Professional Documents
Culture Documents
An Orthotropic Damage Model For Masonry Walls With Consistent Damage Evolution Laws
An Orthotropic Damage Model For Masonry Walls With Consistent Damage Evolution Laws
https://doi.org/10.1617/s11527-023-02220-x (0123456789().,-volV)
( 01234567
89().,-volV)
ORIGINAL ARTICLE
Abstract This research presents a constitutive Keywords Macro scale modeling Continuum
model for the macro scale simulation of masonry damage mechanics Orthotropic damage model
structures. The model is containing an orthotropic Tensor mapping Shear behaviour
plane stress assumption, which appears as an appro-
priate assumption for the in-plane analysis of masonry
walls. The material model is based upon damage
mechanics, split into tensile and compression parts. 1 Introduction
The novelty herewith is the consistent mapping of the
damage evolution laws. Aim of this research is to Ever since the use of masonry in buildings and
develop a simple but accurate constitutive law, monuments, it has been a challenging task to accu-
suitable to simulate large scale structures and build- rately predict the behaviour and durability of the
ings with reasonable computational costs. The devel- construction material. With the introduction of finite
opments are presented and compared with available element methods, numerical methods have eased the
literature examples from laboratory testings. assessment and simulation significantly [1]. Simulta-
neously, different scales of modeling have been raised,
aiming to analyze different details of the structure.
Detailed micro models, which are containing all
T. Teschemacher constituents separately have been developed to show
Chair of Structural Analysis, TUM School of Engineering the heterogeneous composite behaviour. Macro mod-
and Design, Technical University of Munich, Munich, els on the other end shall help estimating the structural
Germany behaviour of the material.
T. Teschemacher (&) K.-U. Bletzinger Macro models overlook the combined kinematics
School of Civil Engineering, The University of of bricks, mortar and their interface to aim a compu-
Queensland, Brisbane, Australia tationally efficient modelling of the large scale
e-mail: tobias.teschemacher@tum.de
structures. The constitutive behaviour might be
P. Kalkbrenner L. Pelà obtained from reference volume elements (RVE),
Department of Civil and Environmental Engineering, where the properties can be gathered experimentally
Universitat Politècnica de Catalunya, Barcelona, Spain e.g. [2, 3], from national codes as the EC 6 [4], or
numerically [5–10]. The macro-scale finite element
R. Wüchner
Institute of Structural Analysis, Technical University of simulations of masonry structures may be idealized as
Braunschweig, Brunswick, Germany solid, shell, beam, truss or spring elements depending
151 Page 2 of 18 Materials and Structures (2023) 56:151
on the local dimensional fidelity requirements. Even damage models for simplicity with less parameters
though solid and shell elements can better represent and fewer internal dependencies.
the system geometry of masonry structures, dimen- The mentioned material models are adequate in the
sionally reduced 1D and spring elements are exten- analysis of isotropic materials with sensible triggers
sively used for seismic and blast simulations because for the softening zones. Thusly, in masonry, its
of their optimized computational costs ([11–15] for applications are so-far mostly for the simulation
some seismic masonry examples). Continuum-based within continuum-based micro models of bricks
masonry macro models are typically accurate in the (without holes) and mortar. However, as the structural
structural simulation (examples for in-plane loading properties of masonry show significant correlating
include references [8, 16] and out-of-plane load- orthotropic behaviour it is insufficient to fulfill the
ing [17–20]), however, estimating non-homogeneous apparent direction-dependent estates. Rather than
material properties as unstructured bonds and detailed updating the yield surface with anisotropic surfaces
crack predictions is, up-to-date, complex. Apart from as the Hill-type criterion [29–31] or Tsai–Wu [32] it is
analytical and continuum based approaches, mesh free preferred to base upon a mapping from the real
or particle based methods find increasing application anisotropic space to an ideal mapped isotropic stress
in masonry classification as well (e.g. discrete element state. This concept is generally simpler with the
modeling [21], or latice discrete particle model [22]), possibility to keep former developments and mostly
which facilitates the latter problem of difficulties in leaner computational costs. The concept of the
crack propagation simulation, however, the computa- implicit orthotropic yield criterion was initiated by
tional costs can be tremendous. Betten [33] and later advanced by Oller et al. [34, 35].
This research aims to facilitate simulations capable Pelà et al. [16, 36] promoted those models to the
to express complex geometries and structures with simulation of masonry structures. The most significant
justifiable computational costs for full scale buildings. improvement from the macro model of Lourenço et al.
