See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/302488618

Masonry Modeling

Chapter · January 2015

DOI: 10.1007/978-3-642-35344-4_153

CITATIONS READS

2 2,204

1 author:

Paulo B. Lourenco
University of Minho
1,105 PUBLICATIONS   20,263 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Initiative TUA View project

ProTimber - Probabilistic Assessment of Existing Timber Structures View project

All content following this page was uploaded by Paulo B. Lourenco on 19 May 2016.

The user has requested enhancement of the downloaded file.

Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Masonry Modeling
Paulo B. Lourenço*
Department of Civil Engineering, University of Minho, ISISE, Guimarães, Portugal

Synonyms
Computational strategies; Finite element modeling; Numerical modeling; Structural analysis

Introduction
This entry deals with modeling of masonry structures. In general, the approach towards the
numerical representation of masonry can focus on the micro-modeling of the individual compo-
nents, units and mortar, or the macro-modeling of masonry as a composite (Rots 1991). Much effort
is made today in the link between the micro- and macro-modeling approaches using homogenization
techniques.
Masonry joints act as planes of weakness, and the explicit representation of the joints and units in
a numerical model seems a logical step towards a rigorous analysis tool (Fig. 1). This kind of
analysis is particularly adequate for small structures, subjected to states of stress and strain strongly
heterogeneous, and demands the knowledge of each of the constituents of masonry (unit and mortar)
as well as the interface. In terms of modeling, all the nonlinear behavior can be concentrated in the
joints and in straight potential vertical cracks in the centerline of all units. In general, a higher
computational effort ensues, so this approach still has a wider application in research and in small
models for localized analysis.
The approach based on the use of averaged constitutive equations seems to be the only one
suitable to be employed a large-scale finite element analyses. Two different approaches are illus-
trated in Fig. 2, one collating experimental data at average level and another from homogenization
techniques. A major difference is that homogenization techniques provide continuum average
results as a mathematical process that includes the information on the microstructure.
These three different approaches, namely, micro-modeling, macro-modeling, and homogeniza-
tion, are reviewed next.

Micro-Modeling
Different approaches are possible to represent heterogeneous media, namely, the discrete element
method, the discontinuous finite element method (FEM), and limit analysis. The finite element
method remains the most used tool for numerical analysis in solid mechanics, and an extension from
standard continuum finite elements to represent discrete joints was developed in the early days of
nonlinear mechanics. Interface elements were initially employed in concrete and rock mechanics,
being used since then in a great variety of structural problems. A complete micro-model must
include all the failure mechanisms of masonry, namely, cracking of joints, sliding over one head or

*Email: pbl@civil.uminho.pt

Page 1 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 1 Examples of cracking at failure using structural micro-modeling: (a) shear wall with opening; (b) wall subjected
to out-of-plane loading; (c) retaining wall

a 750m
m
mm
3300
3404mm

Homogenization

Basic cell Homogenized

continuum

Fig. 2 Constitutive behavior of materials with microstructure: (a) collating experimental data; (b) mathematical process
using geometry and mechanics of components

Page 2 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 3 Multi-surface interface constitutive model

bed joint, cracking of the units, and crushing of masonry, as done in (Lourenço and Rots 1997) for
monotonic loading and in (Oliveira and Lourenço 2004) for cyclic loading.
The typical characteristics of discrete element methods are (a) consideration of rigid or FEM
deformable blocks, (b) connection between vertices and sides/faces, (c) interpenetration usually
possible, and (d) explicit integration of the equations of motion for the blocks using the real damping
coefficient (dynamics) or artificially large (statics). The main advantages are the formulation for
large displacements, including contact update, and an independent mesh for each block, in case of
deformable blocks. The main disadvantages are the large number of contact points required for
accurate representation of interface stresses and a time-consuming analysis, especially for 3D
problems. Masonry applications can be found in (Lemos 2007).
Computational limit analysis received far less attention from the technical and scientific commu-
nity for blocky structures. Still, limit analysis has the advantage of being a simple tool, while having
the disadvantages that only collapse load and collapse mechanism can be obtained and loading
history can hardly be included. A limit analysis constitutive model for masonry that incorporates
nonassociated flow at the joints, tensile, shear, and compressive failure and a novel formulation for
torsion is given (Orduña and Lourenço 2005a, b). The salient characteristics of limit analysis are
(a) rigid blocks, (b) interpenetration not allowed, and (c) mathematical formulation that leads to an
optimization problem (linear or nonlinear). The main advantages of the technique are adequate
formulation for design problems (requires a low number of parameters) and fast analysis. The main
advantages are that only the collapse load and mechanism can be obtained, tensile strength cannot be
included in the analysis, and the introduction of the loading history remains a challenge.
As an example of the possibilities that can be achieved with micro-modeling, a powerful interface
model is applied to illustrative examples. A constitutive interface model can be defined by a convex
composite yield criterion (see Fig. 3), composed by three individual yield functions, usually with
softening included for all modes so that experimental observations can be replicated. The three
functions include the usual Coulomb friction criterion for shear, combined with tension and
compression cutoffs. The adoption of appropriate inelastic constitutive relations and evolution
rules makes possible to reproduce nonlinear behavior during loading, unloading, and reloading,
consisting of exponential softening for tension and shear, and parabolic hardening followed by
exponential softening in compression (see Fig. 4). In this case, isotropic and kinematic hardening
laws are used, and aspects related to the algorithm can be found in (Oliveira and Lourenço 2004).
The ability of the model to reproduce the main features of structural masonry elements is shown in
Fig. 5. Shear walls have been used traditionally to validate constitutive models and to characterize

