TITLE NO. 69-53
Analysis of Reinforced Concrete Membrane
Subject to Tension and Shear
By NICHOLAS 8. DUCHON
‘An analytical procedure is presented to predict
stresies and. deformations in’ reinforced. or pe-
ressed concrete wall. subjected to membrane
forces, and reinforced by a combination of orthog-
‘onal and inclined bers. The directions of the rein.
forcement are taken independently of the principal
directions. The concrete is assumed to act only in
compresion. The effectiveness of incined bars in
Fesisting shear forces and
Application of the method in goneral practice
suggested.
cofindical shelled.
lament method? mathe
forced concrete: shoat sss
ural analysis toe stot! walle
formation
Beste walls strane
IW Tue CONVENTIONAL DESIGN of reinforced concrete
walls subject to membrane forces calls for an
orthogonal basic reinforcement and sometimes an
additional system of reinforcement following the
direction of the principal membrane tension.
However, even if that direction is known and
constant, which is not always the case, the direc-
tions of principal stresses and strains seldom
coincide in cracked reinforced concrete. Thus, the
conventional design method fails.
Nuclear containment structures represent typi-
cal examples of this problem. The walls, usually
cylindrical, are designed primarily to resist in-
ternal pressure resulting in horizontal and vertical,
tensile stresses, but in addition they must resist
shear stresses caused by seismic or other loadings.
‘These load combinations generally result in wide-
ly varying principal directions.
To determine the effectiveness of various com-
binations of orthogonal and inclined reinforce-
ment, the following analytical procedure has been
developed.
‘578
ASSUMPTIONS
‘The computation is based on force equilibrium,
and strain compatibility of an elementary portion
of the membrane. The concrete is assumed to act
only in compression, the reinforcement in tension
or compression and the liner plate, if present, in
any direction. The concrete is thus modeled as a
continuum intercepted by an infinite number of
fine cracks all perpendicular to the direction of
‘the maximum principal tensile strain, The purpose
of the analysis is to predict the highest probable
stresses in each component element. The effects
of repeated stresses, that alternately open and
close the cracks in various directions, are a prob-
lem of fatigue of the whole structure and can be
found only by actual tests.
‘The treatment of the subject is sufficiently gen-
eral to be applicable to any wall-like structure
having similar loading and reinforcement. Also,
the suggested method may become useful in any
reinforced concrete analysis in which the structure
can be properly simulated by a truss system,
which is the most common case, the reinforcement
representing the tensile members and the concrete
the compressive members,
APPLICABILITY OF ASSUMPTIONS
Under increasing internal test pressure, the typi-
cal cylindrical reinforced conerete containment
shell will first develop vertical cracks, since the
hoop tensile force is twice the vertical force, which,
is further reduced by the dead load.
‘ACI member Nicholes 8, Duchon, structural engineer
bechiel "Corporation, San Francicer Calif. received Nt
fnginesring "diploma trom the. Techical” University of
Budapest Hungary. euperiences
Include bridge” design id “design of
‘) Ms. Cuchsn
tsinments for nuclear power plont
ACL JOURNAL / SEPTEMBER 1972After the first cracking, the wall becomes an
assembly of vertical concrete columns enclosed by
a reinforcing bar net. Further increments of the
pressure will increase the width of the cracks
separating these vertical columns and eventually
additional cracking will develop normal to the
vertical tensile stress. Finally, blocks separated by
vertical and horizontal cracks and held together
by the reinforcement will result.
If membrane shear force acts on the wall after
such a crack pattern has developed, the surface
roughness and dowel action will transfer some
shear stress between the concrete blocks, thus
resisting the shear strain of the wall until the
tensile strength of the concrete is once again ex-
ceeded and diagonal cracks begin to form perpen-
dicular to the principal tensile forces. The model
presented herein becomes quite realistic then, for
high membrane shear forces.
