Sie sind auf Seite 1von 18

Behaviour of concrete block masonry

prisms under axial compression


E.H. Fahmy and T.G.M. Ghoneim

Abstract: A nonlinear three-dimensional finite element model was developed to study the complex
behavior of ungrouted and grouted concrete block masonry prisms under axial compression. The model
detects crack initiation and traces crack propagation in the masonry assemblage. Variable strengths for
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

blocks, mortar, and grout were used to study the effect of the mechanical properties of prism
constituents, and their combinations, on the prism strength and modulus of elasticity. The effect of the
number of courses was also investigated. The results of the finite element analysis were used to develop
simplified relationships to predict prism strength and modulus of elasticity. Good agreement was
observed between the available experimental data and the predicted prism strengths.
Key words: compressive strength, concrete blocks, failure modes, finite element, masonry, modulus of
elasticity, prisms.

Resum6 : Un modkle d'ClCments finis tridimensionnels non linCaires a CtC dCveloppC pour Ctudier le
comportement complexe de prismes de ma~onneriede blocs de bCton IiaisonnCe par injection et non
IiaisonnCe, soumis i une compression axiale. Le modkle dCtecte l'amorce de fissures et retrace la
propagation de celles-ci dans l'assemblage de ma~onnerie.Des resistances variables pour les blocs, le
mortier et le coulis ont CtC utilisCes pour Ctudier l'effet des propriCtCs mCcaniques des composantes des
For personal use only.

prismes, et leurs combinaisons, sur la resistance et le module d'ClasticitC des prismes. L'effet du
nombre d'assises a Cgalement CtC analysC. Les rCsultats de l'analyse par la mCthode des ClCments finis
ont CtC utilisCs pour Claborer des relations simples permettant de prCvoir la rtsistance et le module
d'ClasticitC des prismes. Une bonne concordance a Ct observCe entre les donnCes expkrimentales
existantes et les rksistances prkvues.
Mots clis : rrtsistance en compression, blocs de biton, mode d'Cchec, Clement fini, ma~onnerie,module
dlClasticitC, prismes.
[Traduit par la redaction]

Introduction can practically be considered representative for masonry


strength investigation.
Compressive strength is the most important characteristic of
This paper presents the results of an investigation to
masonry. CAN3-S304-M84 and ACI 530-88IASCE 5-88 use (i) study the stress distributions and failure modes of ungrouted
the prism compressive strength as a basis for determining the and grouted concrete masonry prisms under axial compres-
allowable stresses for masonry under different loading con- sion; (ii) determine the effect of the mechanical properties of
ditions. Grout provides large cross-sectional area, which the prism constituents and number of courses on the prism
increases the compressive load carrying capacity of masonry. strength and modulus of elasticity; and (iii) develop simpli-
Hamid and Drysdale (1979) have developed a mathematical fied relationships to predict the prism strength and modulus
model to predict the prism strength depending on the mechani- of elasticity.
cal properties of prism components. Cheema and Klingner The finite element method was used in the present study
(1986) have developed an analytical model for masonry to obtain a detailed description of stress distributions in the
prism considering the constituents as linear elastic materials. three orthogonal directions of the prism.
They accounted for material nonlinearity by using the secant
modulus of elasticity. Boult (1979) and Hamid and Chuk-
wunenye (1986) have shown that a three-course prism model
Finite element model
A nonlinear three-dimensional finite element model was
developed using ANSYS finite element package (Swanson
Received July 18, 1994. Analysis Systems Inc. 1990). A three-dimensional concrete
Revised manuscript accepted January 31, 1995.
element with eight nodes, each having three translation
E.H. Fahmy and T.G.M. Ghoneim. Construction degrees of freedom, was used in this model. This element
Engineering Unit, The American University in Cairo, accommodates material nonlinearity, cracking, and crushing
P.O. Box 2511, Cairo, Egypt. of concrete. To simplify the complexity of the geometry and
Written discussion of this paper is welcomed and will be interaction of prism constituents, the following assumptions
received by the Editor until February 29, 1996 (address were made:
inside front cover). (i) For ungrouted prisms, mortar was placed along the full
Can. J. Civ. Eng. 22: 898-915 (1995). Printed in Canada 1 Imprimt au Canada
Fahmy and Ghoneim

Fig. 1. Stress-strain behavior of an imposed force test, nth iteration.


a'
t
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14
For personal use only.

