Material design of aerated concrete-

An optimum performance design

S. Tada
Materials ScienceLaboratory, Misawa Homes Institute of Research and Development2-4-5, Takaido, Suginarniku, Tokyo.

An optimum design for Autoclaved Aerated Concrete (AAC) panel is proposed on the basis
of the performance requirements for dwellings. With bulk density representing "property"
as an intensive variable and thickness of the wall representing "quantity" as an extensive
variable. Performance requirements such as cost, thermal, acoustic insulation and structural
safety are illustrated in terms of bulk density d and thickness of wall T for AAC. For the
Tokyo area, a combination of d = 5 0 0 (kg/m 3) and T = 0 . 1 2 (m) is desirable while d = 4 0 0
and T=0.18 is recommended for colder regions.

1o INTRODUCTION with each other since the thickness of the wall must be
1.2 m for thermal requirement.
In spite of the vast amount of research devoted to We, hereinafter, will prove that an application of the
the properties of autoclaved aerated concrete (AAC), monolithic use of AAC for dwellings is an example
emphasis has scarcely been laid on its actual characteris- which exhibits the multi-functional characteristics in
tics which fulfill almost all performances in simple question, and will illustrate quantitatively that AAC is
monolithic construction. It is in the last 15 years that oriented to the lower bulk density.
the performance requirements for buildings have been The scope of the present study is limited to the
quantitatively elaborated, with quantitative design performance design for AAC panels, and the design of
methods based on the performance concept still being pore structure which plays a central role in the material
researched and improved upon. design for AAC will be reported on subsequently.
In the case of monolithic use of AAC, however,
the design method will be simplified considerably in
comparison with a multilayer composite wall. It is 2. PROPERTIES AND PERFORMANCE REQUIREMENTS
especially important to notice that almost all per-
formance requirements for dwellings can be represented 2.1. Performance criteria
by both bulk density of AAC, d, and thickness of the
wall, T. The bulk density is a concept representing the 2.1.1. Insulation requirements
property as an intensive variable while the thickness is a The requirement for the insulation of air borne sound
concept representing quantity as an extensive variable. is represented as the value of transmission loss, TL, a
We will begin with a review of research on the commonly adopted practice in the regulations of many
properties of AAC carried out over the last ten years countries. Most of the objects of the control is limited
and will approximate them as functions of the bulk to the party wall between two families, however for
density. Next, substituting these relations with the per- the outer wall, T L > 3 5 (dB) may be sufficient as a
formance requirements which are normally given in the minimum requirement.
form of inequalities, the relations between d and T will For the solid and monolithic walt, well-known mass
be presented, and by illustrating the inequalities on a law acts on the transmission loss TL which is given as
space composed of d and T, a region where possible a function of the surface density and the frequency.
combinations of d and T actually exist will be shown. The following calculation will be carried out at fixed
However, depending on the sort of material, there frequency, 500 Hz, for simplicity. In addition, TL will
may be no such region. A monolithic normal concrete be lowered due to the coincidence effect, which cannot
wall is a rather good example, as it is durable and be covered by mass law. The AAC panel should be
inexpensive. However, common requirements of designed so as to eliminate the coincidence effect or,
transmission loss of 35 dB and thermal transmittance at least, to keep it out of the frequency range sensitive
K-value of 1.0 will never be economically consistent human hearing. The coincidence frequency, f~, is there-

0025-5432/86/01 21 06/$ 2.60/ 9 BORDAS-GAUTItlER-VILLARS 21

Vol. 1 9 - N ~ 1 0 9 - M a t a r i a u x e t C o n s t r u c t i o n s

TABLE I
MINIMUM K-VALUE FOR OUTER WALL.