For the case in which the tensile stresses caus-
ing the initial cracking are uniform, the direction
of cracking is random. Then, when a membrane
shear force is subsequently applied, some of the
random cracking will be perpendicular to the in-
clined maximum tensile strain. Thus in this case,
the proposed analysis is applicable from the be-
ginning of the application of the membrane shear
force.
In general, the proposed analysis conservatively
predicts upper limits for stresses and strains of a
shear wall, since all concrete tensile strength is
neglected.
MATHEMATICAL MODEL
‘The structural model used in the analysis is
‘composed of materials each acting in one direction
only, rotated as necessary to represent the vertical,
horizontal and inclined reinforcement, and the
concrete. If both principal strains are tensile (ex-
panding), the concrete is assumed to have zero
strength. This is accomplished in the mathematical
development by setting the conerete stress ou
equal to zero.
The model thus consists of a vertical, horizontal
and two inclined anisotropic materials having
modulus of elasticity E,, and an anisotropic con:
crete material resisting compressive strain e, only
and having modulus of elasticity E,. The
following indices are used: Vertical direction, 1;
Horizontal, 2; Inclined, and 4; the direction of
maximum principal strain I, and that of the min-
imum strain II, which is also the direction of
the compressive concrete resistance (see Fig. 1)
The boundary forces are V vertical, H horizon-
tal, and T shear. The ratio of reinforcement area
to the gross area is p,, i=1 to 4. The normal
stresses in the elements are « and the strains
#4, where i= 1 to 4 and Directions I and II. Within
the range of validity of this analysis, o; = 0.
ACI JOURNAL / SEPTEMBER 1972
~
‘The angle between Direction I and the vertical
is B, and the angles between 3 and the vertical
and 4 and the vertical are as and a, (Fig. 1).
‘The following three equations represent the
equilibrium conditions of the model and must
always be satistied:
ae [»: cos? + py cos* ay cos* (8 — a)
+ Pecos a cos (Bay) Je
+ [p. sin? B + ps cos as sin® (B — as)
a cos? ay sin? (8 — ay)
BEE] eu o
= [b= sin? B+ ps sin? ay cos? (B — a3)
+ py sin® ay cost (8 — «| 8
+ [reso #4 post gaint a)
+ Pisin? a, sin? (8 — ay)
+ Eden ®
aig; =|[s cos? (B05) sin oy 08 0
+ Pe cos? (B — a4) sin ay cos «] &
3,
2
T
H
—
Idealized crack pattern
Reinforcement
Fig. | —Typical wall element
519+[rvse? 0) smacnen
+ px sin® (8 — ay) sin ay cos ay
__ sin Be 8) ea @
Once the three variables, 8, 1, and ey have
been determined, the stresses in the reinforcing
bars can be computed from:
1 = (ex cos* B + ey sin® B) By “@
o2 = (e sin? B + a cos* B) Ey @)
= [ou cost (p — a) + eu sin® (B— «oe, 6)
% =[« cos* (8 — a4) + e1 sin? (B — «| E,(7)
‘The stress in the concrete is’
ou = tn Ee @
And finally the shear strain is:
y= (1 — en) sin 26 ®
Using these equations, the behavior of rein-
forced concrete or prestressed concrete walls can
be conveniently investigated, and the ratio of
reinforcement in various directions can be opti-
mized. Derivation of Eq. (1), (2), and (3) is
given in the Appendix.
EXAMPLE
The effects of variation of the inclined rein-
forcement are investigated in the following ex-
ample, and the results are plotted in Fig, 2, It is,
believed that the example is sufficiently typical
to indicate a general trend applicable to similar
containment structures.
A cylindrical reinforced concrete containment
structure of 88.0 ft (2682m) inside diameter is
subjected to an internal pressure of 58 psi (407
kgf/em?), and a horizontal force of 18,100 kips
(8.22 x 10? kgf). The wall is 6.0 ft (1.83 m) thick.