cross-sectional area of the block units, as recommended by fied, no cracking or crushing occurs. Otherwise, if any
Harnid and Chukwunenye (1986). Although the Canadian principal stress is tensile, the material will crack with the
code (CAN3-S304-M84) neglects the webs in calculating the crack plane normal to the direction of that principal stress.
ungrouted prism strength, Hamid and Chukwunenye (1986) Crushing will occur if [I] is satisfied and the three principal
have shown that the vertical stress distribution at the face- stresses are compressive.
shell along the height of faceshell-bedded prisms is highly In the finite element program, cracking was modeled
nonuniform compared to that of full-bedded prisms. They using the smeared crack modeling technique, in which an
have also shown that larger lateral tensile stresses are created adjustment of material properties was made by introducing a
in the webs due to the beam action caused by the gap between plane of weakness in a direction normal to the crack face.
the webs. Also, a shear transfer coefficient, P,, was introduced to
(ii) Tapering of faceshells and webs was eliminated with represent the shear strength reduction factor for the subse-
the minimum thickness of faceshells and webs considered. quent loads which induce sliding (shear) across the crack
This assumption may not affect the results of the ungrouted face. If the crack closes, all compressive stresses normal to
prisms, where the contribution of mortar is governed by the the crack plane are transmitted across the crack plane and
minimum cross-sectional area of faceshells and webs. On the only a shear reduction factor, PC, for a closed crack is intro-
other hand, for grouted prisms, it leads to neglecting the duced. In the finite element model, P, and PC were taken as
effect of tapering on the grout column. 0.10 and 0.80, respectively.
(iii) Perfect bond (no slippage) was assumed at the inter- Due to nonlinearity, small load increments were applied
faces between blocks and mortar. It has been shown (Hamid to allow for solution convergence. Convergence is achieved
and Chukwunenye 1986) that the shear stresses developed at when the plasticity ratio, ~ E P ' / E ~ '(Fig.
, 1) at all integration
the block-mortar interfaces are far below the frictional points in the structure is less than a preset value p (Swanson
resistance due to normal compressive stresses, which justi- Analysis Systems Inc. 1990). A large value of p would lead
fies the second assumption. to less accurate computed (converged) stresses; a very small
The nonlinear material behavior was modeled by the value of p would significantly increase the number of itera-
Besseling (sublayer) model that allows for a multilinear tions required for convergence. For the present finite ele-
stress -strain behavior. The criterion for failure of concrete ment model, this value was taken as 0.01. The integrity of
under multiaxial state of stress was expressed in the follow- the structure due to progressive cracking was checked after
ing form (Swanson Analysis Systems Inc. 1990): each load increment. Failure was assumed to occur when the
plasticity ratio was higher than the preset value or if non-
converged concrete elements exist.
A model representing one-eighth of a three-course prism
where F is a function of the principal stress state, and S is (Fig. 2) was used. The values of x-displacements of plane
a failure surface expressed in terms of the material compres- OCGD; Y-displacements of plane BCGF, and Z-displace-
sive strength, f,,and tensile strength, f,. If [1] is not satis- ments of plane OABC were set to zero, as they are planes of
Can. J. Civ. Eng. Vol. 22, 1995

Fig. 2. Finite element model. where f,, is the mortar compressive strength. The secant

z4 e \7
Grout
(If an,)
modulus at 0.2% strain, E0.2m,was taken as %Ei,; and the
confined mortar behavior was described by an idealized
stress -strain curve depending on the block strength (Cheema
and Klingner 1986). Poisson's ratio was taken as 0.21
D (Hamid and Chukwunenye 1986); and the mortar split-
ting tensile strength, Am, was determined as (Cheema and
Klingner 1986)

Web - Hedstrom and Hogan (1990) have shown that a linear


relationship between grout strength, f,,, and its modulus of
elasticity is more appropriate than the ACI 530-88 formula-
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

tion that relates the grout modulus of elasticity to the square


root of its strength. Hence, in the present model, the grout
initial modulus of elasticity, Ei,, was taken as 950f,,; the
secant modulus of elasticity of grout at 0.2% strain, E0,2g,
was taken as %E%;the ultimate strain was taken as 0.3%;
and Poisson's ratio was taken as 0.21. The grout splitting
tensile strength, A,, was taken as

Analytical model verification


To verify the validity of the finite element idealization for
For personal use only.

blocks, five block units (type A) with dimensions 190 x


190 x 390 mm which meet the ASTM C90 requirements
for dimensions, absorption, and strength of normal weight
symmetry. The upper face was subjected to uniform vertical grade N units were tested under compression, and fitted with
displacement and zero lateral restraint to simulate a friction- a strain measurement device, to determine the compressive
less loading platen in an experimental test. strength and modulus of elasticity of the units. The measured
dimensions of the block units were used to define the geome-
I Material characteristics try of a finite element model for one block unit. Using the
Parabolic curves were adopted to represent the stress - strain experimentally determined unit weight and compressive
relationship of blocks, mortar, and grout materials. The strength of the block units, the idealized stress-strain relation-
adopted curves pass through points a, b, and c (Fig. 3a) ship for the block units material was determined. Figure 5
defined as follows: point a is the origin; point b represents shows the comparison between the experimentally obtained
the stress defined by the initial modulus of elasticity, El, at stress-strain curve for block units type A, with its 95 % con-
25 % of the full strain range; and point c is the stress defined fidence interval, and that obtained from the finite element
by the secant modulus of elasticity, E0.2, at 0.2% strain. analysis. The good agreement between the experimental and
The initial modulus of elasticity of block units, Eib, was analytical results (Table 1) verifies the validity of the finite
taken as (ACI 31 8 M-83) element idealization.

