Vous êtes sur la page 1sur 8

Available online at www.rilem.


Materials a n d Structures 38 (March 2005) 249-256

Failure mechanism of masonry prism loaded in axial

compression: computational aspects

L. Berto 1, A. Saetta 2, R. Scotta 1 and R. Vitaliani 1

(1) Department of Construction and Transportation, University of Padova, Italy
(2) Department of Architectural Construction, IUAV - Venice, Italy

Received." 22 April 2004; accepted." 26 June 2004

In this paper, the mechanical behaviour of masonry prism under typical laboratory testing conditions, such as uniaxial compression, was
simulated by means of finite element analysis by both two-dimensional and three dimensional models. The analyses were performed by assuming
damage constitutive law for the component materials (mortar and bricks). The effect of lateral stresses arising from the Poisson's effect is
emphasised and the main failure mechanism is identified, demonstrating the inability of usual plain-stress and plain-strain 2-D analysis to capture
the actual structural response in the most common practical cases. A modified 2-D model accounting for the out-of-plane interaction between
mortar and units is therefore proposed and comparative stress-strain diagrams are presented.
1359-5997 9 2004 RILEM. All rights reserved.

Dans cet article, on analyse le comportement m~canique de briques sous chargement &pique de laboratoire, comme par exemple compression
uniaxiale, grdce (I la m(thode des Ol~mentsfinis avec des modkles 21) et 3D. L 'endommagement est pris en compte pour les dew: mat(riaux (briques et
mortier). On met en avant les effets de tension lat~rale provenant du coeflqcient de Poisson et on identifie les m(canismes principaux de rupture,
d~montrant l 'inccqJacit(des rnodkles 2D de tension plane de reproduire le comportement du mate'riau. On propose donc an modkle modifi~ 21:)qui prend
en eompte les interactions hors du plan entre le mortier et une brique, et on prOsente une comparaison tension-d~formation entre les deux modkles.

1. I N T R O D U C T I O N the bricks. As a consequence, in such a mode of failure the

compressive strength of the masonry may greatly exceed the
The behaviour of masonry subjected to compression in the cube crushing strength of the mortar, while it remains lower
direction normal to the bed joints has widely been studied and than the compressive strength of the unit.
many investigations, based on experimental as well as Some interesting experimental tests about the effect of the
numerical approaches, (e.g. [1-5]) have been carried out. properties of bed materials have been conducted e.g. by the
The failure of masonry in compression depends basically Structural Clay Products Research Foundation in the United
on the interaction of unit and mortar joint as the result of their States [7], by Morsy [8], and by Binda et al. [3].
different deformation characteristics. In particular, the The other basic factor affecting the masonry compressive
difference of the elastic properties of the component materials strength (and the correspondent failure mode) is the transverse
strongly influences the failure mode which can cause either tensile strength of the units. It is influenced by the clay quality,
tension cracks parallel to the direction of loading or a kind of the fLring temperature, the porosity, etc, Schubert in [9] has
shear failure along some lines of weakness [6]. The latter evidenced some other factors which may be fundamental in
occurs in some cases when the mortar mechanical characterizing the tensile strength of the unit (e.g. the inner
characteristics are similar, or even greater than, the unit ones stresses caused by firing and cooling, as well as cracks due to
(i.e. certain types of concrete brickwork), while the first mode draining and firing for clay bricks; the matrix strength -
of failure, i.e. the usual tensile splitting failure, always occurs binding agents, finest material, water - as well as the bonding
when the units, which are stiffer than mortar, restrain the with the aggregates for calcium silicate and concrete units; the
lateral defornaation of the mortar leading to a state of triaxial percentage and pattern of perforation in case of perforated
compression in the mortar and compression/biaxial tension in units). All these parameters together with the texture, and the

1359-5997 9 2004 RILEM. All rights reserved.

