for eac.h mat-
erial are dedued for use with design farula*. The results are analyzed
to determine the correlation of the limiting deflection* with the thicknesses
and diamters of the iaba and the elastic propertlea of the wttials
used.

Vllinmis, G. R., 'NasurIzg Yun' a Noftlus by the Resonence Nothod", Intstr.

Practice, &, Pp 319-321, April 1965.

I 2 feV, taI relationship betweenun' obusa 1 he velocity off

sound Is considered, wAm a wr instrument in described vhich uses this re-
latioNship as a mows of asm
nt.

Winbornes .A.3 -SiPAled dCre o elw hdol

nalysis%
Atomics Internations3l, Auzat 1, l9&.

Design of reliable bellows reiuires (1) solution of the prbblac, to obtain

the required aunal dimnsmions by use of the best the*Wreal methods
available, and (2) eliaitinn or stress-raising iftfluvces which can pro-
&we stresses much biloe then those calculated. OUItified tbeomieical
design procedures, beem-theary eqiations with chats to appl1y cuvtr
corrections, awe presented in this repor for velded plate, convolute,, and
toroidal bellows. Factors In design vtech, in lowge convolute-ehaped bellows,
pro=* stresses uth higbe then those calculated, awe discussed In detail
In another report. A digital-coqputer program. was developed, based an
aqitatic shell theory, to obtain design charts correlatinag stresses ini
shellw shell-shaped (welded plate) bellows with those In amr simple strw,-
turel el]muts. Mnese design charts were developed to permit a proaentAtioi
of fou shell-theary parsters an one graph seaet, wa thus avoid use of
comlicted noographic chwts.

_ _ _ _ _ _ 100D_ _
r 7 - = -~~~

Winborne, R. A., -Stress and Elevated Temperature Fatigue Characteristics of

Lsage Sellows", Attmics International, September 15, 1964 (R6!i-29352Y.
Ten large 20-in.-diameter 304&stainless steel bellows were tested to or near
fatigue failure; nine to 1200 F sodium, and ame in air' at 70 F. The dsta,
Indicated that, at 1200 F, fatigue life of axially deflected large convoluta-
shaped bellows, withouat internal pressure influence, was mainly determine1 by
bellows geometry, and was associated with a factor pi, used frequently In as~Wa-
totic shell theory of c onvolute- shaped bellchrI. Chu a log-log chart there qpared
to be a linear relationship between 1/i and. cycles to failure of axially de-'
flected, large convolute- shaped bellows at 120 F. The fatigue life of e
bellows with a high IL value (p = 6")was less then l1l100 that of a bel~ows with
a low P value (p = 0.25). Conventional, high m (p ->2) bellows always failed
at, or near,. the welded Juncatures or the bellows-to-pipe ends where stresses
were highest. With stresses very uniformly distributed over their profiles,
low P (p = 0.25) bellws appeared to be relatively insensitive to stress raisinc.
influenceg. The fatigue test p>erformance of one toroid-shaped bellows was
excellent.* Charts in this report show axial stress distribution over exterior
bellowts surfaces induced by bellows axial deflection, and by internal pressuri-
zation. The influence of root rings on stress distribution is presented
graphically.

Wintergerst,, E., and Lintz, H.,. "Properties and Applications of Metal Diaphra-exs",
Regelungstechnische Praxis (Nmeich), I~ (5) pp 160-165% MW~ 1959 (in German).
Introductory remarks are followed by a survey of the method of calculation and
also of the properties and applications of mtal diaphragms. A short reffereace
is made to the vellknown formulas for calculating plain metal diaphragms. How-
ever, the methods of approximate calculationi of concentrically convoluted din-
phrag.s, hich have been developed only recently, are dealt with more fully.
Toe article contains, furtherwore, a brief report on empirically roud values
and examples of the applicationi of metal diapbrwas in neawvIng and control
devices, with particular consideration of the arin of safety at excess
pressures,

Wolley., L.W.G., -Manufacture of Brometric Bellows and Like Corrugated Tubes",

Nuclear Bag., 2, p !494., Noveber 1957 (British Patent T17,1146).

The main cbject of this invention is to form a -orrugated tube or metal bel-lows
in which the wall thickness is substantially constant in the corruSated parts,
that is, within a tolerance of 10.031 inc-h with a wall thickness of the order of
0.020 inch. According to the invention a protess for forming a tube of uniform
wall thickness Into a corrugated tube of uniform wall th'.ckness In the corru-
gated parts comprises forming the corrugations successively with the unformed
41 part of the tube unrestrained from moving axially toward the corrugations as
they are formted.