[2] El, = 0 . 0 4 3 ~ ;56 Analytical investigation


where wb is the unit weight of block a n d h b is the compres- Having proved the validity of the finite idealization for block
sive strength of the block unit. Furthermore, for the finite units, the model was extended by assuming two other types
element model, the parabolic curve was replaced by a multi- of block units (B and C). The two types were taken such that
linear five-segment curve (Fig. 3b). Poisson's ratio was con- the initial modulus of elasticity was increased over that of
sidered as 0.18 (Hamid and Chukwunenye 1986). The block type A by 25 % and 50%, respectively. The three types had
splitting tensile strength, Ab, was taken as (Cheema and the same geometry, thus the block strength and modulus of
Klingner 1986) elasticity would be the only parameters affecting the prism
characteristics. Three types of mortar (Nl, N2, and S) con-
forming with ASTM C270, and CSA A224, requirements
and two types of grout (GW and GN) complying to ASTM
whereLb is the compressive strength of the material of the
C1091 requirements were used in the analysis. Table 2 shows
block unit. The stress-strain relationship for mortar was
the properties of blocks, mortar, and grout.
idealized using the same procedure for blocks (Fig. 4). The
initial modulus of elasticity, Eim, was taken as (Cheema and The prism specimens were grouped into four groups to
determine the effect of the properties of the constituent mate-
Klingner 1986)
rials on the prism behavior, and to study the effect of the
number of courses on the prism, for both ungrouted and
Fahmy and Ghoneim

Fig. 3. Stress-strain relationship of block units.


Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

Strain (xloP3)
Fig. 4. Stress-strain relationship of mortar.
20
i
For personal use only.

Strain (x 1O-R)

grouted prisms. Tables 3-5 show the details of the prism was under nearly equal lateral tensile stresses, as shown in
groups and the results of the finite element analysis for Fig. 7. For both ungrouted and grouted prisms, the lateral
ungrouted and grouted prisms, respectively. tensile stresses along the webs were higher than those for
the faceshells, especially near the interior corners, indicating
that cracks would initiate in the webs. Mortar, in both cases,
Results and discussion
and grout were under a state of triaxial compression. For
Stress distribution in the precracking stage both ungrouted and grouted prisms, the prism maximum ver-
For both ungrouted and grouted prisms, the block units were tical strain, and accordingly the maximum vertical stress,
under tensile stress in both transverse directions. For before cracking increased with the increase of mortar strength.
ungrouted prisms, the stresses normal to the thickness were For grouted prisms, the prism maximum vertical strain before
larger than those in the direction of the thickness. The full cracking decreased, while the maximum stress increased,
height of the middle block was under tension in both trans- with increasing grout strength for the same type of blocks
verse directions, while for the top and bottom blocks, the and mortar. Generally, the prism maximum vertical strain
stresses decreased to nearly zero at the mid-height of the before cracking was less for the grouted prisms than for the
blocks (Fig. 6). For grouted prisms, the full height of blocks ungrouted prisms for the same type of blocks and mortar
Can. J. Civ. Eng. Vol. 22, 1995

Fig. 5. Comparison between analytical and experimental depending on the block-grout modular ratio as explained
stress -strain curves for block units type A. earlier. However, the vertical stress along the mortar joints
can be divided into two main categories according to the
block-mortar modular ratio, nbm (EiblEim):
1. Prisms of nb, 5 2.4: The vertical stress along the
mortar joints is slightly below the average stress in the block
units, especially for the internal faces. The maximum devia-
tion of the vertical stress in mortar from the average vertical
stress in the block units was -5% (Fig. 9).
2. Prisms of nbm > 2.4: The deviation of the vertical
stress in mortar from that in the block units reached nearly
- 15% for the internal faces, as shown in Fig. 10. This is
attributed to the effect of the grouted core where, from com-
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

patibility of strains, the vertical stress in mortar is governed


by the grout-mortar modular ratio, ngm,which leads to an
increase in the grout vertical stress and a decrease in the mor-
tar vertical stress at the bed joint level.

Stress distribution in the postcracking stage


All crack planes were essentially vertical and normal to the
direction of the major principal tensile stress (x-direction for
webs and y-direction for faceshells). The maximum deviation
of crack planes from being vertical was f2". The results
showed that crack patterns depend on the interaction of the
prism constituents (blocks, mortar, and grout). Therefore,
because of the cumulative effect of mortar and grout on the the modular ratios were used to describe the prism behavior
For personal use only.