Auxiliary elements and beams typically have increas- [31] has been achieved by the split in compressive and
ingly challenges when modelling complex structures. tensile damages, which allowed to model an exceed-
Whereas, solid and particle based modeling allows ingly vast amount of problems.
almost infinite possibilities in the geometrical speci- Within this research the focus is on damage-based
fication of constructions, typically the computational models, while plastic zones would be expressed by the
efforts prevent the application in structural analyses damage parameter. This is a fast and generally reliable
processes. Thusly, the larger scope shall be on the approach once a one directional loading is applied. For
description of surface-based shell structures, which seismic and changing loading cases, plasticity effects
allow the expression of many geometrical shapes, should be considered separately for a better accuracy
including reasonable computation costs. (see e.g. [23, 27, 37]).
Continuum based damage models define a common Subsequently, a model which merges the latest
way to describe the structural properties of masonry improvements of the damage evolution law from
buildings. Initially, one damage parameter was Petracca et al. [14, 38] and Pelà et al. [16] into a novel
applied to describe the softening of the construction material model is presented. The presented research
material [23]. Where later, those models have been proposes a consistent mapping of the corresponding
split up into compression and tension damages to damage evolution laws with more sensitive triggers in
allow deeper insights in highly loading dependent the softening within an appropriate orthotropic
materials [24–26]. Those concepts have been behaviour.
enhanced further with damage parameters per direc- This publication is structured as following:
tion [27]. Other concepts do not make the damage split
• Section 2 presents the applied continuum based
within the stress state but the direction [28] with inter-
d þ /d -damage material model.
dependent damage parameters, which is a powerful
• Section 3 shows the damage model with the
approach but less appropriate for the apparent
orthotropic to a mapped isotropic state.
masonry properties. Following, within this publication
• Section 4 gives some explanatory examples with
the focus shall be kept on two parameter dþ /d -
the proposed constitutive model.
Materials and Structures (2023) 56:151 Page 3 of 18 151
The stress thresholds r þ and r are the maximum Furthermore, I1 is the first invariant of the effective
uniaxial equivalent stresses which can be carried stress tensor and J2 the second invariant of the
within a system. The applied threshold yield surface deviatoric stress tensor. rmax is used as the maximal
shall be based upon a Drucker–Prager criterion, principal effective stress, respectively rmin being the
enhanced by Lubliner et al. [23] and thoroughly smallest principal effective stress. The shear reduction
defined by Petracca et al. [38]. The two equivalent parameter k1 is introduced to control the compressive
stress surfaces in tension s and compression sþ , are strength in the shear state, which properties are shown
graphically displayed in Fig. 1 and discussed in the Fig. 1. Petracca [14] and Petracca et al. [17] provide
following: comprehensive studies on the structural effect of this
parameter. Finally, the factor kb is defined by the ratio
from the biaxial strength of the applied material (see
Fig. 1).
151 Page 4 of 18 Materials and Structures (2023) 56:151
As the thresholds are defined for any stress state, in The characteristic length lch is used to regularize the
some cases it may happen that the damage surface fracture zone and is dependent on the geometry of the
advances for tension in uniaxial compression or in finite element. Multiple approaches exist to approxi-
reverse. To avoid this, the Heavyside function HðxÞ ¼ mate the characteristic length [10, 39–41], where it is
possible to accurately calibrate the fracture dissipation
0 for x 0 length within structured quadrilaterals, more complex
is used to ensure following criteria:
1 for x [ 0 non-aligned geometries and high order form functions
• tensile damage only evolves if at least one will typically produce discrepancies.