250 L. Berto et al. / Materials' and Structures 38 (2005) 249-256

geometrical characteristics of masonry panel influence its

compressive behaviour, so that various experimental results
have been carried out during the years.
Anyway, the response of masonry under ulaiaxial
compression is the object of a long-going debate among
researches and is out of the aims of this study. A discussion
can be found in [6, 9, 10].
The goal of the present paper is to investigate whether the
proposed numerical analyses can reproduce the actual response
of masonry under compression. Moreover, the effect of the
presence of out-of-plane stresses oz is carefully analysed, by
considering the usual case with mortar weaker than units and its ~y
complementary case, although this one rarely occurs. Therefore
finite element analyses were performed to characterize the Fig. 1 - Brick-mortar prism subjected to compression orthogonal
mechanical and structural behaviour of a masonry prism in to the bed joints.
compression, with particular attention to the lateral stress
effects. The study has been carried out by assuming a non linear component materials and a uniform stress state holding in
behaviour for both the mortar and bricks. the specimen.
It is worth nothing that the importance of the out-of-plane With such hypotheses the lateral strains in the bricks and
stresses in the vertical uniaxial compression of masonry and in the mortar joints in the x and z directions can be easily
the benefit derived from numerical methods able to capture derived:
them have been underlined by Pegon and Anthoine in [11],
b_ 1 b
who have preferred the generalized plane stress analysis
instead of the plane stress, and by Lourengo [12].
Within the isotropic damage approach, firstly a fully 3D b 1 b
analysis is carried out in order to capture the real behaviour of (1)
the sample; then a usual 2D stress plane analysis is considered.
Finally the same 2D analysis is enhanced by implementing in gxm = F o1mr n[-E • _ v r n ( ~ y + O . 2 ) l
the finite element code the formulae which provide the lateral
stresses ~z in mortar and brick. The obtained results are then s = FOmEm
1 [ z -vm(Oy+Om)l
These analyses have been performed for two cases where Eb and ETM are the Young's moduli of brick and
characterised by a different ratio between the elastic properties mortar respectively, and v b and v m the corresponding
of the masonry components. The comparison of the obtained Poisson's ratios.
results shows that only a numerical analysis which takes into As the lateral strains in the bricks and mortar at their
account the presence of stresses crz is able to reproduce the real interface are the same, we have:
compressive strength and failure mode of masonry in
whichever situation, while the usual 2D plane stress analysis b m and b m
ex = gx ~z = ez (2)
may significantly underestimate the peak strength and
incorrectly predict the failure mode in the more common Besides, for the equilibrium, the total lateral force in the
situations in which the mortar is weaker than the units. mortar must be equal to the ones in the bricks; hence, by
Obviously, the 2D plane strain analysis approach is not assuming that each of the component materials is subjected
suitable for the analysis of masonry, since it leads to to a uniform stress state, it follows:
overestimate the compressive stresses % in both the mortar m b
and bricks. Finally, it is worth underlining that the proposed % = - ~ ~x (3)
analyses deal with the mechanical behaviour of brickwork type m b
(5"z = -Or (5"z
masonry, characterized by a ratio of joint thickness to unit
height ranging from 0.15-0.2. where c~ is the ratio of the height of the unit and the
thickness of the mortar joints.
The substitution of the formulae(l) together with (3) in
2. COMPRESSION BEHAVIOUR BASED Equations (2), with the introduction of the dimensionless
ON ELASTIC ANALYSIS parameter t3 = Eb/E ra, gives:

b _ b O'Y( -13Vm + v b )
A simple and very intuitive way to estimate the lateral
(5"x -- r =
stresses induced by a vertical compression load in a 1 + a[~ --Vb -- al~V m (4)
masonry prism can be derived from the classical continuum
conditions (e.g. [6, 13]). m _ m =

Let us consider a brick-mortar prism subjected to an ~x -- ~ z

1 + otl3 --V b -- U,~V m
axial compressive test along y direction (Fig. 1) by
assuming a linear elastic isotropic hehaviour for the that provide the relationships between the lateral stresses in the
L. Berto et al. / Materials and Structures 38 (2005) 249-256 251