101
Vrstel T.,, "Riniature Pressure Cel Mnstrmets, , PP 7364T39,* M
1953.
Righ-speed aerod~namic testing requirns the use of l. prssure calls having
good 1namic characteristics. Bah cells umfly emp)AW a fixed-eWs 4ingibre
and ake use of sensitive electronic detecting equimnt. Comtrutics detas
and general characteristics are given for tUrsa t7We of miniature prese
cells developed and used by the WA. ~Mw thre e. reasivw, ceeitive,
and Inductive. &affeting: and flutter research requires the -1mIzremnt1 of
fluctuating air loads uo that vith Increased ampbsis an this kind of testing
such cells vii]. play an inareasingly Imortant role an a research tool.

?akharov,. Yu. 0., "Measuring Pulsating Pressures With the Aid of Mjwren
Transducers', AMS, 'PAfl, PTD-2-63-73TA AD 1133076 (traslatoa).
In the investigationi of transient regime and In a numer of other -erimic
prcbfleas it become ncessary to insure a pe re wich champes with ta.
Th Instruments usually used for this kind of wmseur t are memrane tromma-
ducers, which transform a variable pressure Into an electrical Yolt!Ca or
current,, the chage In which can easily be recorded with the zVL of an oncillo-
graph. Th different kinds of electrical transducers - qaItives inauctanceo
te---- rtric - are widel used at the present tliz in aeratry. This article
a aqs to determine ezei~alytub- correction factors for the measured
amplitude and pbase of a pmlaatlm pressure, considering the transducer, and the
Inlet tube as a single cacillating myatea.

ziemwa, 1. 7.,P and Uaohiloy, V. V., 3 %motric totinof a Toroidal

allPrlkl. 16t. 11ekh., (5), pp 521-530,j 1951 (in Russiain).
This paper gives the asymptotic solnUon for the bg 1 n-o prchlen for the
toroidal shell In terms of hakel functions. Tha partticular solution is developed
In terms of a rgomtric series.

IM
 -Satwfo"-aw Oka- tte~ma tftt .4-~

SMONKTINS~~
cai~wm&m ainTY . 06sed WCFT CkA P RATO
Pattele Nkmorial institute I Unelassified.
505 Kin Avenue 2 60
COlvmbuko Mio .43M1 ____________

Hulbert, L. B., X-tithp R. It.; and Trainer, T. X.

C Mfrff AST TaL ". o alp . lb.W

I&a. or %twa
November 1965 171 1 6+ 376 In Anntated blio -.
0. TRACT 00 MOAf? me. 94L Ofte@aTws "80"T wime6ws)
A? 04(61)-10532
ILP*O "no Al'S No. 750 G

AFSC Project *j. 675 96. &r Pf~ *Q AP-

etewaO -o besolO
ASC Task No. 6T5304 UT
A_ _ __ _ _ __ _ __ _ _ __ _ __ _ _ AP PPL- t -65 -
it. AVAMSAWWWUTIUU110
NFlu

Qualified requestors my obtain cies of this report fromn the Defense

* Documentation Center
11. Supftem"TART moU it. weSPONSCM OurANr aCrM 1Y
Air 1bree Rocket Propulsion Lab ortory
Research and Technology Division
Air Fbirce. Systems Cinnd

A 2-1/2-year program, has been under taken to establish analytical

procedvres, stress analysis methods, techniques for manufactiuig control,
and other factors essential to the successful. design and fabrication of
metallic bellows and diapragas. The initial phase of the progam has In-
cluded a state-of-the-art survey to assist in the deteraimitIcn or the best
means of accomlishing the over-all progran objectives. This report sun-
marites the results of the survey and presents recomendstti Enfl for the
remidr of the progrm. An annotated bibliography of 376 references is
included.

DD SM"1473_ _ _ _ _ _
 Aneroid 'Capsules
Axisyumetric Shells
Bellows
Bellows Analysis
Bellows Conftigurat ion
Bellows Design
Bellows Evaluation
Bellows Manufacturing
Computer Solutions
Diaphragms
Diaphragm Azaysis
Diaphragm Configuration
Diapbragm Design
Diasphragm Salustioti
Diaphragm naufacturlng

Direct Integration

&hd-Face Seals
Expansion Joints
Finite Difference
Flexible Comectors
Meretic Seals
*draulic Actuators
Rldraulic Notors
Liaer.aihefl ArAlysis
Mtal Bose
Metallic Bellows
Mettallic Diaphragms
Notion franaducers
$ NolirNear-Sbell Analysis

36ewiy CledWAI
 L~ Uml _

* haft sals
sels or Revolution
Shock Abscbers
Ts rature Sesrs
Vibration Iup ers
Volum Cmuators

iI

(I

.. m$mmimi~m~im