principal stress is positive, As proposed by Petracca et al. [17, 38], the
• compressive damage only evolves if at least one compression damage progression is analyzed on an
principal stress is negative. ad hoc basis under a segmented Bézier curve, which is
controlled by the elastic limit strength fc0 , the yield
The effect of the applied Heavyside is deciphered in strength fcp , its corresponding yield strain cp and the
Fig. 1 by the active and inactive threshold surfaces. residual strength fcr . The damage evolution is defined
by the hardening variable RðnÞ with its strain-like
2.3 Damage evolution counterpart n ¼ rE (follow ‘‘Appendix 1’’ for a
detailed extraction of the Bézier curve):
The previously described equivalent stress thresholds RðnÞ
shall be used as parameters for the irreversible damage d ðr Þ ¼ 1 : ð2:9Þ
r
evolution of the material. It is expected that the tensile
softening is defined by the strength and the fracture
energy with a non-linear softening, whereas the 2.4 Total stresses
compression shows a more complex behaviour with
multiple inputs for the analysis (see Fig. 2 for possible Including the convoluted damage parameters d þ /d
damaging zones). the total stresses are defined by:
The tensile damage growth is obtained by the r ¼ ð1 dþ Þrþ þ ð1 d Þr : ð2:10Þ
subsequent equation:
þ
rþ r rþ
d þ ðr þ Þ ¼ 1 0þ exp 2Hdis 0 þ : ð2:7Þ 2.5 Tangent constitutive tensor
r r0
Hdis is defined as following: the fourth order tangent constitutive tensor Ctangent can
lch 2EGt be computed by:
Hdis ¼ with lmat ¼ : ð2:8Þ
lmat lch ft02 Ctangent ¼ ð1 d þ ÞP þ ð1 d ÞðI PÞ: ð2:12Þ
This tensor can directly be used within most finite
element implementations. Checkout [42] for alterna-
tive approaches for a more consistent implementation
of the constitutive tensor.
In the following the recently described isotropic considering the split into tensile rþ and compressive
damage model shall be enhanced to enable orthotropic r the mapping relations read:
behaviours. Rather than updating the yield surface þ
rþ ¼ Ar : rþ and r ¼ Ar : r :
with anistropic surfaces as the Hill type criterion
[29–31] or Tsai–Wu [32] it is preferred to base upon a ð3:2Þ
mapping from the real anisotropic space to an ideal
The transformation tensors Arþ and Ar map from the
mapped isotropic stress state [33–35], as this is the
actual orthotropic space and its actual limit strength to
simpler and thusly computationally more efficient
the mapped isotropic space. As the mapped isotropic
approach. Furthermore, it allows the application of
space is an independently chosen space, f þ and f
advanced threshold surfaces for isotropic materials, as
can be selected arbitrary. However, Pelà L et al. [36]
e.g. the one discussed in Sect. 2. 1
suggests to use f ¼ f11 to simplify the transforma-
tion, which are defined by following matrices accord-
3.1 Transformation between actual orthotropic
ing to Voigt notation (see [34] for a detailed derivation
and mapped isotropic space
of the constituents):
0 þ 1
The concept of the implicit orthotropic yield criterion was f
initiated by Betten [33] and lateron advanced by Oller et al. Bf 0 0 C
B t0;1 C
[34, 35]. Pelà et al. [16, 36] brought those models to the B þ C
þ B f C
application in simulation of masonry structures, with Ar ¼ B B 0 0 C;
C ð3:3Þ
B ft0;2 C
enhancing the previous models with the split in compres- B þ C
sive and tensile parts. The models are based upon a @ ft;12 A
0 0
transformation between the actual anisotropic, or in this ft;12
case orthotropic stress state to an arbitrary mapped 0 1
isotropic space (notation ðÞ ) and its corresponding f
B fcp;1 0 0 C
strain-based reverse mapping, as shown in Fig. 3. B C
B f C
The mapping from the actual stress state to the B 0 C
Ar ¼B 0 C: ð3:4Þ
isotropic space is defined by the transformation tensor B fcp;2 C
B C
Ar is defined as following: @ A
f12
0 0
fc;12
Additionally to f , this model also requires f12
,
which shall impress the mapped space shear force.
Accordingly, the last entries of the mapping matrices
shall devote the relation between the two shear
strength.2 Considering Eq. 2.2, the stress split is still
valid in the mapped isotropic space:
r ¼ rþ r : ð3:5Þ
1
Within this section f11 is the generic form for ft0;1 or fcp;1 .
2
Pelà et al. [36] considers a Drucker–Prager damage surface,
which itself modifies the threshold r0 with the mapped strength:
pffiffi pffiffiffi
r0 ¼ 33 ðkb 2Þf . The Lubliner damage surface is reduc-
Fig. 3 Mapping between actual anisotropic and isotropic space ing the uniaxial stress with the contribution of kb (see Sect. 2).
(adapted from [34]) This makes the mapping simpler and more consistent.