Heating degree-days (~ day)

Countries References Enforcement
2,000 3,000 4,000
The Building
England. 0.60 0.60 0.60 Regulation. 1982
Part F.
France. 0.57 0.41 Decret No. 74-306 1974
Decret No. 74-306
W. Germany. 0.76 Gesetz zur Einsparung 1977
von Energie in Gebafiden
Sweden.. 0.30 Svensk Byggnorrn 1977
U.S.A.. 0.45 0.40 0.30 HUD-FHA 1979

fore necessary to fulfill the fr > 3,000 or f~ < 300 require- AAC panel, i. e., o < F b. In addition, excessive deflec-
ments. tion of the panel due to the external force may make
The requirement for the insulation of heat has been the occupants feel uneasy and cause the detachment of
represented as the coefficient of thermal transmittance, components, so that the deflection 6 should be limited
K, and adapted widely to the regulations of countries. to the range of ~ < a / 6 0 0 where a is the span of the
The minimum value of K varies according to the wall. Of course, AAC can be reinforced by steel bars
heating degree-days given in table I, since the climatic and the admissible stress m a y vary according to the
conditions of a region in which the houses are built degree of reinforcement. Here, the calculation will be
differ from one another. In the following calculation, carried out without regard to the steel reinforcement
K < 0 . 4 (W/mK) and K < 0 . 6 will be employed, respecti- and the section will be assumed to be dependent only
vely, for relatively cold regions and for relatively warm on the AAC matrix. Bending strength F b and Young's
regions. modulus of elasticity E are necessary to be approxima-
ted as a function of the bulk density of AAC for our
The thermal conductivity X (W/InK) and Young's calculation.
modulus of elasticity E (Mpa) of AAC have therefore
to be given as a function of its bulk density for our 2.2. Properties
calculation.
Among many engineering properties of AAC,
2.1.2. Economical requirements Young's modulus of elasticity, thermal conductivity,
Assuming that T is the thickness of the thermal bending strength, compressive strength and frost resis-
insulator, a well-known method for calculating the opti- tance are selected on the bases of the above examina-
mum thickness of the insulator can also be adopted to tion of the performance requirements, and the depen-
AAC. The theoretical bases of the method are derived dency of the properties on the bulk density are presen-
from the fact that a minimum yearly cost for heating ted from figure 1 to figure 3 accompanied by the
may certainly exist when the yearly repayment of the experimental equations.
insulator increases with greater thickness while the run- Excellent linearity can be seen in each equation with
ning cost for heating and the initial cost of heating a correlation coefficient of more than 0.96, for the data
equipment are eventually reduced. The optimum from various reports are employed only if the testing
thickness of the insulator will be given as a function method, the size of the specimen and the moisture
of its thermal conductivity. content are clearly described.
A more direct economical factor is the cost of the The data for E are derived from the works of Short
AAC panel per unit area. W (yen/me), which is given and Kinniburgh [1], H u n k e and Eliasson [2] and
by W=d.T.C. If C, cost of AAC per unit weight, is Ozawa [3], for thermal conductivity, data by Schuele
assumed to be 70 yen/kg, then d . T < 5 7 and d . T < 7 2 [4], Parjak [5] and L o u d o n [6] are employed in addition
will be given according to the variation of W from to [1] and [2].
4,000 yen/m 2 to 5,000 yen/m 2. Of course C also varies
Since the aim of the examinations of compressive
according to conditions: the relation of d and T should
strength and frost resistance is to check the lower limit
be considered merely as a measure for the time when
of bulk density, no experimental equations have been
W changes in 1,000 yen steps.
formulated. It seems necessary to comment on the Scr
2.1.3. Structural requirements which has been employed as a measure of frost resis-
The outer wall in question is assumed to be non- tance of cement-based porous materials. The critical
load bearing and to act mainly as the resistance to degree of saturation, Scr, is defined in the R I L E M
wind pressure. Cracks in an AAC panel due to the Recommendation reported by Fagerlund [8]. Scr has
bending stress caused by wind pressure should be been considered to be a threshold moisture content
avoided. The bending stress c borne in the AAC panel, over which the material m a y be implicitly damaged in
should always be less than the bending strength. Fn, of the presence of moisture and sub-zero temperature.

22
 S. Tada

Here, the data measured by Larsson and Purins [9],

the authors [10] and Roulet [11] are employed. It may
be concluded from figures 4 and 5 that the lower limit 3.

of the bulk density is approximately 300 kg/m 3.

2.