block units. in the postcracking stage. For ungrouted prisms, the prisms
For ungrouted prisms, the vertical stress distribution can be categorized into three main categories according to
along the faceshells can practically be considered uniform the block- mortar modular ratio, nbm:
and equal to the average vertical compressive stress for all 1. Prisms of low modular ratio (nbm 5 1.5): Cracks
cases (Fig. 8a). However, the vertical stress in the webs was initiated at the interior corner of the webs near the interface
slightly above the average near the interfaces with mortar, with mortar. Considerable crack propagation occurred
while across the mortar joints it was slightly below the aver- before failure covering the full height of the web of the
age. The maximum deviation from the average vertical stress middle block and about 75 % of the web height for the top and
was +_5% (Fig. 8b). bottom blocks. The faceshells suffered less crack propaga-
For grouted prisms, the vertical stress distribution is tion before failure. Failure mode was splitting of block units.
governed by the interaction of the three prism components. 2. Prisms of moderate modular ratio (1.5 < nbm < 3.3):
For compatibility of vertical deformation, the relationship Sudden cracking in both webs and faceshells occurred along
between the vertical stress in the block units, Szb, and the the full length of the block - mortar interfaces. However, the
grout, S,,, is (Hamid and Drysdale 1979) lateral stress distribution just before cracking indicates that
cracking would occur first at the interior corner of the web
near the interface with mortar. No considerable crack propa-
where nbg is the block-grout modular ratio. Increasing the gation occurred before failure and cracks covered approxi-
grout strength, and accordingly its modulus of elasticity, mately 25 % of the height of blocks. Failure mode was, also,
leads to an increase in the vertical stress in the grout, causing splitting of blocks.
higher induced lateral tensile stresses in the block units and, 3. Prisms of high modular ratio (nbm 1 3.3): Failure
hence, cracking of blocks at a lower strain level. occurred suddenly and was initiated by crushing of mortar.
For equilibrium in the vertical direction, the vertical No cracking occurred before failure. However, due to the
stress in the shell (block units and mortar joints) and the large lateral expansion of mortar, large tensile stresses are
grout is related to the prism average vertical stress, Szp, by induced in the blocks at the interfaces with mortar causing
the following equations (Hamid and Drysdale 1979): successive cracking of block units.
For grouted prisms, a relative modular ratio, n,, repre-
senting the three prism components was used, where

where v is the ratio of net-to-gross cross-sectional area of


block units. It can be seen from the equations that the vertical According to the value of n,, prisms can be divided into
stress in the blocks is higher than the prism average vertical two main categories:
stress for nb, > 1, and vice versa. Based on the analyti- 1. Prisms of n, < 2.4: Cracks initiated in the blocks at
cal results, the vertical stress distribution along the block the interior corner of webs near the interface with mortar.
units can practically be considered uniform with a value Considerable crack propagation occurred before failure cover-
Fahmy and Ghoneim

Table 1. Comparison between experimental and analytical results for block units type A.
- -- -

Experimental
Analytical
Standard Analytical Experimental
Property Results deviation results (%)

Gross compressive strength, fgross (MPa) 7.71 0.619 7.65 0.992


Net compressive strength, f,,,(MPa) 13.22 1.070 13.11 0.992
Initial modulus of elasticity, Eib (GPa) 16.81 0.825 14.83 0.882
Ultimate strain at failure, E, ( X 9.64 0.229 9.11 0.945
Secant modulus of elasticity at failure,
Esb @Pa) 13.70 0.795 14.28 1.050
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

Table 2. Properties of prism constituent materials.

Block units Mortar Grout

Property A B C N1 N2 S GW GN

w (kg/m3) 2143 2240 2380 - - -


Ei (GPa) 14.83 18.53 22.24 5.60 8.40 12.40 14.08 19.31
E0.2 (GPa) 9.88 12.36 14.83 2.80 4.20 6.20 7.04 9.66
f, (MPa) 13.11 16.69 20.00 5.60 8.40 12.40 14.77 20.29
ft (MPa) 1.87 2.09 2.29 1.38 1.69 2.05 1.60 1.92
For personal use only.

Table 3. Constituents and analytical results of groups I and 11. of the web near the interface with mortar. Limited crack
propagation occurred before failure and cracks covered
Prism Block Mortar No. of f,, Ei, ESP about 25% of the height of all block units. No cracking
group Code type type courses (MPa) (GPa) occurred in the grout prior to failure.
I A-N1 A N1 3 7.02 14.10 13.24
For all cases, the failure mode was splitting of block units.
A-N2 A N2 3 9.00 14.29 13.83
Mortar did not suffer any damage before the prism failure.
A-S
B-N 1 Effect of constituent materials and number of courses
B-N2 (a) Block units
B-S For both ungrouted and grouted prisms, prism strength, f,,,
C-N1 increased with increasing the block strength.
C-N2 For ungrouted prisms, increasing the block strength by
C-S 50% resulted in an average increase in the prism strength of
I1 AS2 approximately 15%. Block strength has two opposite effects
AS3 on the prism strength: (i) the tensile strength of the block
AS4 increases with the increase of its compressive strength,
AS5 accordingly higher load carrying capacity of the prism would
AS7 be expected; and (ii) the increase in block strength is accom-
AS9 panied by an increase in its modulus of elasticity leading to
ASW higher induced lateral tensile stresses in the blocks due to
mortar confinement and, consequently, reduction of prism
strength is expected. It is clear from Fig. 11 that the rate of
increase in the prism strength decreases with increasing the
ing the full height of the middle-block web and from 75 % to block strength except for very low mortar strength (speci-
nearly the full height of the webs of the top and bottom mens B-N1 and C-N1) where the failure mode changes from
blocks. For faceshells, cracks near the corners propagated up block splitting to mortar crushing. In this case, increasing the
to nearly the full prism height in most cases, while less crack block strength leads to higher lateral confining stresses in
propagation took place away from the corners of the face- mortar and consequently an increase in its triaxial compres-
shells. Also, considerable crack propagation occurred in the sive strength. In spite of such increase in mortar strength, the
grout at the interface with blocks before failure. prism strength is still below what it should be if failure were
2. Prisms of n, L 2.4: Cracks initiated in the webs of the due to block splitting.
blocks along the full length of the block-mortar interfaces. The percentage of increase in the prism strength with
However, the lateral stress distribution just before cracking increasing the block strength was less for the grouted prisms
indicates that cracking would occur first at the interior corner than for the ungrouted prisms. For weak grout (GW), a 50%
904 Can. J. Civ. Eng. Vol. 22, 1995

Table 4. Constituents and analytical results of Group 111.