151 Page 6 of 18 Materials and Structures (2023) 56:151
Eqs. 2.4 and 2.5 and with those the damage parameters
f22 =f11
2
i
cp ¼ cp;1 cos2 ðhÞ þ i cp;2 sin2 ðhÞ; ð3:17Þ
i;2 i;2
f f
E 2 E
¼
fcp;1 cp;1
cos ðhÞ þ
fcp;2 cp;2
sin2 ðhÞ: ð3:18Þ
E1 E2
ci ¼ ci cos2 ðhÞ þ ci sin2 ðhÞ for i ¼ 1; 2; 3; Fig. 4 Parameter definition within this scope
ð3:20Þ
Within Table 1 are summarized the material proper-
equally as the shear reduction parameter k1 and the ties including the additional Bézier curve parameters
factor kb defined by the ratio of biaxial strength9 : (see ‘‘Appendix 1’’).
The obtained results for compression and tension
k1 ¼ k11 cos2 ðhÞ þ k12 sin2 ðhÞ; ð3:21Þ
are presented within Fig. 5. It demonstrates that the
damage evolution in both directions can be expressed
kb ¼ kb1 cos2 ðhÞ þ kb2 sin2 ðhÞ: ð3:22Þ
appropriately. Therefore, all control parameters are
mapped consistently between the apparent spaces. It
shall be noted that the graphs act independently as the
4 Numerical example additional mapping is employed. If this were not
provided, the relaxation would always conform
This section demonstrates the applicability of the dependently in the two directions. The additionally
proposed model using two numerical examples. First,
this model is used to represent the directional masonry
Table 1 Masonry material properties from experimental
material features from [3]. Next, the shear wall tests
tests [3]
from [43] are examined with varying pre-compression
and different geometries (without and with hole). E (MPa) m (–) G (MPa) Gt (N/mm) ft0 (MPa)
Within this publication the parameter convention in Elastic and tension properties
the directions parallel and normal to bed joints is given 5000 0.15 1340 1:9 103 0.091
in Fig. 4. E1 corresponds to the direction e1 and E2 to 3100 0.1 960 3 0.272
4:5 10
the direction e2 . Same accounts for the remaining
Gc (N/mm) fc0 (MPa) fcp (MPa) r (–) fcr (MPa)
parameters. e3 refers to the out of plane direction,
which is not discussed within this scope. See also [3], Compression properties
which refers to the same parameter definition. 1.2 5.2 7.38 0.0018 1.3
1.2 2.9 4.05 0.002 0.0
4.1 Uniaxial numerical tests of [3]-masonry
c1 c2 c3
(–) (–) (–)
In this study, the uniaxial behaviours within both
directions shall be examined with the apparent mate- Be´zier curve parameters
rial model, including a perspective to the damage 0.65 0.8 1.2
growth. This is conducted to prove that the orthotropic 0.65 0.5 1.5
mapping is working properly, with a focus on the
variated strengths within compression and tension.
151 Page 8 of 18 Materials and Structures (2023) 56:151
4.2 Shear wall Due to the high amount of references, the example
delivers great possibilities for comparisons and thus, it
This study considers a masonry wall which is first appears as a good benchmark. Most of the in-plane
subjected to a vertical load (I) at different magnitudes. masonry phenomena are apparent: significant ortho-
In a second step it is fixed vertically at this deforma- tropic material, shear failure and biaxial stiffness
tion (IIa) and sheared horizontally (IIb). Geometri- effects. This helps to justify if the model is suitable to
cally this problem appears within two different express the major in-plane masonry characteristics.
variations: once as a whole wall (see Fig. 6a) and It shall be noted that the apparent literature,
once with a window opening (see Fi. 6b). Both walls typically focuses on either of the geometrical options:
have a dimension of 1 0:99 m. full wall or wall with window. This shows that
This example [43, 44] is well known in literature, covering this example entirely remains challenging.
among others shall be mentioned: The material properties of the tests are summarized
• Lourenço [45]: micro and macro models in Table 2. Some of the macro scale parameters are
Materials and Structures (2023) 56:151 Page 9 of 18 151
known directly from test data t ðÞ by Vermeltfoort and loading with 0.3 MPa appears in 2 consecutive
Raijmakers [43] or analyzed by micro scale analyses experimental trials.