3. PERFORMANCE D E S I G N uJ 1.

3.1. Mass law and the coincidence effect 200 400 600 800 1000

The transmission loss TL of air borne sound through Fig. I. - - Dependence of E-module on bulk density.
a rigid, monolithic wall such as AAC is known to be
governed by Mass law. TL is given as a function of
frequency f and surface density m of the component 0.25
9 Short /
x Schuer 9 9
and is presented as follows in the case of diffused 0 Author
0.20 9 Hunke and Er
incidence ([12], [13]): Loudon

Y-2. 43~10"4X+4.62=10-3
0.15
TL = 20 log (2nfm/2 pc)
- 10 log [In [1 + (2 n fro~2 pc)2]], (1) 0.10

where m = surface density (d.7) kg/m2), pc = specific O. 05 .Y

impedance of air (414;kg/m2.sec.) and f = frequency r l ~imturl conttnt-l.G-S,O Ivo~X)
0
(500; Hz) are denoted. Upon substituting the condition GO0 400 600 800 1000
T L > 3 5 examined in the previous section 2.1.1 into Fig. 2. -- Dependence of thernml conductivity on bulk density.
equation (1), we obtain:
d. T > 50 [Transmission loss > 35 dB]. (2) 1.5 . . . .
[] I ~ n k e s n d s ~-2. 3-'g. e O f
9 H~nk. and [klas~n. e-33. t-39. 2 / . a*~
Actually, the sound insulation performance is not 9 oza.i. *,-25.1-46.3

dependent merely upon the Mass law but also upon

Y-2. 5 0 " I 0-3x-O. 2 5 7 ( d r y ) ~-"
the rigidity and internal resistance of a wall. A well- Q j:.-"

known coincidence effect makes the calculated value O. 5 ~ ,-" y-I. 7s*lo-3x-9. 88*10"2 (wIt }
based on Mass law slip down towards an undesirable
value. The coincidence frequency fr can be given as .~ ~-~aoisture content In u

follows: 200 400 600 800 1000

L = ,/a. rm = r. (3) Fig. 3. -- Dependence of bending strength on bulk density.

where c = velocity of sound in air (340; m/sec.), B = ben-

ding rigidity of wall per unit area
J
.0....
U Hunkm and E ~ | a s ~ n "
0 Author
A Hebek ho.db~k
!/ 9 9
(B = ET3/12 (1 - cr2) = ET3/12; N-m), E = Young's 9 Author n - 1 3 5

modulus of elasticity given in figure 1 g

(E=5.31d--853;Mpa), d=bulk density (kg/m ~) and
T = thickness of wall (m) are denoted. Upon substitu- //"
ting the conditions f~> 3000 and f , < 300 examined in
the previous section 2.1.1 into equation (3), we obtain:
j."
0"~ 9 moisture content -1.8-1. ? (VOLX)
T2<85d.lO-6/(d-160) [fr > 3,000 Hz], (4) 0 200
I I
400
I
500 800
I
1000

T2>85d.lO-'/(d-160) brr< 300 Hz]. (5) Fig. 4. -- Dependence of compressive strength on bulk density.

Upon combining the inequalities (2), (4) and (5),

a shaded region is illustrated in figure 6 where the
combinations of d and T, capable of satisfying the
acoustical performance, can exist exclusively. 9 0

3.2. Coefficient of thermal transmittance

If the outer wall is composed, for example, of a 0.6 20
//x (au o

O
9
O
Hunk. and E~Lasm)n
Author
L a r l l o n and Purinm
(m) wide panel, the overall K-value may be increased x RouLet

slightly due to the thermal bridge effect of the joints. O0 200 400 600 9800 1000
However, in this calculation, it is assumed that the
panel is room-size wide and the surface distribution of Fig. 5. -- Dependence of critical degree of saturation for freeze-thaw
redstance on bulk density.