the masonry structures as being in a state of plane stress, which

~y is a typical condition for panels and shear walls.
The properties of masonry to be introduced in the macro-
Cyb models may be derived from specific experimental tests as
(Yx z well as from homogeneization techniques or by means of
b ~x micro-modeling analyses. In the two last cases, the evaluation
of masonry properties should be performed by considering the
Gy importance of the 3D effect for masonry subjected to uniaxial
in compression, even if full three- dimensional analyses can not
O'y in
in (3"z
be afforded because too expensive in computing terms. In the
(Yx In case of homogeneization approach, such a concept is stressed
m O"x also by Pegon and Anthoine in [11] and Lourengo in [12].
in In the following the micro-modelling technique is adopted
to analyse the response of masonry under compression.
Bricks and mortar are represented by different elements
Fig. 2 - State of stress in the case in which the mortar is weaker whose non linear behaviour is described by means of an
than the units. isotropic damage model. To this aim a full 3D analysis is
component materials and the vertical compressive stress Cry. carried out so that the effect of the out of plane stresses can
In the usual cases characterised by mortar weaker than be captured; then, an enhanced 2D stress plane analysis is
units, from the formulae (4) we can obtain %b = %b >0 and proposed, which includes the 3D effect without turning to a
%m = crm<0 , that lead to a triaxial compression state in the more expensive three dimensional analysis.
mortar and compression/biaxial tension state in the bricks, as
shown in Fig. 2. As a consequence the failure occurs by
splitting in the units with the usual development of tension
cracks parallel to the axis of loading.
In order to investigate the compressive response of
It is worth noting that a compressive vertical load
masonry the non linear behaviour of the component
induces lateral stresses crx (in the plane of masonry) and Crz
materials (mortar and bricks) has been simulated by means
(out of plane of masonry) that are equal to each other.
of a phenomenological isotropic damage model based on
Hence both of them have to be taken into account to
the hypothesis of"strain equivalence"([29-31]).
properly simulate the behaviour of the material.
This model is able to describe the different behaviour of the
material under tensile and compressive loading by introducing
3. NUMERICAL M O D E L L I N G OF two internal independent damage variables d+ and d-(which
vary between 0 and 1). According to this formulation, the
effective stress tensor is split into two components N+ and
N- related to its tensile and compressive components:
Depending on the level of accuracy and the
computational effort required, the numerical analysis of 3
masonry can be performed by following two different
approaches: the micro and the macro-modelling. i=1
Micro-modelling is probably the best tool available to 3 (5)
analyse and understand the real behaviour of masonry, ~- ---- ( - - ~ ) : Z (--~i)Pi | P i = cr -O'+
particularly concerning its local response. Within such an i=1
approach is possible to characterise separately mortar,
blocks and their interfaces, adopting suitable constitutive where Pi ([[Pill = 1) is the unit vector for the direction of the im
laws for each component, which take into account their principal stress, and symbol /. ./. is the MacAuley operator,
different mechanical behaviour ([ 14-19]). which returns the value of the enclosed expression if positive,
On the other side, when large real structures have to be a zero value if negative.
studied in order to capture their global response, the macro- Moreover two equivalent effective stresses are introduced,
modelling is often the only effective option. Within such an one in tension:
approach, masonry is regarded as an equivalent material,
where mortar and blocks are melted together, and appropriate z = X/~+ : ~+ (6)
relations are established between averaged masonry strains and
averaged masonry stresses. A number of such models have and one in compression:
been developed (e.g. [20-23]), the most complete reproduce
the masonry as a macroscopically orthotropic material (7)
characterised by a non linear behaviour for each materials axis.
Some of these ([24-27]) have been developed in the where ~o-ct and Z-oct are respectively the octhaedreal normal
framework of damage mechanics, other ones (e.g. [12, 28]) are
stress and the octhaedreal shear stress obtained fi'om N - , and
based on the plasticity theory. Generally these models consider
252 L. Berto et al. / M a t e r i a l s and Structures 38 (2005) 2 4 9 - 2 5 6

K is a material property which is related to the strength depicted in Fig. 5.

increase observed in the biaxial and triaxial compression state. The compressive test is simulated by imposing a history
This set of equations gives a sort of Drucker -Prager failure o f monotonically increasing vertical displacements on the
criterion in the triaxial compression field with a cap - closure top face, while the nodes on the bottom face are restrained
in the tensile field. The intersection o f the damage surface with in the direction o f the load.
the ~3 = 0 plane is depicted in Fig. 3. In the following two different cases will be studied
The assumed constitutive law can be written in the form: characterised respectively by mortar weaker or stronger than
the units. The first case (i.e. mortar weaker than bricks) is the
most common in classical brickwork; it will be analysed by
using the previously described damage model and by
The evolution laws o f the damage variables d + and d are assuming for the constituent materials the properties
expressed respectively as function of ~+ and ~- : summarised in Table 1. The second analysis is performed by
considering the strength and the Young's modulus E m o f the
+ mortar 25 % higher than the units ones, and the Poisson's
d + = l-r~ -e ~" ~" tO)) moduli v m = v b =0.1.
u (9)
5.1 F i r s t case : m o r t a r w e a k e r t h a n u n i t s
rE l x
d- = 1- ~--x--.,I-B-,-B-.e -
5.1.1 319 analysis
where A + is dependent on some characteristics of the material In order to accurately simulate the real response o f the
and of the adopted mesh, A and B are material parameters specimen a three-dimensional numerical analysis has been
which describe the compressive constitutive law.
Fig. 4 shows the constitutive curves ~ - e for both cases of
uniaxial tension and compression.
All the details about the formulation o f this model and the
strategy adopted to control the mesh-dependency are described 115
in [31-33].