Prism Block Mortar Grout No. of EIP ESP
group Code type type type courses (GPa) (GPa)

I11 AWD
AWE
AND
ANE
ASD
ASE
BWD
BWE
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

BND
BNE
BSD
BSE
CWD
CWE
CND
CNE
CSD
CSE

Table 5. Constituents and analytical results of Group IV.


For personal use only.

Prism Block Mortar Grout No. of f,, EiP ESP


group Code type type type courses (MPa) (GPa) (GPa)

IV ASD2 A S GW 2 10.63 14.45 13.12


ASD3 A S GW 3 10.62 14.44 13.11
ASD4 A S GW 4 10.60 14.43 13.11
ASD5 A S GW 5 10.59 14.42 13.09
ASD7 A S GW 7 10.53 14.42 13.10
ASD9 A S GW 9 10.47 14.41 13.11
ASDW A S GW 12 10.37 14.41 13.10

increase in the block strength led to an average increase in ungrouted and grouted prisms, respectively.
that of the prism of approximately 8 % , while for normal
grout (GN), the percentage average increase in the prism (b) Mortar
strength was only 6 % . It can be seen from Fig. 12 that the For both ungrouted and grouted prisms, the prism strength
rate of increase in the prism strength is less for the lower increased with increasing mortar strength. Figure 15 shows
mortar strength and the higher grout strength. The decrease the relationship between mortar strength and the prism
in mortar strength is accompanied by a decrease in its modu- strength for ungrouted prisms. It can be seen from the figure
lus of elasticity, leading to higher induced lateral tensile that the relation between the prism strength and mortar
stresses in the blocks. Also, increasing the grout strength, strength is nearly bilinear with the point of change in slope
and accordingly its modulus of elasticity, leads to a higher representing the change of failure mode from crushing of
vertical stress in the grout ([8] and [9]) causing higher lateral mortar, for lowf,,, to splitting of blocks, for highf,,. To
tensile stresses in the block units. verify the trend of this relationship, some auxiliary prism
The initial modulus of elasticity of the prism, Eip, specimens with very low mortar strength were analyzed
increases with increasing that of the blocks. The modulus of (marked with an X in Fig. 15). The increase in the uniaxial
elasticity of the prism is linearly proportional to that of the mortar strength results in a corresponding increase in its
block units for the same type of mortar and grout. For modulus of elasticity which, in turn, reduces the confining
ungrouted prisms, increasing the modulus of elasticity of stresses. As a result of this, the percentage increase in the tri-
blocks by 50% resulted in an average increase in that of the axial strength of mortar is less than that in the uniaxial
prism of approximately 45%. For grouted prisms, a 50% strength, which leads to the observed small rate of increase
increase in the initial modulus of elasticity of blocks caused in the prism strength. In this case, increasing the mortar
an average increase in that of the prism of approximately strength by 40% led to an average increase in the prism
26%. Figures 13 and 14 show the relationship between strength by 12%. In the region where failure is due to split-
blocks initial modulus of elasticity and that of the prism for ting of blocks, the effect of mortar is more significant.
Fahmy and Ghoneim

Fig. 6 . Typical lateral stress distribution for ungrouted prisms: (a) along the mid-width
of external web (internal face of section A-A); (b) along the mid-length of faceshell
(internal face of section B-B).

HOLLOW BLOCK PRISM


I
- (A-S)

1
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

! JUST BEFORE JUST AFTER

CT~-
ccKtNG'
For personal use only.

Stress (MPa)

Stress (MPa)

Higher mortar strength leads to less induced lateral tensile prisms, the relationship between the prism strength and mor-
stresses in the blocks, which is the main cause of the prism tar strength varied with the type of grout (Fig. 16).
failure in this case. Increasing the mortar strength by 120% For weak grout (GW), the relation was nearly bilinear
caused an average increase in that of the prism by 75% with the point of change in slope being at a block-mortar
(specimens A-Nl vs. A-S and B-Nl vs. B-S). For grouted modular ratio of nearly 2.4. As explained earlier, before this
Can. J. Civ. Eng. Vol. 22, 1995

Fig. 7. Typical lateral stress distribution for grouted prisms: (a) along the mid-width
of external web (internal face of section A-A); (b) along the mid-length of faceshell
(internal face of section B-B).

(a) I GROUTED PRISM


Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14
For personal use only.

Stress (MPa)

GROUTED PRISM
69400 4 I , Type (A-S-D

'
B
312.50
0
JUSTAFTER
.+
cn CRACKlNG
247.50

Stress (MPa)

limit the vertical stress in mortar is about 15% less than the auxiliary prism specimens with very low mortar strength
average vertical stress in the blocks and, consequently, the were analyzed (marked with an X in Fig. 16). For low mor-
effect of mortar strength on that of the prism is less pro- tar strength (nb, > 2.4), increasing the mortar strength by
nounced. To verify the trend of this relationship, some 110% led to an average increase in that of the prism by 9%,
Fahmy and Ghoneim

Fig. 8. Typical vertical stress distribution for ungrouted prisms: (a) along the
mid-length of faceshell (internal face of section B-B); (b) along the mid-width of
external web (internal face of section A-A).
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14
For personal use only.