23
Vol. 19 - N ~ 109 - Mat6riaux et Constructions

0.3 the thickness of an insulator, the optimum thickness is

given by the following equation [14]:

"~ o. 2
T= Lx/1/a.pl (DfE/rl +p2JA0) - LR, (9)

where L=thermal conductivity of AAC given in

figure 2 (L=2.43 x 10-4d+4.62 x 10-3W/mK),
o
0.1 l
~',~.,..,.~, TL>35 D=heating degree-second (1.69x108 for Tokyo,
3.46 x 108 for Sapporo; K.sec.), f = c o s t of fuel per unit
0
calorification (1.54x 10-6; yen/J), E=coefficient of
0 200 400 600 800 1000
operation of the central heating system (0.85;-),
l~g. 6. -- Combinations of d and T for the acoustical requirements. 11=heating efficiency of the central heating system
(0.70, oil base;-), p l = coefficient of yearly repayment
0.3 for AAC (0.095;-), p2=coefficient of yearly
repayment for central heating system (0.134;-),
J=cost of central heating system per unit calorification
0.2 (50; yen/W), AO= temperature difference of design base
(20; degree), a=cost of insulator per unit area and unit
thermal resistance (L.70.d; W/mK.yen/m3) and
0.1
R=thermal resistance factor, except for the insulator
(0.15;m2K/W), are denoted. The example calculation
J is carried out on the basis of the climatic conditions of
0 I000
0 200 400 600 800 the Tokyo and the Sapporo area as follows:
Fig. 7. -- Combinations of d and T for the thermal requirements. T=0.0161 x//'~-.-A-0.15L [for Tokyo area], (10)
T=0.0224x//-~-.d-0.15 L [for Sapporo area]. (11)
moisture content is uniform. The coefficient of thermal These equations are also illustrated in figure 7 together
transmittance is given as follows: with the results of the K-value. It is apparent that the
K= 1~(1loci+ T/L+ 1/eto), (6) K-value required in each of the regions do not overlap
the curves of optimum thickness. This means that it is
where ed, cto=surface coefficient of heat transfer" difficult to treat the AAC as an exclusive insulator.
(9.3, 23, respectively: W/m2K), L= thermal conductivity
of AAC given in figure 2 (L = 2.43 3.4. Crack and deflection due to wind pressure
x 10-*d+4.62x 10-3;W/mK) and T= thickness of
wall (m) are denoted. Upon substituting the conditions The AAC panels with a 0.6 m width are normally
K < 0 . 4 and K < 0 . 6 examined in the previous section joined, in situ, so as to form one body. However, in
2.1.1 into equation (6), we obtain: this calculation, it is assumed as the first step, that the
panel acts as a single panel for safety consideration.
T>5.71 x 10-4d+ 10.85 x 10 -3 The bending stress cr charged on the tension side of
[K < 0.4 W/maK], (7) the panel simply supported at each end is given by the
following equation:
T>3.68 x 10-4d+7.00 x 10 -3
G=a.q.a2/T 2, (12)
[K < 0.6 W/m2K], (8)
where at=coefficient of stress at the width of 0.6 m
Upon combining the inequalities (7) and (8), a sha- (1.25;-), a = s p a n (3; m), T = thickness of wall (T; m)
ded region is illustrated in figure 7. It can easily be and q = uniform load caused by wind pressure specified
seen that the desirable d is the smaller one and T is in the Japanese Building Law as 60x/~ with h = 6 m in
the thicker one. kgf/m 2 (1,440; pa) are denoted. Upon substituting the
condition cr < Fb examined in the previous section 2.1.3,
3.3. Optimum thickness of thermal insulation and the experimental equation of
F b = 1 . 7 6 x 1 0 - a d - 9 . 8 8 x 1 0 -2 in the wet condition
For the calculation, it will be assumed that the houses given in figure 3 into equation (12), we obtain:
are equipped with a central heating system using Kero-
sene with an overall heating efficiency of 0.7. The price T2 > 9.20/(d- 56.1) [or <Fb] (13)
rises at a rate of 8 ~ and the interest on money are Next, under the same structural condition, the deflec-
based on a rate of 9 ~. A preliminary calculation is tion of the panel due to wind pressure will be examined.
carried out for the mean cost of fuel during the 6 years The deflection fi at the center of the panel is given by
repayment period of the heating system and for the the following equation:
yearly repayment rate of the AAC wall pl and of the
heating system p2 on the bases of their life cycle, 35
= [3.q.a3/ET a, (14)
and 13 years, respectively. Assuming that T is equal to where [3=coefficient of deflection at the width of 0.6 m