Page 3 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 4 Comparison between experimental and numerical results for uniaxial testing: (a) tension; (b) compression;
(c) shear; (d) tension–compression

masonry structural solutions, due to the fact that the most demanding demand for a masonry building
with box behavior usually leads to a combination of vertical and horizontal loading.
For the first wall considered, it was found that the geometric asymmetry in the microstructure
(arrangement of the units) influenced significantly its structural behavior. Note that, depending in the
loading direction, the masonry course starts either with a full unit or only with half unit. Figure 5a
shows that the monotonic collapse load is 112.0 kN in the LR direction and 90.8 kN in the RL
direction, where L indicates left and R indicates right. The cyclic collapse load is 78.7 kN, which
represents a loss of about 13 % with respect to the minimum monotonic value but a loss of about
30 % with respect to the maximum monotonic value. This demonstrates the importance of cyclic
loading but also the importance of taking into account the microstructure. It is also clear from these
analyses that masonry shear walls with diagonal zigzag cracks possess an appropriate seismic
behavior with respect to energy dissipation.
The second wall considered (Fig. 5b) is a tall wall that simply rocks in both ways. The highly
nonlinear shape of the load–displacement curve is essentially due to the opening and subsequent
closing, under load reversal, of the top and bottom bed joints. Similar deformed patterns, involving
the opening of extreme bed joints, were observed during the experimental test. Numerical results
show that the cyclic behavior of the wall is controlled by the opening and closing of the top and
bottom bed joints, where damage is mainly concentrated. The model also shows low (static) energy
dissipation, which is a consequence of the activated nonlinear mechanism (opening–closing of joints
and the absence of a diagonal shear crack). As shown, the failure of this wall is much different from
the previous wall, stressing the relevance of the internal structure.

Page 4 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 5 Results obtained with interface cyclic loading model for shear walls, in terms of force–displacement diagram and
failure mode: (a) wall failing in shear; (b) wall failing in bending

Macro-Modeling
Difficulties of conceiving and implementing macro-models for the analysis of masonry structures
arise especially due to the fact that few experimental results are available (either for pre- and post-
peak behavior), but also due to the intrinsic complexity of formulating anisotropic inelastic behavior.
Only a reduced number of authors tried to develop specific models for the analysis of masonry
structures (Dhanasekar et al. 1985; Seim 1994; Lourenço et al. 1998; Berto et al. 2002; Pelà
et al. 2011), always using the finite element method.
Formulations of isotropic quasi-brittle materials behavior consider, generally, different inelastic
criteria for tension and compression. The new model introduced in (Lourenço et al. 1997), extended
to accommodate shell masonry behavior (Lourenço 2000), combines the advantages of modern
plasticity concepts with a powerful representation of anisotropic material behavior, which includes
different hardening/softening behaviors along each material axis. The model includes the combina-
tion of a Rankine-like yield surface for tension and a Hill-like yield surface in compression (see
Fig. 6). The behavior of the model in uniaxial tension and uniaxial compression, along two
orthogonal directions, is given in Fig. 7. This composite surface permits to reproduce the results
obtained in uniaxial tests, in which different behaviors are obtained along different directions. The

Page 5 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 6 Continuum failure surface for masonry (plane stress representation)