We are focusing on the 115 m m thick masonry pr~fn

Fig. 5 - Case study: masonry prism.
Table 1 - Material parameters of the tested specimen
mortar brick
Young's modulus (MPa) E = 6000 E = 15000

Poisson's ratio v =0.30 v = 0.10

Uniaxial linear-elastic limit fD =3.5 f ~ =14.9

Fig. 3 - Initial damage surface adopted in the model. in compression (MPa)
Uniaxial compressive strength ~'lV =5 fl~) = 14.9
Uniaxial tensile strength (MPa) Jo+ =1.3 J'0+ = 1.4

Fracture energy (N/mm) Gf=0.06 G.f= 0.08

Compression parameter A A- =1 A- = 0.9

Compression parameter B B- =1.1 B_ = 1.6

Fig. 4 - Constitutive law adopted in the model. Rate fzD /flD R0 = 1.25 R0 = 1.1
L. Berto et al. / Materials and Structures 38 (2005) 249-256 253

Fig. 6 - 3D finite element mesh. Fig. 8 - Bi-dimensional analysis. Finite element mesh.

Fig. 7 - First case. 3D analysis - Contours of the damage variables at

failure: (a) tensile damage d+; (b) compressive damage d. Fig. 9 - First case. Plane stress analysis - Contours of the damage
variables at failure: (a) tensile damage d§ (b) compressivedamage d-.
The modelling of the sample is carried out with a structured take into account the occurrence of the out of plane stresses Cyz.
mesh made up of with 3400 8-noded brick element. The As a consequence, this technique underestimates the real
adopted finite element mesh is depicted in Fig. 6. compressive strength of the mortar (which is, in this case,
The damage contours at failure are shown in Fig. 7. As subjected to a triaxial compressive state of stress) and delays
expected, although the mortar joints are rather damaged in the tensile splitting failure of the bricks.
compression, the failure is governed by the tensile damage 5.1.3 21) analysis- enhanced plane stress approach
in the central brick, while the concurrent lower compressive
damage is due to the tensile-compressive behaviour of the On the basis of the previous considerations a new 2D
brick. Hence the resulting failure mode is tensile splitting analysis has been carried out by implementing in the numerical
on the vertical plane in the middle of the central brick, code the formulae (4) in order to consider the out of plane
which is the usual mode for the case characterised by stresses ~z. In such a way we have developed a method, which
mortar weaker than units. could be called "enhanced plane stress", able to macro-
scopically reproduce in the 2D framework the triaxial effect.
5.1.2 2D analysis -plane stress approach The damage contours obtained by this third numerical
The same analysis has been repeated by following a 2D analysis are given in Fig. 10. It is evident the similarity
approach. In particular, as usual in the numerical modelling between these results and the ones obtained with the "3D
of masonry structures, initially a plane stress analysis has analysis".
been performed. The tensile damage is concentrated in the middle of the
The mesh adopted for the bi-dimensional analysis is central brick, which fails when its tensile strength is
depicted in Fig. 8. It consists of 340 rectangular four noded exhausted, in a completely different way with respect to the
elements. usual 2D analysis.
The obtained results are shown in Fig. 9, in terms of 5.1.4 Results and discussion
damage contours at failure. It is evident that the failure is
connected with the compression damage of the bed joints, A comparison between the results of the three numerical
while the tensile damage is almost negligible and localised tests is given in Fig. 11 in terms of stress-strain curves and
in some edge zones. the following remarks can be made.
It is worth noting that the plane stress analysis, due to its Firstly, as expected (see section 2), the compressive
own characteristics, may not be able to capture the real strength of the sample, which is carried out by the 3D analysis,
behaviour of the masonry in compression, because it does not is nearly 8.5 N/mm 2, which is lower than the compressive
254 L. Berto et al. /Materials and Structures 38 (2005) 249-256

Fig. 10 - First case. Enhanced plane stress - Contours of the Fig. 12 - Second case. 3D analysis - Contours of the damage
damage variables at failure: (a) tensile damage d+; (b) compressive variables before the failure: (a) tensile damage d+; (b) compressive
damage d-. damage d-.