E,= 5 X 10° psi (0.352 x 10° kgf/em*).
‘The wall is reinforced by a steel area of 18.0 sq
in, per ft (275em*/m) vertical, 8.6 sq in, per ft
(182.em/m) horizontal and a variable amount of
inclined bars at angles aij = + 45 deg,
‘The resultant maximum membrane forces at
an elevation of about 20 ft (6.1m) above the base
slab are: Vertical, 1225 kips per ft (182,000 kgf/
m), horizontal, 369.0 kips per ft (549,000 kgf/m),
and shear 122.5 kips per ft (182,000 kgf/m).
Consequently the model used in this analysis
has the following applied stresses: Vertical, 142
580
psi (10 kgf/em*), horizontal, 427 psi (30 kgf/em*),
and shear 142 psi (10 kgt/em*).
‘The reinforcement ratios are: Vertical 0.015 and
horizontal 0.01. The ratio for the inclined bars is
varied. In Fig. 2 are plotted reinforcement stresses,
concrete stress, strains, and the angle g for steel
ratios ps and p, varying continuously between
zero and 0,025.
In Fig. 3 to 5, all dimensions and forces are the
same except for the applied membrane shear,
which is varied from zero to 500 psi (35.2 kgt/em*)
In these three figures the stresses and strains are
plotted as functions of the applied membrane
shear force.
DISCUSSION OF RESULTS
In Fig. 2 the tensile stress in the vertical rein-
forcement is rapidly reduced as the area of
inclined reinforcement is increased, since the in-
clined reinforcement system tends to compress
the wall in the vertical direction while permitting
ine)
has
vert.
@ pk Tinaiiay :
oor]
Strains
8
stress, pst.
8
8
then
7
F
709| =
60°]
}-Angle of inclination jo
Eo 005, Ol 015 02
Arnot Incl. reint /grors area, By b¢
6
Fig. 2 — Stresses, srains, and angle B
‘ACI JOURNAL / SEPTEMBER 1972
a.004
8
[Shear strain
the horizontal extension under the action of the
relatively large horizontal force.
In the absence of inclined bars the compressive
principal stress in the concrete reaches a level
twice that of the applied boundary shear stress.
As the area of inclined reinforcement is increased,
concrete stress decreases and the direction of the
compressive principal strain approaches the ver-
tical
‘The concrete stress reaches a minimum value
as the ratio of inclined reinforcement becomes
heavy, but then further increase of the inclined
reinforcement results in greater concrete stress,
‘This is the result of the vertical compression ex-
erted by the inclined reinforcement system under
the action of the relatively large horizontal force.
In Fig. 3, the high initial shear strain predicted
by this analysis in the absence of inclined bars
can be observed. Increasing the area of inclined
reinforcement results in decreased shear deforma-
tion and also in a more nearly linear variation of
that strain,
A close relation can be found between the shear
strains and the direction of the principal tensile
strain. Without inclined reinforcement the shear
3 deformation is not prevented until the previously
3 cracked conerete is compressed. As the applied
$ 5° “00 (iby 200 o}300 400180) 500
- Applied shear stress, pst (kgf/cm) p= Ps, Pa cy
Fig. 3 — Shear strain, principal stress, angle 8
~409) 1
[4
(za) T=
ll
ar
S$
es Ee or |]
z F009) :
= 409, ; 5 Fg
= is
= ee 3 Bleo00)
z —| 2 2
100 z
fr. 0) 7005 BOLT = $1009)
fe al Es
i <= 3—e
= 400 :
i * ogg
(2a)
-200 _@000)
3 (10) z 20
= # 0000)
> “100 (16) 200 (@0)800 400 (80) 500 2
Aviad seat stot, atm") puts 04
Fig. 4 —Conerete stress components
‘ACI JOURNAL / SEPTEMBER 1972
’ 10 Go) 200 (20)300 400180) 500
Applied shear stress, pal (hgt/em*) P=Ps, Pe
Fig. 5 — Reinforcement stresses
581shear stress increases, angle f defining the direc-
tion of the principal tensile strain rapidly turns
from the original 90 deg to about 60 deg and then
slowly approaches 45 deg. The curves which rep-
resent heavy inclined reinforcement exhibit a
greater resistance to shear strain, and also a slow-
er change of the direction of the principal strains
as the applied shear force is increased.