[45]. While others are acquired through speculation The stress–strain graphs of this example are
s
ðÞ and reverse engineering. These parameters are presented within Fig. 7. The strengthens of the
either obtained from literary works or extrapolated 0.3 MPa pre-loading examples are matching well.
with the apparent demands. It should be noted that The experimental test at 1.21 MPa indicates material
certain speculated values change across several pub- characteristics that are noticeably different since the
lications since the employed models make modest first elastic zone seems stiffer than in the previous test.
adjustments and apply parameters in various manners. That is quite usual with this heterogeneous material. In
Unfortunately, not sufficient experimental data is both test setups, the softening of the specimen seems
available to mechanically justify either of the selected
parameters. Additionally, various parameter sets could
be providing the correct outcome for the apparent
model. As a further challenge shall be noted that the
material parameters seem to vary significantly
between the experimental tests. However, within this
numerical study only one parameter set shall be
considered. Although there may be larger disparities
compared to some of the results, this should be a step
in the direction of homogenization, allowing predic-
tions rather than just result fitting by calibration.
Compression properties
s t s s s
6.8 10 0.002 2.3 2.0
s t s s s
5.2 8.8 0.002 3.0 1.96
c1 (–) c2 (–) c3 (–) kb (–) k1 (–) ft;12 fc;12
ft;21 (–) fc;21 (–)
Figure 8 displays the findings and compares them to are presented the corresponding Cauchy stresses, the
several existing references. The graph shows good principal stresses in their direction and the max and
agreement between both test results. The first soften- min principal stresses in the system. The colours of the
ing is adequately addressed by the suggested model, principal stresses are mostly kept similar, however,
however the first major softening is exaggerated. peaks are highlighted slightly, which can be compared
The comparison to some literature in Fig. 8 show to the neighboring figures.
that most results contain an almost continuous soften- The evolving damage results from Fig. 9 denote
ing after reaching the peak. The presented solution is crack initiations at the corners of the window, that are
showing a jump in the softening. It is believed that this opening towards the supports of the walls amid
is in relation to the shear cracking which is not covered increasing displacements. At the top left and basal
in such detail in the other references. This damaging right part of the specimen is shown a damage, which is
would allow a movement, which otherwise would not not displayed within the referenced crack pattern. This
be possible. However, this is based on speculation and is showing a separation of the specimen from the
it is suggested to conduct further research on this topic. concrete top and bottom bars, which can also be
Furthermore, it shall be mentioned that with certain observed within the experimental results. Further-
calibrations this jump can be smeared. more, in the damage patterns can be seen that in both,
It should be emphasised that while each model from the experimental tests and the numerical model, a
the literature that has been cited has some merit. crack happens in the left upper third and right bottom
However, typically distinct zones are highlighted. third. This proves quite effectively that a certain trust
Some results match mostly the initial limit state. can be given to the macroscopic models and that not
Others focus more on the accurate representation of only micro models would display this local phenom-
the softening. ena. Finally, the cracks which initiate at the corners of
the windows do not go towards the corners of the
4.2.4 Qualitative analysis of results specimen, but towards a close point at top or bottom.
The reason of this is that most of the forces, which are
Within Fig. 9 are shown the displacements in refer- supplied from top to bottom go through the parts left
ence to the aimed current top displacements. Further- and right of the window. Accordingly those areas are
more are provided the tensile damage and compressive largely under compression and thusly do not fail by
damage contour plots and the resulting maximum tensile forces.
principal strains of the digital model. Within Fig. 11
Materials and Structures (2023) 56:151 Page 11 of 18 151
Fig. 9 Observed displacements, damage, and strains at [43]-wall with window tests. Computed with 50 50 elements
The tensile damage indicates good agreement with instead of thin cracks (corners of the window towards
the reference solutions, see Fig. 10, which displays the the boundaries of the specimen).
experimental test results within the solutions from the Lourénço et al. [31] and Pelà et al. [16] are
numerical model at a displacement of 20 mm. It shows reporting a compressive damage in the upper left and
thin crack lines where only single cracks appear (top lower right corner of the specimen. This is not
third left and bottom third right). At the zones of observed in the current model. The apparent com-
multiple crack lines, the model shows damage zones pressive stresses in those corners are higher than in the
151 Page 12 of 18 Materials and Structures (2023) 56:151
Fig. 11 Observed stresses at [43]-wall with window tests. Computed with 50 50 elements
processes compared to micro models. The computa- One of the improvements of the proposed model is
tional costs seem to be fairly low, whereby a the capability to cover the biaxial stress states and the
comprehensive quantitative comparison would exceed resulting internal forces in an orthotropic manner,
the scope of this publication, as it would require a which allows the analysis of masonry in different pre-
comparison to micro models with comparative loading scenarios. Furthermore, the model has a large
implementations. bandwidth of possibilities to trigger the stress strain
151 Page 14 of 18 Materials and Structures (2023) 56:151
Fig. 12 Observed damage at [43]-wall with window tests with a mesh refinement from 10 10–100 100 elements
behaviour, which brings stability within the simulation tests. It has to be noted that a large amount of material
if the material response is known. As contrast, properties needed to be speculated as not sufficient test
comparatively many parameters need to be obtained, data was provided. Seeing the performance of the law
which is up-to date infrequent, as typically only few within a fully comprehensive test setup [3] would be
parameters are studied within a series of experimental required to specify the entire generality of the model.