Fig. 7 Behavior of the model for (a) tension and (b) compression, along two orthogonal directions

application of the model in structural modeling of masonry structures leads to excellent results, both
in terms of collapse loads and in terms of reproduced behavior (see Lourenço et al. 1998; Lourenço
2000).
Figure 8 shows the results of modeling another shear wall with an initial vertical pre-compression
pressure. The horizontal force F drives the wall to failure and produces a horizontal displacement d at
top. The wall is confined by two concrete slabs (top and bottom) and two masonry flanges (left and
right). This confinement and the large size of the wall make it appropriate for continuum modeling.
Initially, cracking occurs well distributed in the panel and finally concentrates in a single shear band
from one corner of the panel to the other. The compressive stresses are well below the crushing
strength of masonry, i.e., failure is dominated by tension. A complete discussion of the numerical
results has been given in (Lourenço et al. 1998). Figure 9 shows the results of modeling a panel with
out-of-plane pressure. The panel is simply supported on two sides (left and right), fully clamped on
one side (bottom), and free on the other (top). The central opening simulates a window, and the panel
was loaded with an air bag with a uniformly distributed load. The predicted form of collapse includes
diagonal cracks from each lower corner of the panel up to the opening, which were also observed in
the experiments. This form of yield line collapse does not mean that yield line design is safe due to
the quasi-brittle behavior of the material. A complete discussion of the numerical results has been
given in Lourenço (2000). A discussion on the application of these models to large historic
constructions is given in Lourenço (2002).

Page 6 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 8 Results for masonry shear wall: (a) load–displacement diagram; (b) predicted cracking pattern at peak and
ultimate load

Fig. 9 Results for a panel subjected to uniform out-of-plane loading: predicted cracking pattern at (a) bottom and (b) top
face of the panel

Homogenization Approaches
Homogenization techniques permit the establishment of constitutive relations in terms of average
stresses and strains from the geometry and constitutive relations of the individual components. This
can represent a step forward in masonry modeling because of the possibility to use standard isotropic
material models and data for masonry components, instead of the rather expensive approach of
testing large masonry specimens under homogenous loading conditions.
The complex geometry of the masonry representative volume, i.e., the geometrical pattern that
repeats periodically in space, means that no closed-form solution of the problems exists for running
bond masonry. One of the first ideas presented was to substitute the complex geometry of the basic
cell with a simplified geometry, so that a closed-form solution for the homogenization problem was
possible. This approach, rooted in geotechnical engineering applications, assumed masonry as
a layered material (Lourenço 1996). This simplification does not allow including information on

Page 7 of 13
 DOI 10.1007/978-3-642-36197-5_153-1
Encyclopedia of Earthquake Engineering

# Springer-Verlag Berlin Heidelberg 2014

Fig. 10 Proposed micro-mechanical model: (a) elementary cell; (b) subdivision in layers along thickness and subdivision of each layer in sub-domains; (c) imposition
of internal equilibrium, equilibrium on interfaces and anti-periodicity

Page 8 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 11 Masonry infill wall subjected to blast pressure: comparison among deformed shapes at t ¼ 400 ms for
(a) homogenized limit analysis and heterogeneous 3D elastic–plastic FE approach; (b) comparison between maximum
out of plane displacements and limit analysis failure mode

the arrangement of the masonry units and provides significant errors in the case of nonlinear
analysis. Moreover, the results depend on the sequence of homogenization steps. To overcome the
limitation, micro-mechanical homogenization approaches that consider additional internal defor-
mation mechanisms have been derived (Zucchini and Lourenço 2002, 2009). The implementation of
these approaches in standard macroscopic finite element nonlinear codes is relatively simple, and the
approaches can compete with macroscopic approaches (see Lourenço et al. (2007) for a detailed
review on homogenization approaches).
A powerful micro-mechanical model for the limit analysis for masonry has been recently
developed (Milani et al. 2006a, b). In this model, the elementary cell is subdivided along its
thickness in several layers. For each layer, fully equilibrated stress fields are assumed, adopting
polynomial expressions for the stress tensor components in a finite number of sub-domains. The
continuity of the stress vector on the interfaces between adjacent sub-domains and suitable anti-
periodicity conditions on the boundary surface are further imposed. In this way, linearized homog-
enized surfaces in six dimensions for masonry in- and out-of-plane loaded are obtained. Such
surfaces are then implemented in a finite element limit analysis code for simulation of 3D structures
and including, as recent advances, blast analysis, quasiperiodic masonry internal structure, and FRP
strengthening. Figure 10 shows a masonry wall constituted by a periodic arrangement of bricks and
mortar arranged in running bond. For a general rigid-plastic heterogeneous material, homogeniza-
tion techniques combined with limit analysis can be applied for the evaluation of the homogenized
in- and out-of-plane strength domain.