Fig. l 1 - First case. Curve r comparison between the

numerical results.
Fig. 13 - Second case. Enhanced plane stress analysis - Contours
strength of the bricks (14.9N/mm2), but higher than the of the damage variables before the failure: (a) tensile damage d+;
uniaxial compressive strength of the mortar (5 N/mmZ). (b) compressive damage d-.
Moreover, we can observe that, in spite of its simplified
formulation, the enhanced plane stress analysis is able to Contrary to the first case, in this situation the mortar,
properly capture the compressive masonry behaviour both in because of its higher stiffness, tends to reduce the lateral
terms of failure mode (as it is proved by the damage contours) strains in the bricks, leading to a state of compression-biaxial
and of peak strength (as proved by the stress-strain diagram). tension in the mortar mad of triaxial compression in the units.
Finally, we can remark that the application of the usual According to such a stress state the numerical results before
"plane stress analysis" in the simulation of the compressive the failure, which are given in terms of damage contours in
response of the masonry may lead to completely incorrect Figs. 12 and 13 respectively for the 3D and the 2D analysis,
results. In particular, in this first case such a kind of show a significant tensile damage in the mortar joints.
analysis not only has underestimated the real compressive The increasing of this damage in the joints is related to a
strength of the sample (nearly one third lower) but has reduction of the stiffness characteristics of the mortar
predicted a failure mode substantially different from the resulting in a reduction of its restraining action. As a
real one (failure due to compression of the bed joints consequence, the collapse is not directly connected with the
instead of tensile splitting of the central brick). damage in the mortar, but, according to what is stated in [6],
a kind of shear failure occurs, as evidenced by the damage
contours in Figs. 14 and 15 and by the deformed mesh at
5.2 S e c o n d case : m o r t a r stronger than units
failure given in Figs. 16 and 17.
To complete this study the three types of numerical Such a behaviour can be ascribed to the phenomenon of
analyses (i.e. 3D - 2D plane stress- 2D enhanced plane stress shear failure which can be found in multiaxial compression
analyses) have also been carried out for the case in which the test of brittle materials, e.g. [34].
mortar is stronger than the units. This situation rarely occurs It is worth noting that the presence of mortar with strength
in the common construction where masonry is usually greater than unit strength, has the effect of restraining lateral
characterised by mortar properties lower or similar to those deformation of the bricks and failure becomes similar of that
of the units. For such a reason in this case we have assumed attainable in a compression test on a brittle material, e.g. [6].
the strength and Young's modulus of the 25% higher than the Similarly Priestley [35] has found that, incorporation of
bricks ones. steel confining plates in the mortar beds dramatically changed
L. Berto et al. /Materials and Structures 38 (2005) 249-256 255

Fig. 17 - Second case. Enhanced plane stress analysis -

Fig. 14 - Second case. 3D analysis - Contour of the damage
Deformed mesh at failure.
variables at failure: (a) tensile damage d+; (b) compressive
damage d-.

Fig. 18 - Second case. Curve ~y-e: comparison between the

Fig. 15 - Second case. Enhanced plane stress analysis - Contours numerical results.
of the damage variables at failure: (a) tensile damage d+;
that, because of the triaxial compression state in the units, the
(b) compressive damage d-.
compressive strength o f the sample is nearly 16.3 N/ram 2,
which is higher than the tmiaxial compressive strength o f the
brick (14.9 N/ram2).
As in this case the results of the plane stress analysis are
substantially identical to the ones o f the "enhanced plane stress
analysis", for the sake o f brevity the damage contours related
to the first analysis have been omitted, and only the ones
obtained with "the enhanced model" and the "3D model" have
been shown.
The good agreement found between the numerical
results allow to conclude that in such a case the effect o f the
out-of- plane stresses is almost negligible, and for this
reason a simple 2D "plane stress analysis" is able to capture
the real behaviour o f masonry both in terms o f stress-strain
curves and o f failure mode.
Fig. 16 - Second case. 3D analysis - Deformed mesh at failure.