Stress (MPa)

HOLLOW BLOCK PRISM

Stress (MPa)
while for high mortar strength (nbm I2.4), increasing the linear with the rate of increase in the prism strength being
mortar strength by 77 % led to an average increase in the less than that for the weak grout. This is attributed to the high
prism strength by 23 % . For normal grout (GN), the relation- grout -mortar modular ratio, ngm, which leads to less verti-
ship between mortar strength and the prism strength was cal stress in mortar and hence reduces its contribution to
Can. J. Civ. Eng. Vol. 22, 1995

Fig. 9. Typical vertical stress distribution for grouted prisms of n,, I2.4: (a) along
the mid-width of external web (internal face of section A-A); (b) along the
mid-length of faceshell (internal face of section B-B).
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14
For personal use only.

Stress (MPa)

(b)
GROUTED PRISM
690.00 . Type (A-S-D)
642.50 .
495.00 -

447.50 .

- *a0.40
590.40
--
-g ,a,:
295.00 - JUSTBEFORE JUSTAFTER
s - CRACKING
.-
CRACKING
$ 247.60 .

200.00
frn.40
;
1a50.

96.00 .
47.50 -

%-
,. a
; : g j z z
m
0 0 a

Stress (MPa)
prism strength. For this case, increasing the mortar strength significant effect on that of the prism, especially for mortar
by 120% led to an average increase in that of the prism strength higher than 6 MPa which is the practical range for
by 22%. mortar strength. For ungrouted prisms, increasing the initial
The initial modulus of elasticity of mortar, Ei,, has no modulus of elasticity of mortar by 120% caused an average
Fahrny and Ghoneirn

Fig. 10. Typical vertical stress distribution for grouted prisms of n,, > 2.4: (a) along
the mid-width of external web (internal face of section A-A); (b) along the

60400
,
mid-length of faceshell (internal face of section B-B).

(a)
.
GROUTED PRISM

64Z50.

4500 .
447.50 .

-aa g z -
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

S4Z50.
0
$ m.W JUST BEFORE
m
i--' CRACKING
.A
p, U7.M .
IW.00 -
180.W -
14ZM.

9600.

47.60 .

BSOBPBSO
For personal use only.

dee665==

Stress (MPa)

Stress (MPa)

increase in that of the prism by 5 % (Fig. 17). For grouted (c) Grout
prisms, increasing the mortar initial modulus of elasticity by Figure 19 shows the effect of grout strength on the prism
120% led to an average increase in that of the prism by only strength. It can be seen that for weak mortar (N1 and N2),
2 % (Fig. 18). the prism strength is linearly proportional to grout strength.
Can. J. Civ. Eng. Vol. 22, 1995

Fig. 11. Relationship between block strength and ungrouted Fig. 13. Relationship between blocks initial modulus of
prism strength. elasticity and the initial modulus of elasticity of ungrouted
prisms.
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

- Mortar Type N1
--- Mortar Tppe N2
-.- Mortar Type S
3,
10 12 14 16 18 20
fcb ( D a )

Fig. 12. Relationship between block strength and grouted prism


strength.

Fig. 14. Relationship between blocks initial modulus of elasticity


For personal use only.

and the initial modulus of elasticity of grouted prisms.

On the other hand, for strong mortar (S), the relationship


between the prism strength and grout strength is nonlinear
with the rate of increase in the prism strength decreasing with
the increase in grout strength. Grout contributes to the prism hand, for strong mortar the third effect is less pronounced
strength in three ways: (i) increasing the grout strength leads and the prism strength is governed by the counter balance
to an increase in the load carrying capacity of the prisms; between the first and the second effects causing the reduction
(ii) increasing the grout strength, and accordingly its modu- of the rate of increase in prism strength as shown in Fig. 19.
lus of elasticity, decreases the block-grout modular ratio, For weak mortar, increasing the grout strength by 37%
nbg,and hence higher vertical stress is induced in the grout caused an average increase in the prism strength by 13%,
causing higher induced lateral tensile stresses in the block while for strong mortar the percentage of increase in the
units due to grout confinement; and (iii) the increase in grout prism strength ranged from 3 % for the strongest blocks
strength leads to an increase in the grout-mortar modular (type C) to 7 % for the weakest blocks (type A).
ratio, n,,, which reduces the vertical stress in mortar and The prism initial modulus of elasticity is linearly propor-
accordingly decreases the lateral stresses induced in the tional to that of grout (Fig. 20). Increasing the grout modulus
blocks due to mortar confinement. of elasticity by 37 % led to an average increase in that of the
For weak mortar, the third effect is more pronounced and prism by 13 %.
a balance between the second and the third effects is
achieved, thus leading to the observed linear relationship (d) Number of courses
between the prism strength and grout strength. On the other The finite element results show that the prism strength and
Fahrny and Ghoneirn 91 1

Fig. 15. Relationship between mortar strength and ungrouted Fig. 18. Relationship between mortar initial modulus of
prism strength. elasticity and the initial modulus of elasticity of grouted prisms.