Without inclined reinforcement the concrete
principal compressive stress increases nearly
linearly with the applied shear. Light inclined re-
inforcement (p= 0.005) carries the total shear
force even at very small shear deformation. As
the applied shear stress increases, at some point
the compressive component of the shear stress
overcomes the normal tensile stress in the direc-
tion of the minimum strain and thereafter the
compressive concrete stress increases approxi-
mately linearly. Changes in the principal direc-
tions correspond to those in the concrete stresses.
With heavier inclined reinforcement the conerete
{is in compression in the vertical direction even
without any applied shear force.
Fig. 4 shows the vertical, horizontal and shear
stresses in the conerete, as found by projecting
the prineipal stress into these directions. The
variation of the vertical stresses with increasing
applied shear seems to be similar to that of the
principal stresses, while variation of the horizontal
and especially the shear components is more near-
ly linear.
Fig. 5 shows that without inclined reinforcement
the stress in the horizontal bars increases to very
high values as the applied shear force is increased,
and the provision of inclined reinforcement sig-
nificantly reduces this stress.
A similar effect on the vertical reinforcement
can be observed. Heavy inclined reinforcement
even causes compression stresses in the vertical
steel at the lower values of the applied shear.
At zero applied shear force the stresses in the
inclined bars are the same for both Directions 3
and 4. With increasing applied shear force bars in
Direction 3 (Fig. 1) are subject to increasing ten-
sion, while those in Direction 4 lose tension and
pass finally into compression. Tension in the
heavier inclined reinforcement varies more linear-
ly with applied shear, as the concrete becomes
relatively less important in resisting the shear
force.
SUGGESTED USES
1, The described analytical method is recom-
mended in the design of shear walls, and specif
ically nuclear containment structures. The design
582
parameters are, in effect, only the compressive
strength of the concrete and the tensile strength
of the reinforcement, which makes for a simple
and dependable design method.
2, When the cracking of concrete is taken into
consideration in a finite element program pre-
pared for axisymmetric structures, shear analysis,
such as this is needed because the action of i
clined reinforcement renders the structure non-
axisymmetric.
3. The prediction of the shear strain, and hence
the horizontal deformation of cracked structures,
makes this method useful in dynamic analyses. By
varying the amount of inclined reinforcement the
shear stiffness of the wall, and therefore the dy-
namie response, may be influenced,
4. The curves shown in Fig. 2 to 5 are valid
only for the structure proportions and loads stated.
Nomographs or tables may be prepared for fre-
quent and rapid application. A short computer
rogram has proved useful.
5. Using this analysis the need for inclined re-
inforcement in addition to the orthogonal system
can be determined, based either on stress levels
or strain limitations. Generally, excessive concrete
compressive stress or horizontal deformation of
the structure exceeding acceptable limits would
call for an inclined reinforcement system.
6. To apply this method in the design of pre-
stressed concrete structures it is only necessary to
include the prestressing forces in the applied
boundary loading.
CONCLUSIONS
1. If large shear forces are acting, conerete
compressive stresses are significantly reduced by
a sufficient area of inclined reinforcement.
2. The addition of inclined reinforcement con-
trols the shear deformation more effectively than
the orthogonal system alone, especially at small
shear stresses,
3. In a typical nuclear containment structure
the role of the vertical reinforcement diminishes
greatly when inclined reinforcement is added.
REFERENCES
1. ACI Committee 249, “Criteria for Reinforced Con-
crete Nuclear Power Containment Structures,” ACI
Journat, Proceedings V, 69, No. 1, Jan. 1972, pp. 2-28.