24
 S. Tada

(0.26;-), a = s p a n (3;m), T = thickness of wall (T;m), 0.3~

q = u n i f o r m load caused by wind pressure (1,440;pa)

and E = Young's modulus of elasticity given in figure 1
(E = 5.31 d - 853; Mpa) are denoted. U p o n substituting
the condition 6 <a/600 examined in the previous sec-
tion 2.1.3 into equation (14), we obtain: i

Ta>O.381/(d-160) [6 <a/600] (15) Oefor~ tion

U p o n combining the inequalities (13) and (15), a 00 200 400 600 800 1000
shaded region is illustrated in figure 8 as the criteria
for structural requirements of AAC. It seems that a Fig. 8. -- Combinations of d and T for the mechanical requirements.
greater bulk density and a thicker wall are more favora-
k<O. 4 k<O. 5
ble. Of course, some modifications on the bending /
/ :/ :/
rigidity of the panel should be made with respect to
the safety coefficient, moisture conditions and the width
~ el\\ .........
,:/
of the panel.

4. C O N C L U S I O N S

It has been shown in the above examinations how

the combinations of d and T of AAC can satisfy the i IL>J~
200 400 600 800 I000
insulative, economical and structural requirements for
dwellings. In addition to this criteria, the conditions Fig. 9. - Optimum combinations of d and T for performance require-
ments of dwellings.
for the cost of the A A C panel per unit area, examined
in the previous section 2.1.2, should be included. The
integrated figure, including the conditions of the cost
per unit area, d . T < 5 7 a n d d . T < 7 2 , is shown in
o.3i
II 1'' ......... [
figure 9. I k<l,0
0-21
It is in a very small area that the promising combina- I
tions of d and T can exist, and it is seen that the bulk
0.11
density must be less than 400 kg/m 3 and the thickness
must be greater than 17 cm even for the relatively
warm region. A challenge in manufacturing to the
lower limit of bulk density may eventually be required 0 200 400 600 800 1000

when a target is placed in the B or C region.

Fig. 10. - Optimum combinations of d and T for the p e r f o r m a n c e
The condition for crack formation circumscribing requirements of dwellings (warm regions).
the lower limit of A, B and C regions is based on an
assumption that the width of the A A C panel is 0.6 m.
Actually, the panels are joined so as to act as one
REFERENCES
body. Moreover a room-size panel has recently been
produced [15]; and requirements for thermal insulation
[1] SHORTA., KrNNIBURGHW. -- Lightweight concrete, CR
may be reduced to K < 1 taking the climatic conditions Books, London, 2nd and 3rd edition, 1968 and 1981.
of Asian countries into consideration. Accordingly, [2] FUWKE K., ELIASSONB. - - Hoegisolerande laettbetong
figure 10 can be obtained, where the acoustical require- sore baerande element, (High-insulation aerated concrete
ment of T L > 3 5 and f c < 3 0 0 circumscribe the lower as bearing elements), Byggdok, Stockholm, 1978,
limit of the optimum region, an economical require- pp. 369.
ment of A < 4 , 0 0 0 border the upper limit, and the A [3] OZAWAM. -- Internal report, Misawa Homes Institute
of Research and Development, Tokyo, 1981 and 1983.
or B region can be chosen according to the thermal
[4] SCHUELEW. - Waelmeschutz mit Gasbeton, Deutsche
-

requirement. Bauzig., 1974, No. 10, pp. 1787.

The combinations of d and T within the shaded areas [5] PAP,JAg I. -- Ddtermination de la conductivit6 thermique
denoted as A and B can satisfy all the requirements with des mat~riaux de construction en Sudde, Annales de
a lower cost. A higher degree of thermal performance, ITBTP, 1979, No. 376, pp. 92.
however, will be required in the presence of an energy [6] LOUDON A.G. - The effect of moisture content on
thermal conductivity, in Autoclaved Aerated Concrete-
shortage, and AAC will be oriented to a lower bulk Moisture and Properties, Elsevier, 1983, pp. 131.
density and a thicker wall. The conclusions drawn [7] HEBELTECHNICALHANDBOOK, Asahi Kasei Kenzai, 1982,
here are based on the conditions peculiar to Tokyo. pp. 274.
Therefore it may be necessary to modify the optimum [8] FAGERLUND G . - - T h e critical degree of saturation method
diagram according to the climatic conditions and the of assessing the freeze/thaw resistance of concrete, M a t e -
regulations of the countries. riaux et Constructions, 1978, Vol. 10, pp. 217.