Materials and Structures (2023) 56:151 Page 15 of 18 151
It shall be noted that the model contains 17 Open Access This article is licensed under a Creative Com-
parameters per direction, making it 34 in total. mons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any med-
Unfortunately, mostly barely that much information ium or format, as long as you give appropriate credit to the
is known about the material. Once the uniaxial original author(s) and the source, provide a link to the Creative
evolution law would be simplified, this number could Commons licence, and indicate if changes were made. The
be reduced. However, this would limit the possibilities images or other third party material in this article are included in
the article’s Creative Commons licence, unless indicated
for a correct adjustment of the model. otherwise in a credit line to the material. If material is not
The model was studied within examples which included in the article’s Creative Commons licence and your
cover uniaxial compression and tension, shear, biaxial intended use is not permitted by statutory regulation or exceeds
stress states with a simple and a more complex the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this licence, visit
geometry. The test cases have been in-plane 2D http://creativecommons.org/licenses/by/4.0/.
problems in a plane-stress state, while the expected
behaviour could be simulated properly. However,
more complex shapes would need to be considered to Appendix 1: Uniaxial compression Bézier-curve
allow a more generic evaluation of the proposed model based relaxation
with the behaviour of the sophisticated construction
material masonry. Accordingly, an avenue for future Within this appendix, the uniaxial compression
research shall be the application of the constitutive law Bézier-curve based relaxation introduced by Petracca
within 3D solid formulations for complex shapes with et al. [17], which is shown within Fig. 13 shall be
largely varying thickness. Another possibility would deciphered. The model itself requires as input E, fc0 ,
be the imposition within thin-walled or thick/solid- fcp , cp and fcr . Furthermore, the locations of the
based shell formulations to investigate the bending remaining required stress/strains can be controlled by
and the out-of plane performance. Thereby, shells with a set of scalars, namely the Bézier controllers c1 , c2
a thickness integration would consider multiple enti- and c3 . However, if preferred, the input can also
ties of the material per in-plane integration point (see contain the remaining stresses and strains, rather than
[51, 52] for eventual applications). controllers.
With the input of the Bézier controller inputs, the
Acknowledgements The authors thank the reviewers whose
constructive comments improved the quality of the article. The complete set of initial control parameters shall be
authors gratefully acknowledge support from the Deutsche defined as following:
Forschungsgemeinschaft (DFG) as part of the project ‘‘Eine
Methode zur effizienten Simulation großer rj ¼ fc0 ri ¼ rp ¼ rj ¼ fcp ru ¼ rr ¼ fcr ð6:1Þ
Mauerwerksscheiben unter exzentrischer und/oder zyklischer
biaxialer Beanspruchung auf Grundlagewirklichkeitsnaher
Kleinkörperversuche’’ (BL306/34-1) and the Australian
Government Research Training Program Scholarship and The
University of Queensland as part of the ‘‘Global Strategy and
Partnerships Seed Funding Scheme, 2019’’.