the physical appearance o f the failure mechanism: vertical 6. CONCLUSION

splitting of the blocks was virtually eliminated, with a shear-
compression failure occurring. In lmmerical applications it is commonly adopted a
In this second case, due to its own peculiarity (i.e. mortar description of masonry walls via a plane-stress model. In this
stronger than units) and to the reduced difference existing paper the capability o f such an approach to simulate the
between the elastic properties o f the two component materials, masonry response in compression has been investigated. The
the occurrence of the out o f plane stresses ~z is not essential to results emphasise the need to consider the presence o f the out o f
be considered, as it is proved by the good agreement between plane stresses in the usual case characterised by mortar weaker
the results o f the three numerical analyses given in Fig. 18 in than units. It has been shown that in such a case neglecting
terms of ~-~ curves. Moreover this diagrams allow to note these effects can lead to unacceptable prediction both in term of
256 L. Berto et al. / Materials and Structures 38 (2005) 249-256

peak strength and failure mode. In order to overcome this [16] Loffi, H.R. and Shing, B.P, 'Interface model applied to
problem, without giving up the effectiveness and the efficiency fracture of masonry structures', J. Struc. Engng, ASCE 120 (1)
of the bi-dimensional analysis, a modified 2-D model has been (1994) 63-80.
introduced and validated with the results of a fully 3D analysis. [17] Louren~o, P.B. and Rots, J.G., 'Multisurface interface model
for analysis of masonry structures', J. Engng Mech., ASCE
123 (7) (1997) 660-668.
ACKNOWLEDGEMENTS [18] Zucchini, A. and Louren~o, P.B., 'A micro-mechanical model
for the homogenisation of masonry', Int. J. Solids and
Structures 39 (2002) 3233-3255.
The financial support of M I U R PRIN 2003 is gratefully [19] Gambarotta, L., Lagomarsino, S. and Morbiducci, R., 'Two-
acknowledged. dimensional finite element simulation of a large scale brick
masonry wall through a continuum damage model' In
'Experimental and numerical investigation on a brick masonry
REFERENCES building prototype' (Report 3.0 CNR GNDT, 1995).
[20] Page, A.W., Kleeman, P.W. and Dhanasekar, M., 'An in-plane
[1] Francis, A.J., Horman, C.B. and Jerrems, L.E., 'The effect of f'mite element model for brick masonry', in 'New Analysis
joint thickness and other factors on the compressive strength Techniques for Structural Masonry' (S.C.Anand, Ed., Proc.
of brickwork', Proc. of the 2nd Int. Brick Masonry Conf., 1971 Struct. Congress, ASCE, Chicago, Ill., 1985) 1-18.
(eds. H.W.H.West & K.H. Speed, British Ceramic Research [21] Andreaus U., 'Failure criteria for masonry panels under in-
Association, Stoke-on-Trent, 1971) 31-37. plane loading', J.Struct.Engrg., ASCE 122 (1) (1996) 37-46.
[2] Hilsdorf, H.K., 'An investigation into the failure mechanism [22] Luciano, R. and Sacco, E., 'A damage model for masonry
of brick masonry loaded in axial compression', in 'Designing, structures', Eur. J. Mech., A/Solids 17 No 2 (1998) 285-303.
Engineering and Constructing with Masonry Products' (ed. [23] Syrmakezis, C.A. and Asteris, P.G, 'Masonry failure criterion
F.B. Johnson, Gulf, Houston, Tex., 1969) 34-41. under biaxial stress state ', J. Materials in Civil Engineering 13
[3] Binda, L., Fontana, A. and Frigerio, G., 'Mechanical No.1 (2001) 58-64.
behaviour of brick masonries derived from unit and mortar [24] Papa, E. and Nappi, A., 'Numerical modelling of masonry
characteristics', in Proc. of the 8th International Brick/Block structures subjected to cyclic loads', In 'Masonry mechanics
Masonry Conf, Dublin, 1 (1988) 205-216. between theory and practice', Proc. of the National Conf.,