25
Vol. 1 9 - N ~ 1 0 9 - M a t 6 r i a u x et C o n s t r u c t i o n s

[9] LARSSONL.-E. P~R~NS E. -- Investigation on the frost [12] KAMURA S. -- Building acoustics and noise reduction
resistance and pore structure of lightweight concrete, Publ. design, Shohkokusha, Tokyo, 1977, pp. 95.
77.15, Chalmers Institute of Technology, G6teborg, Swe- [13] CgEs'ro~s Ph. -- Comportement acoustique des patois en
den, 1977, pp. 242. b~ton I~ger, Cahiers de CSTB, 1978, No. 190, pp. 63.
[10] TADAS., NAKANOS. -- Microstructural approach to pro- [14] The Ministry of International Trade and Industry,
perties of moist cellular concrete, in Autoclaved Aerated Annual report 1975; Research and development of the
Concrete-Moisture and Properties, Elsevier, 1983, performance of dwellings, Japan Housing Equipment
pp. 71. System Association, 1976, pp. 203.
[11] ROULET C.A. -- Expansion of aerated concrete due to [15] HANAOr,A T., TADA S. -- House 55 Misawa system and
frost - Determination of critical saturation, ibid., PALC technology, Prec. Innovating Technology in Buil-
pp. 157. ding, 1983, Paris, pp. 45.

R~Sl.nvlg peuvent ~tre repr~sent~es par ~ d ~ pour la densit~ appa-

rente et e T ~ pour F~paisseur du mur.
Pour atteindre la performance optimale, l'analyse de
Pr6d6termination des mat&iaux dans le b&on cellu- la r~gression des propri~t~s du B C A : r~sistance fi la
laire autoclav6 : conception performantielle compression, r~sistance ~ la flexion, module E, conducti-
optimale. - - N o u s proposons ici une ~tude concernant la vit~ thermique et le degr~ critique de saturation pour la
conception performantielle optimale du B~ton Cellulaire r~sistance au gel par rapport ~ la densit~ apparente d,
Autoclav~ (BCA) bas~e sur l'~tude des caract&istiques sent d'abord repr~sent~s comrae des ~quations exp&imen-
enregistr~es au cours de ces dix derni~res ann~es. tales. Ensuite, les exigences de performance pour les
L' utilisation monolithique du B C A pour les tours ext~- comportements acoustiques, thermiques et m~caniques du
rieurs est caract~ris~e par une composition simple et une tour ext~rieur sent class~es sous forme d'in~galit~s pour
r~duction importante du co~t. Cependant, ram~lioration d e t T.
de l'isolation thermique en r~duisant la densit~ apparente En combinant les ~quations exp&imentales avec les
tend ~ ~tre incompatible avec celle de l'isolation phoni- in~galit~s, on peut exprimer les exigences de performan-
que. N~anmoins, les demandes de r~duction de poids ces par d et T et faire une v~rification par les limites
s e n t souvent incompatibles avec les ameliorations des de la r~sistance fi la compression et la durabilitY. La
propri~t~s m~caniques. Tous ces facteurs doivent ~tre representation graphique des exigences de performances
pris en compte pour d~terminer quelle conception peut pour les habitations pr~sente une d~termination quantita-
aboutir aux meilleures performances. tive du rapport entre d et T en fonction des conditions
A v e c l a densit~ apparente repr~sentant en principe la climatiques des pays. Pour la r~gion de Tokio, un rapport
~ propri~t~ ~ comme variante intensive et l'~paisseur du d = 5 0 0 (kg/m 3) et T = 0 , 1 2 (m) est recommand~ alors
tour repr~sentant la ~ quantit~ ~ comme variante exten- que pour les r~gions plus froides un rapport d = 4 0 0 et
sive, les exigences de performances pour les habitations T = O, 18 est recommand~.

26