Declarations
3. Teschemacher T, Wilson P, Wüchner R, Bletzinger K-U 22. Angiolilli M, Pathirage M, Gregori A, Cusatis G (2021)
(2022) Macroscopic characterisation of modern masonry. Lattice discrete particle model for the simulation of irreg-
Eng Struct. https://doi.org/10.1016/j.engstruct.2022. ular stone masonry. J Struct Eng 147(9):04021123
114630 23. Lubliner J, Oliver J, Oller S, Oñate E (1989) A plastic-
4. Eurocode 6 (2013) Design of masonry structures—Part 1-1: damage model for concrete. Int J Solids Struct
general rules for reinforced and unreinforced masonry 25(3):299–326
structures 24. Cervera M, Oliver J, Faria R (1995) Seismic evaluation of
5. Lourenço PB, Rots JG (1997) Multisurface interface model concrete dams via continuum damage models. Earthq Eng
for analysis of masonry structures. J Eng Mech Struct Dyn 24(9):1225–1245
123(7):660–668 25. Faria R, Oliver J, Cervera M (1998) A strain-based plastic
6. Zucchini A, Lourenço P (2002) A micro-mechanical model viscous-damage model for massive concrete structures. Int J
for the homogenisation of masonry. Int J Solids Struct Solids Struct 35(14):1533–1558
39(12):3233–3255 26. Wu JY, Li J, Faria R (2006) An energy release rate-based
7. Milani G, Lourenço P, Tralli A (2006) Homogenised limit plastic-damage model for concrete. Int J Solids Struct
analysis of masonry walls, part I: failure surfaces. Comput 43(3):583–612
Struct 84(3):166–180 27. Berto L, Saetta A, Scotta R, Vitaliani R (2002) An ortho-
8. Milani G, Lourenço P, Tralli A (2006) Homogenised limit tropic damage model for masonry structures. Int J Numer
analysis of masonry walls, part II: structural examples. Methods Eng 55(2):127–157
Comput Struct 84(3):181–195 28. Lapczyk I, Hurtado JA (2007) Progressive damage model-
9. Zucchini A, Lourenço P (2009) A micro-mechanical ing in fiber-reinforced materials. Compos Part A Appl Sci
homogenisation model for masonry: application to shear Manuf 38(11):2333–2341 (CompTest 2006)
walls. Int J Solids Struct 46(3):871–886 29. Hill R, Orowan E (1948) A theory of the yielding and plastic
10. Petracca M, Pelà L, Rossi R, Oller S, Camata G, Spacone E flow of anisotropic metals. Proc R Soc Lond Ser A Math
(2016) Regularization of first order computational homog- Phys Sci 193(1033):281–297
enization for multiscale analysis of masonry structures. 30. Hill R (1990) Constitutive modelling of orthotropic plas-
Comput Mech 57:257–276, 02 ticity in sheet metals. J Mech Phys Solids 38(3):405–417
11. Pelà L, Aprile A, Benedetti A (2009) Seismic assessment of 31. Lourénço PB, De Borst R, Rots JG (1997) A plane stress
masonry arch bridges. Eng Struct 31(8):1777–1788 softening plasticity model for orthotropic materials. Int J
12. DeJong MJ, Belletti B, Hendriks MA, Rots JG (2009) Shell Numer Methods Eng 40(21):4033–4057
elements for sequentially linear analysis: lateral failure of 32. Tsai SW, Wu EM (1971) A general theory of strength for
masonry structures. Eng Struct 31(7):1382–1392 anisotropic materials. J Compos Mater 5(1):58–80
13. Pelà L, Aprile A, Benedetti A (2013) Comparison of seismic 33. Betten J (1988) Applications of tensor functions to the
assessment procedures for masonry arch bridges. Constr formulation of yield criteria for anisotropic materials. Int J
Build Mater 38:381–394. 25th Anniversary session for ACI Plast 4(1):29–46
228—building on the past for the future of NDT of concrete 34. Oller S, Car E, Lubliner J (2003) Definition of a general
14. Petracca M (2016) Computational multiscale analysis of implicit orthotropic yield criterion. Comput Methods Appl
masonry structures. PhD thesis, Universitat Politècnica de Mech Eng 192:895–912, 02
Catalunya. Departament d’Enginyeria Civil i Ambiental 35. Oller S, Botello S, Miquel J, Oñate E (1995) An anisotropic
15. Silva L, Lourenco P, Milani G (2017) Rigid block and elastoplastic model based on an isotropic formulation. Eng
spring homogenized model (HRBSM) for masonry sub- Comput 12:245–262
jected to impact and blast loading. Int J Impact Eng 36. Pelà L, Cervera M, Roca P (2011) Continuum damage
109:14–28, 11 model for orthotropic materials: application to masonry.