.2.

-
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

Block Type A
--- Block Type B
Block Type C

0 3 6 9 12 15 12j I
f c m (MPa) 4 6 8 i0 12 14
E (GPa)
Fig. 16. Relationship between mortar strength and grouted prism
strength.
Fig. 19. Relationship between grout strength and prism strength.
For personal use only.

0 5 10 15 2'0 25
fc, (MPa)

Fig. 17. Relationship between mortar initial modulus of Fig. 20. Relationship between grout initial modulus of elasticity
elasticity and the initial modulus of elasticity of ungrouted and prism initial modulus of elasticity.
prisms.
22
I

- Block Type A
--- Block Type B
-.- Block Type C
91 2 Can. J. Civ. Eng. Vol. 22, 1995

Fig. 21. Effect of number of courses on ungrouted prisms: Fig. 22. Effect of number of courses on grouted prisms:
(a) prism strength; (b) prism initial modulus of elasticity. (a) prism strength; (b) prism initial modulus of elasticity.
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

0 3 6 9 1;
Number of courses
Number of courses
For personal use only.

0 3 6 9 12
Number of courses Number of courses

initial modulus of elasticity decrease with increasing the use the relative strength ratios to predict prism strength and
number of courses for both ungrouted and grouted prisms initial modulus of elasticity.
(Figs. 2 1 and 22, respectively). However, for ungrouted For ungrouted prisms, a block-mortar strength ratio,
prisms, insignificant effect was observed for number of fbm, was used, which is given by
courses more than 5. The prism strength for a 12-course
prism is 98% of that for a 5-course prism and 92% of that [I1] fbm = f c b / f c m

for a 3-course prism. For grouted prisms, the continuity pro- Figure 23 shows the best-fit curve, using the least squares
vided by the grouted cores reduces the effect of number of method, for the relationship between fbm and the relative
courses. The prism strength for a 12-course prism is 98% of prism strength, fpb, given by
that for the 3-course prism.
For both ungrouted and grouted prisms, the initial modu- [I2] fpb = f c p / f c b
lus of elasticity of a 12-course prism is approximately 99.7 %
of that for the 3-course prism. The figure also shows the best-fit curve for fpb related to
EiP/&,. It can be seen from Fig. 23 that for fbm > 3.0
Prediction of prism strength and modulus (which is equivalent to a nbmvalue of 3.3), the decrease in
the relative prism strength is insignificant as the failure mode
of elasticity changes from splitting of blocks to crushing of mortar. This
Based on the above discussion, it can be seen that both prism curve can be used as a tool to predict the prism strength and
strength and initial modulus of elasticity depend on the initial modulus of elasticity for a given strength of blocks and
modular ratios of the prism constituent materials. Since the mortar, as shown in the example in Fig. 23.
compressive strength of all prism components is proportional For grouted prisms, a strength ratio, f,, was used, which
to their modulus of elasticity, it would be more practical to is given by
Fahmy and Ghoneim

Fig. 23. Relationship between fpb and f,,


(ungrouted prisms). Fig. 24. Relationship between Pr and f, (grouted prisms)
-, L P l f , b ; - --, E i p K p .
E ip /fcp (x1000)
3.0 2.5 2.0 1.5 1.O
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

Fig. 25. Relationship between Er and f, (grouted prisms).


For personal use only.

fbrn = f c b / f c r n

Also, relative prism strength and modular ratios, PI and


E,, respectively, were used, which are given by

Figure 24 shows the best-fit relationship between f, and


P,, while Fig. 25 shows the relationship betweenf, and E,.
The upper limit of both curves is atf, = 1.26, which repre-
sents the weakest blocks and grout with the strongest mortar
(prism ASD); the lower limit is atf, = 12.19, which repre-
sents the strongest blocks and grout with the weakest mortar
(prism CWE). These limits (plotted in the figures) represent prisms under axial compression. Simple curves were devel-
the practical combinations of grouted prisms. Results of pub- oped to predict the compressive strength and initial modulus
lished experimental work were compared with the predicted of elasticity for both ungrouted and grouted prisms.
prism strengths using Fig. 24. Table 6 shows a comparison Based on the analysis presented in this paper, the follow-
between the predicted prism strength and the available ing conclusions may be drawn:
experimental data in literature (Cheema and Klingner 1986; 1. The developed model can be used to describe the
Drysdale and Hamid 1983; Hamid et al. 1978; Hegemier behavior of the prisms under axial compression in the pre-
et al. 1978) for both ungrouted and grouted prisms. It may cracking and postcracking stages.
be concluded from the agreement between the experimental 2. Both prism strength and initial modulus of elasticity
and predicted results that the assumptions made in the finite decrease with the increase in the number of courses. How-
element model led to the less pronounced effect of neglecting ever, an insignificant effect was encountered for both
the tapering of faceshells and webs on the grouted prisms. ungrouted and grouted prisms of more than three courses.
3. Analysis has shown that crack pattern, failure mode,
Conclusions compressive strength, and initial modulus of elasticity of
concrete masonry prisms depend on the relative rather than
A three-dimensional finite element model was developed to the individual strength and modulus of elasticity of the prism
study the behavior of ungrouted and grouted concrete masonry constituent materials.
91 4 Can. J. Civ. Eng. Vol. 22, 1995

Table 6. Comparison between experimental and predicted prism strengths.