[4] Houston, J.Y. and Grimm, C.T., 'Effect of the brick height on Messina, Italy, 1996 (L .Gambarotta ed. 1996) 441-450 [only
masonry compressive strength', J. Mater., ASTM 7 (1972) available in Italian].
388-397. [25] Callerio, A. and Papa, E., 'An elastic-plastic model with
[5] Page, A.W., Simundic, G. and Han Xie, 'A study of the damage for cyclic analysis of masonry panels', In 'Computer
relationship between unit, prism and wall strength for hollow Methods in Structural Masonry-4' (Pande, Middleton and
masonry loaded in compression', in Proc. of the 9th Int. Kralj eds., published by E & FN Spun, London, 1998).
Brick/Block Masonry Conf., Berlin, Germany (1991). [26] Berto, L., Saetta, A., Seotta, R. and Vitaliani, R., 'An
[6] Hendry, A.W., 'Structural Masonry', 2 nd Edn. (MacMillan orthotropic damage model for masonry structures', Int. J. for
Press LTD, London, 1998). Numerical Method in Engineering 55 (22) (2002) 127-157.
[7] Monk, C.B., 'A historical survey and analysis of the [27] Massart, Th., Bouillard, Ph., Geers, M.G.D. and Peerlings,'
compressive strength of brick masonry', research Rep. No12 R.H.J., 'A 2D anisotropic damage model for masonry walls',
(Structural Clay Products Research Foundation, Geneva Ill., Proc. of V European Conf. on Computational Mechanics,
1997). Krakow, Poland, (2001).
[8] Morsy, E.H., 'An investigation on mortar properties [28] Louren~o, P.B., Rots, J.G. and Blaauwendraad, J., 'Continuum
influencing brickwork strength', PhD Thesis (University of model for masonry: parameter estimation and validation', J. of
Edinburgh, 1968). Struc. Engng. 124 (6) (1998) 642-652.
[9] Schubert, P., 'Compressive and tensile strength of masonry', [29] Lemaitre, J. and Caboche, J.L., 'Mrcanique des Matrriaux
in Proceedings of the 8ta International Brick/Block Masonry Solides' (Dunod, Paris, 1988) 343-444.
Conference, Dublin, 1 (1988)406-420. [30] Simo, J. C. and Ju, J. W., 'Strain and stress based continuum
[10] Shrive, N.G., 'Compressive strength and strength testing of damage models - part I: formulation', Int. ~ Solids and
masonry ', in Proceedings of the 7th International Brick/Block Structures 23 (7) (1987) 821-840.
Masonry Conference, Melbourne (1985) 699-710. [31] Saetta, A, Scotta, R. and Vitaliani, R. 'Mechanical Behaviour
[11] Pegon, P. and Anthoine, A., 'Numerical strategies for solving of Concrete under Physical-Chemical Attacks', J. Engng.
continuum damage problems with softening: application to the Mech., ASCE 124 (10) (1998) 1100-1109.
homogenization of masoiary', Computer and Structures 64 (1-4) [32] Farja, R. and Oliver, J., 'A strain-based viscous-plastic-
(1997) 623-642. damage model for massive concrete structures', Int. J. Solids
[12] Louren~o, P.B., 'Computational Strategies for Masonry and Structures 35 (1998) 1533-1558.
Structures' (Delft University Press, The Netherlands, 1996). [33] Berto, L., 'Damage models for orthotropic materials:
[13] Del Piero, G., 'Constitutive equations' in 'Masonry application to masonry', Dissertation, (University Padua,
constructions' (by G. Del Piero, Intern. Centre for Mechanical Padua, Italy, 2003) [only available in Italian].
Sciences, Udine, Italy, 1984) [only available in Italian]. [34] Vonk, R.A., 'Softening of concrete loading in compression',
[14] Page, A.W, 'Finite element model for masonry', J. of the Dissertation, (Eindhoven University of Technology,
Structural Division, ASCE 104 (No ST8) (1978) 1267-1285. Eindhoven, The Netherlands, 1992).
[15] Ali, S.S. and Page, A.W., 'Finite element model for masonry [35] Priestley, M.J.N. and Elder, D.M., 'Stress-Strain Curves for
subjected to concentrated loads', a~ of Structural Engineering, Unconfined and Confined Concrete Masonry', ACI Journal
114 (8) (1988) 1761-1784. May-June (1983) 192-201.