16. Pelà L, Cervera M, Roca P (2013) An orthotropic damage Comput Methods Appl Mech Eng 200(9):917–930
model for the analysis of masonry structures. Constr Build 37. Cornejo A, Jiménez S, Barbu L, Oller S, Oñate E (2022) A
Mater 41:957–967 unified non-linear energy dissipation-based plastic-damage
17. Petracca M, Pelà L, Rossi R, Oller S, Camata G, Spacone E model for cyclic loading. Comput Methods Appl Mech Eng
(2016) Multiscale computational first order homogenization 400:115543
of thick shells for the analysis of out-of-plane loaded 38. Petracca M, Pelà L, Rossi R, Zaghi S, Camata G, Spacone E
masonry walls. Comput Methods Appl Mech Eng 315:11 (2017) Micro-scale continuous and discrete numerical
18. Noor-E-Khuda S, Dhanasekar M, Thambiratnam DP (2016) models for nonlinear analysis of masonry shear walls.
An explicit finite element modelling method for masonry Constr Build Mater 149:296–314
walls under out-of-plane loading. Eng Struct 113:103–120 39. Bazant Z, Oh B (1983) Crack band theory for fracture of
19. Abdulla KF, Cunningham LS, Gillie M (2017) Simulating concrete. Mat Constr 16:155–177, 05
masonry wall behaviour using a simplified micro-model 40. Cervera M, Hinton E, Hassan O (1987) Nonlinear analysis
approach. Eng Struct 151:349–365 of reinforced concrete plate and shell structures using
20. Noor-E-Khuda S, Dhanasekar M (2018) Masonry walls 20-noded isoparametric brick elements. Comput Struct
under combined in-plane and out-of-plane loadings. J Struct 25(6):845–869
Eng 144(2):04017186 41. Bazant Z, Planas J (2019) Fracture and size effect in con-
21. Lemos JV (2007) Discrete element modeling of masonry crete and other quasibrittle materials, vol 3. CRC Press,
structures. Int J Archit Herit 1(2):190–213 Boca Raton
151 Page 18 of 18 Materials and Structures (2023) 56:151
42. Cervera M, Tesei C (2017) An energy-equivalent d?/d- seismic assessment of masonry structures. Comput Struct
damage model with enhanced microcrack closure-reopen- 252:106561
ing capabilities for cohesive-frictional materials. Materials 51. Teschemacher T (2022) CAD-integrated constitutive mod-
10(4):433 elling, analysis and design of masonry structures. PhD
43. Vermeltfoort A, Raijmakers T (1993) Deformation con- thesis, Technische Universität München, Lehrstuhl für
trolled tests in masonry shear walls, part 2 Statik
44. Vermeltfoort A, Raijmakers T (1992) Deformation con- 52. Teschemacher T, Wüchner R, Bletzinger K-U (2023)
trolled tests in masonry shear walls. Research report tno- Parametric cad-integrated simulation of masonry structures
bouw. report b-92-1156 (in Dutch) based on the isogeometric analysis. Comput Struct
45. Lourenço P (1996) Computational strategies for masonry 53. KratosMultiphysics
structures. PhD thesis, TU Deflt, Civil Engineering and 54. Dadvand P, Rossi R, Oñate E (2010) An object-oriented
Geosciences environment for developing finite element codes for multi-
46. Milani G (2011) Simple homogenization model for the non- disciplinary applications. Arch Comput Methods Eng
linear analysis of in-plane loaded masonry walls. Comput 17(3):253–297
Struct 89(17):1586–1601 55. Teschemacher T, Bauer AM, Aristio R, Messmer M,
47. Fu Q, Qian J, Beskos D (2018) Inelastic anisotropic con- Wüchner R, Bletzinger K-U. Cocodrilo
stitutive models based on evolutionary linear transforma- 56. Teschemacher T, Bauer AM, Aristio R, Messmer M,
tions on stress tensors with application to masonry. Acta Wüchner R, Bletzinger K-U (2022) Concepts of data col-
Mech 229:02 lection for the cad-integrated isogeometric analysis. Eng
48. Bilko P, Małyszko L (2020) An orthotropic elastic–plastic Comput 38:1435–5663
constitutive model for masonry walls. Materials 57. Teschemacher T, Bauer AM, Oberbichler T, Breitenberger
13(18):4064 M, Rossi R, Wüchner R, Bletzinger K-U (2018) Realization
49. Pulatsu B, Erdogmus E, Lourenco P, Lemos J, Tuncay K of cad-integrated shell simulation based on isogeometric
(2020) Simulation of the in-plane structural behavior of b-rep analysis. Adv Model Simul Eng Sci 5(1):1–54
unreinforced masonry walls and buildings using dem.
Structures 27:2274–2287, 10
Publisher’s Note Springer Nature remains neutral with
50. Maria D’Altri A, Lo Presti N, Grillanda N, Castellazzi G, de
regard to jurisdictional claims in published maps and
Miranda S, Milani G (2021) A two-step automated proce-
institutional affiliations.
dure based on adaptive limit and pushover analyses for the