Lb Am f,, f,,(observed)
Test (MPa) (MPa) (MPa) Observed Predicted f,,(predicted)
Cheema 25.86
25.86
Drysdale 27.90
25.10
25.10
25.10
20.20
Hamid 19.66
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

19.66
19.66
19.66
19.66
19.66
19.66
19.66
19.66
Hegemier 35.18
51.04
25.60
25.60
For personal use only.

References Hegemier, G.A., Krishnamorthy, G., Nunn, R.O., and Moorthy,


T.V. 1978. Prism tests for the compressive strength of concrete
American Concrete Institute. 1983. Building code requirements for masonry. Proceedings, North American Masonry Conference,
reinforced concrete structures. Detroit, Mich., ACI 318 M-83. University of Colorado, Boulder, Colo., Vol. 18, pp. 1 -7.
American Concrete Institute. 1988. Building code requirements for Swanson Analysis Systems Inc. 1990. Theoretical manual. ANSYS
masonry structures. Detroit, Mich., ACI 530-88lASCE 5-88. engineering analysis system, Version 4.4-a. Huntington Beach,
American Society for Testing and Materials. 1991. ASTM stan- Calif.
dards in building codes. 28th ed. Philadelphia, Pa.
Boult, B.F. 1979. Concrete masonry prisms testing. Journal of the
American Concrete Institute, 76: 5 13-535. List of symbols
Canadian Standards Association. 1970. Masonry code of practice.
Rexdale, Ont., CSA A224-1970. P C shear strength reduction factor for closed crack
Canadian Standards Association. 1984. Masonry design for build- Pt shear strength reduction factor for open crack
ings. Rexdale, Ont., CAN3-S304-M84. ~EP' plastic strain increment
Cheema, T.S., and Klingner, R.E. 1986. Compressive strength of Eel elastic strain
concrete masonry prisms. Journal of the American Concrete Eu ultimate strain at failure
Institute, 83: 88 -97. E0.2 secant modulus of elasticity at 0.2% strain
Drysdale, R.G., and Hamid, A.A. 1983. Capacity of concrete E0.2b secant modulus of elasticity of block units at 0.2%
block masonry prisms under eccentric compressive loading. strain
Journal of the American Concrete Institute, 80: 102- 108. secant modulus of elasticity of grout at 0.2% strain
Ghoneim, T.G.M. 1993. Behavior of concrete masonry prisms E0.2,
under axial compression. M.Sc. thesis, The American Univer- E0.2m secant modulus of elasticity of mortar at 0.2% strain
sity in Cairo, Cairo, Egypt. Ei component initial modulus of elasticity
Hamid, A.A., and Chukwunenye, A.O. 1986. Compression Eib initial modulus of elasticity of block units
behavior of concrete masonry prisms. ASCE Journal of Struc- Eig initial modulus of elasticity of grout
tural Engineering, 112: 605 -613. Eim initial modulus of elasticity of mortar
Hamid, A.A., Drysdale, R.G., and Heidebrecht, A.C. 1978. Effect EiP initial modulus of elasticity of prism
of grouting on the strength characteristics of concrete block Er relative prism modular ratio
masonry. Proceedings, North American Masonry Conference, secant modulus of elasticity of block units at failure
University of Colorado, Boulder, Colo., Vol. 11, pp. 1-17.
Harnid, A.A., and Drysdale, R.G. 1979. Suggested failure criteria
for grouted concrete masonry under axial compression. Journal
3
Esb
secant modulus of elasticity of prism at failure
function of the principal stress state
of the American Concrete Institute, 76: 1047- 1061. fbm block-mortar strength ratio
Hedstrorn, E.G., and Hogan, M.B. 1990. The properties of fc component compressive strength
masonry grout in concrete masonry. Masonry components to fcb compressive strength of block units
assemblages. American Society for Testing and Materials, fc, compressive strength of grout
Philadelphia, Pa., ASTM STP 1063. fcm compressive strength of mortar
Fahmy and Ghoneim

f,, compressive strength of prism block -grout modular ratio


fgroSsblock compressive strength based on the gross cross- block -mortar modular ratio
sectional area grout -mortar modular ratio
f,,, block compressive strength based on the net cross- relative modular ratio
sectional area grouted prism relative strength
f ungrouted prism relative strength tolerance of plasticity ratio at convergence
fjb relative strength ratio failure surface
f, material tensile strength vertical stress along the block units
ftb tensile strength of block unit material vertical stress along the grout
f,, tensile strength of grout vertical stress along the prism
f, tensile strength of mortar unit weight of block unit
Lb ultimate compressive strength of block unit material
v net-to-gross cross-sectional area of block units (based
Can. J. Civ. Eng. Downloaded from www.nrcresearchpress.com by 70.53.64.150 on 06/04/14

on the minimum cross-sectional area of block unit)


For personal use only.

Das könnte Ihnen auch gefallen