Page 9 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Fig. 12 Church of Gondar (Portugal), FE discretization adopted: (a) heterogeneous random mesh vs. mesh for running
bond regular heterogeneous and homogenized random analysis; (b) ECDF of the failure load provide through a direct
heterogeneous approach and homogenized limit analysis simulations

The homogenized failure surface obtained has been coupled with finite element limit analysis.
Both upper- and lower-bound approaches have been developed, with the aim to provide a complete
set of numerical data for the design and/or the structural assessment of complex structures. The finite
element lower-bound analysis is based on an equilibrated triangular element, while the upper bound
is based on a triangular element with discontinuities of the velocity field in the interfaces. Recent
developments include the extension of the model to blast analysis (Milani et al. 2009), to quasipe-
riodic masonry (Milani et al. 2010), and to FRP strengthening (Milani and Lourenço 2013).
An enclosure running bond masonry wall subjected to a distributed blast pressure is considered
first (Milani et al. 2010). The wall is supposed simply supported at the base and on vertical edges and
free on top due to the typical imperfect connection between infill wall and RC beam. A full 3D FE
heterogeneous elastic–plastic dynamic analysis has been also conducted, in order to have a deep
insight into the problem and to collect alternative data to compare with. For the 3D model, a rigid
infinitely resistant behavior for bricks was assumed, whereas for joints, a Mohr–Coulomb failure

Page 10 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

criterion with the same tensile strength and friction angle used in the homogenized approach for
joints was adopted. Eight-noded brick elements were utilized both for joints and bricks, with
a double row of elements along wall thickness. A comparison between the deformed shapes at
t ¼ 400 ms obtained with the present model and the commercial software is schematically depicted
in Fig. 11a. As it is possible to notice, the models give almost the same response in terms of
deformed shape for the particular instant time inspected, confirming that reliable results may be
obtained with the model proposed. On the other hand, it is worth underlining that the homogenized
rigid-plastic model required only 101 s to be performed on a standard PC Intel Celeron 1.40 GHz
equipped with 1Gb RAM, a processing time around 10–3 lower than the 3D case. Comparisons of
time-maximum displacement diagrams provided by the two models analyzed are reported in
Fig. 11b, together with the evolution of the deformation provided by the homogenized model
proposed.
Recently, two different classes of problems have been investigated (Milani et al. 2010), the first
consisting of full stochastic representative element of volume (REV) assemblages without horizon-
tal and vertical alignment of joints and the second assuming the presence of a horizontal alignment
along bed joints, i.e., allowing block height variability only row by row. The model is characterized
by a few material parameters, and it is therefore particularly suited to perform large-scale Monte
Carlo simulations. Masonry strength domains are obtained equating the power dissipated in the
heterogeneous model with the power dissipated by a fictitious homogeneous macroscopic plate.
A stochastic estimation of out-of-plane masonry strength domains (both bending moments and
torsion are considered) accounting for the geometrical statistical variability of blocks dimensions is
obtained with the proposed model. The case of deterministic block height (quasiperiodic texture) can
be obtained as a subclass of this latter case. As an important benchmark, the case in which joints
obey a Mohr–Coulomb failure criterion is also tested and compared with results obtained assuming
a more complex interfacial behavior for mortar. In order to show the capabilities of the approach
proposed when dealing with large-scale structures, the ultimate behavior prediction of
a Romanesque masonry church facade located in Portugal are presented. Comparisons with finite
element heterogeneous approaches and “at hand” calculations show that reliable predictions of the
load bearing capacity of real large-scale structures may be obtained with a very limited computa-
tional effort (see Fig. 12).

Conclusions
Constraints to be considered in the use of advanced modeling are the cost, the need of an experienced
user/ engineer, the level of accuracy required, the availability of input data, the need for validation,
and the use of the results. As a rule, advanced modeling is a necessary means for understanding the
behavior and damage of masonry constructions. For the study of isolated structural elements and
problems controlled by the geometry of a few units, micro-modeling strategies are available. For
large-scale applications, average continuum mechanics is usually adopted, and homogenization
techniques represent a popular and active field in masonry research. The assessment and design of
unreinforced masonry structures subjected to seismic loading is particularly challenging when
masonry box behavior is not present. In this case, macro-block analysis seems adequate for
simplified assessment and design of strengthening measures.