2, Holley, M. J. Jr, “Provision of Required Seismic
Resistance,” Seismic Design for Nuclear Power Plants,
the MIT Press, Cambridge, Mass, 1970, pp. 401-437,
5. Falconer, Bruce H., “Theory of Stresses Induced
in Reinforced Concrete by Applied Two-Dimensional
Stress," ACI Jounwat, Proceedings V. 53, No. 3, Sept.
1986, pp. 277-294,
ACL JOURNAL / SEPTEMBER 19724. Peter, J,, “Zur Bewehrung von Scheiben und
Schalen fir Hauptepannungen sehiefwinklig zur
Bewehrungsrichtung,” Die Bautechnik (Berlin-Wil-
mersdorf), V. 43, No. 5, 1966, pp. 149-164, and V. 43,
No. 7, 1966, pp. 240-248.
8, Kuyt, B, “Zur Frage der Netzbewehrung von
Flichentragwerken,” Beton und Stahlbetonbau (Berlin-
Wilmersdorf), V. $9, No. 7, 1964, pp. 158-163,
6. Aas-Jakobsen, A, “Die Berechnung der Zylinder-
schalen,” Gottingen, Berlin-Heidelberg, 1958, pp.
1e-t18.
7, Scholz, G., “Zur Frage der Netzbewehrung von
Flichentragwerken,” Beton und Stahlbetonbau (Berlin-
Wilmersdorf), V. 58, No. 10, 1958, pp. 250-255.
‘APPENDIX
Notation
Ay = total eross sectional area
Ee = modulus of elasticity of concrete
E, = modulus of elasticity of reinforcing steel
i index referring to direction in plane of
membrane,
thus:
T= vertical
2 = horizontal
3 = inclined (as)
45 inclined (aa)
1 = maximum principal (most tensile)
minimum principal
n ‘modulus of elasticity ratio: Es/Ee
Pi = ratio of area of reinforcement in direction
{to total area of section
H_ = membrane horizontal force per unit length
T = membrane shear force per unit length
V = membrane vertical force per unit length
as angle between direction i = 1 andi
4 = angle between direction i = 1 andi = 4
B = angle between directions i = 1 and i = 1
* shear strain
a normal strain in direction i
= normal stress in direction 4
Derivation of Eq. (1), (2), and (3)
‘The force equilibrium of the model may be estab-
lished as follows:
= prox + pos cos! aa + poss cos? ae + on sin? B
= pron + pron sin? us + pics sin? a4 + on cos?
= aos sin as €08 as + pros sin 04 608 as
as as 2
—onsin p cos p
‘ACI JOURNAL / SEPTEMBER 1972
‘The strains in any direction can be expressed as the
projections of the principal strains, as follows:
er = cos? Bert sin’ Ba
2 Bert cost Ben
ca = cos? (8 — as) er + sin? (B — as) arr
ost (B — a1) ex + sin? (B — a) enn
Using the relations « = EiBy' and on = en Ee’ and
substituting the preceding values of into the equi-
Ubrium equations, we obta
+[ am (peo +t en]
x prcont + [cov (B— a+
eusint (p — «| Pe cost ay +H int p
athe p+ are p) + weeeip a
4 want a) airs fasonip—ed
+ a sint (p09 ] pein a + 8 ce
[ ]
x ps sina co 0 + [ co¥ (B — a0)
‘taint (B — 00) J pe sina cos a
sin f cos B
Rearranged in terms of and et, Eq. (1), (2), and
(2) are obtained, from which a, «1, and p may be
found conveniently by an iterative process, such a the
following:
For any given value of , «1, and sn can be ex-
pressed from any two equations which represent a
nonhomogeneous system. By successively increasing,
the values of the three variables can be determined
so that the third equation is also satisfied.
rod by the Institue Aug, 22, 1971