Page 11 of 13
Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Summary
Masonry joints act as planes of weakness, and the explicit representation of the joints and units in
a numerical model seems a logical step towards a rigorous analysis tool (Fig. 1). This kind of
analysis is particularly adequate for small structures, subjected to states of stress and strain strongly
heterogeneous. In general, a higher computational effort ensues, so this approach still has a wider
application in research and in small models for localized analysis. The approach based on the use of
averaged constitutive equations seems to be the most suitable to be employed a large-scale finite
element analyses, either by collating experimental data at average level (macro-modeling) or from
homogenization techniques.
A micro-model for masonry joints is usually composed by three individual yield functions,
consisting of the usual Coulomb friction criterion for shear, combined with tension and compression
cutoffs. A macro-model for the masonry composite must permit to reproduce the results obtained in
uniaxial tests, with distinct behavior obtained along different directions. Finally, homogenization
techniques require as input the geometry of the units and joints, as well as their mechanical
properties. The most successful applications are based on micromechanical models that perform
a simplified single-step homogenization because a close form solution is not available.
This entry provides a review of different modeling approaches for masonry structures with
illustrative applications.

References
Berto L, Saetta A, Scotta R, Vitaliani R (2002) Orthotropic damage model for masonry structures.
Int J Numer Methods Eng 55(2):127–157
Dhanasekar M, Page AW, Kleeman PW (1985) The failure of brick masonry under biaxial stresses.
Proc Inst Civ Eng, Part 2 79:295–313
Lemos JV (2007) Discrete element modeling of masonry structures. Int J Archit Herit 1(2):190–213
Lourenço PB (1996) A matrix formulation for the elastoplastic homogenisation of layered materials.
Mech Cohes-Frict Mater 1:273–294
Lourenço PB (2000) Anisotropic softening model for masonry plates and shells. J Struct Eng
126(9):1008–1016
Lourenço PB (2002) Computations of historical masonry constructions. Prog Struct Eng Mater
4(3):301–319
Lourenço PB, Rots JG (1997) Multisurface interface model for the analysis of masonry structures.
J Eng Mech ASCE 123(7):660–668
Lourenço PB, de Borst R, Rots JG (1997) A plane stress softening plasticity model for orthotropic
materials. Int J Numer Methods Eng 40:4033–4057
Lourenço PB, Rots JG, Blaauwendraad J (1998) Continuum model for masonry: parameter estima-
tion and validation. J Struct Eng 124(6):642–652
Lourenço PB, Milani G, Tralli A, Zucchini A (2007) Analysis of masonry structures: review of and
recent trends of homogenization techniques. Can J Civ Eng 34(11):1443–1457
Milani G, Lourenço PB (2013) Simple homogenized model for the non-linear analysis of FRP
strengthened masonry structures. Part I: theory. J Comput Const ASCE 139(1):59–76
Milani G, Lourenço PB, Tralli A (2006a) Homogenised limit analysis of masonry walls. Part I:
failure surfaces. Comput Struct 84(3–4):166–180

Page 12 of 13
 Encyclopedia of Earthquake Engineering
DOI 10.1007/978-3-642-36197-5_153-1
# Springer-Verlag Berlin Heidelberg 2014

Milani G, Lourenço PB, Tralli A (2006b) A homogenization approach for the limit analysis of
out-of-plane loaded masonry walls. J Struct Eng ASCE 132(10):1650–1663
Milani G, Lourenço PB, Tralli A (2009) Homogenized rigid-plastic model for masonry walls
subjected to impact. Int J Solids Struct 46(22–23):4133–4149
Milani G, Lourenço PB, Tralli A (2010) A simplified homogenized limit analysis model for
randomly assembled blocks out-of-plane loaded. Comput Struct 88(11–12):690–717
Oliveira DV, Lourenço PB (2004) Implementation and validation of a constitutive model for the
cyclic behaviour of interface elements. Comput Struct 82(17–19):1451–1461
Orduña A, Lourenço PB (2005a) Three-dimensional limit analysis of rigid blocks assemblages. Part
I: torsion failure on frictional joints and limit analysis formulation. Int J Solids Struct
42(18–19):5140–5160
Orduña A, Lourenço PB (2005b) Three-dimensional limit analysis of rigid blocks assemblages. Part
II: load-path following solution procedure and validation. Int J Solids Struct
42(18–19):5161–5180
Pelà L, Cervera M, Roca P (2011) Continuum damage model for orthotropic materials: application
to masonry. Comput Methods Appl Mech Eng 200(9–12):917–930
Rots JG (1991) Numerical simulation of cracking in structural masonry. Heron 36(2):49–63
Seim W (1994) Numerical modeling of the failure of biaxially loaded masonry walls with consid-
eration of anisotropy (in German). PhD thesis, University of Karlsruhe
Zucchini A, Lourenço PB (2002) A micromechanical model for the homogenisation of masonry. Int
J Solids Struct 39(12):3233–3255
Zucchini A, Lourenço PB (2009) Validation of a micro-mechanical homogenisation model: appli-
cation to shear walls. Int J Solids Struct 46(3–4):871–886

Page 13 of 13

View publication stats