Since the aforementioned peculiarit

Since the aforementioned peculiarities make AAC different from
conventional masonry materials, it is of fundamental importance
to provide a detailed experimental characterization of its
http://dx.doi.org/10.1016/j.conbuildmat.2015.08.121
0950-0618/? 2015 Elsevier Ltd. All rights reserved.
⇑ Corresponding author.
E-mail addresses: [e-mail] (D. Ferretti), [e-mail]
(E. Michelini), [e-mail] (G. Rosati).
Construction and Building Materials 98 (2015) 353–365
Contents lists available at ScienceDirect
Construction and Building Materials
journal homepage: www.elsevier.com/locate/conbuildmat
mechanical behavior. Test results allow indeed to calibrate sophis-
ticated numerical models to be used in structural analyses, both for
the design of new structures and for the assessment/retrofit of
existing ones [12].
As known, masonry assemblages such as shear walls, infill
walls in framed construction or walls supported on beams are
generally subjected to a biaxial state of stress, due to the presence
of normal stresses parallel and perpendicular to the bed joints, as
well as of shear stresses along the joints themselves. Moreover,
unreinforced conventional masonry generally exhibits anisotropic
properties due to its composite structure, with mortar joints act-
ing as planes of weakness. Therefore, its failure cannot be
described solely in terms of the two principal stresses, but a third
variable – related to bed joint orientation – must be also consid-
ered. For these reasons, several researches in the past focused
their attention on the experimental determination of reliable
parameters of masonry strength, as well as on the development
of failure criteria for masonry elements subjected to in-plane
loading (e.g. [13–16]). One of the most complete experimental
campaigns relative to masonry subjected to proportional biaxial
loading was carried out by Page [17,18]. These tests were per-
formed on half-scale square panels made of solid clay units to
investigate the influence exerted on failure mode and strength
by bed joint orientation (with respect to the vertical principal
stress direction), as well as by the applied principal stress ratio.
Based on these experimental data, biaxial failure surfaces were
first derived in terms of the two principal stresses and their orien-
tation to bed joints [17,18] and subsequently in terms of the
stress system related to the direction of the joints [19], which is
better suited for finite element modeling. Moreover, the same test
results were also used in [20] to determine macroscopic elastic
and non-linear stress–strain relations. However, it is worth notic-
ing that the strength envelope obtained by Page is of limited
applicability for other types of masonry, characterized by different
materials, unit shapes and/or geometry. For example, the influ-
ence of joint orientation was found to be less significant for
grouted concrete masonry, whose experimental behavior under
biaxial stresses seems to be essentially isotropic [21]. Further
experimental investigations were also carried out on masonry
panels subjected to in-plane forces, by considering different unit
geometries and materials (concrete blocks, calcium-silicate blocks
and clay bricks [22], or grouted unreinforced brick masonry [23]),
so as to define suitable failure criteria.
It should be also remarked that the results provided from the
abovementioned experimental programs could be hardly extended
to AAC masonry, also because this latter belongs to ‘‘thin bed
masonry” typology. The units are indeed connected together
through thin glue layers, with thickness usually ranging between
0.5 and 3 mm. Researches carried out on thin bed masonry (among
others, e.g. [24–27]) have shown that joint thickness significantly
affects masonry behavior. As an example, the compressive strength
of thin bed masonry is higher with respect to conventional
masonry, since it tends to approach the strength of the blocks.
Moreover, its shear and flexural strengths are not significantly
affected by the interface bond behavior and therefore thin bed
masonry performs more similarly to a continuum under loading,
without excessive localization of the failure path along the joints.
Biaxial compression tests carried out by Vermeltfoort [28] on thin
bed masonry panels with different joint orientations showed that
their failure mechanism was characterized by the three following
phenomena: spalling of the units with fragments of approximately
20 mm, vertical splitting, and bending of the sample.
It is not clear if AAC masonry displays the same behavior. The
most of the experimental programs carried out in recent years on
this specific type of masonry were indeed mainly focused on the
assessment of its seismic performances and were devoted to the
development of seismic design provisions to be included in Design
Codes (among others, e.g. [8–10,29–31]).
Aim of this research work is to provide a complete description
of AAC masonry behavior under in-plane static loading, with par-
ticular attention to the softening regime. To the purpose, several
experimental tests are performed on both AAC masonry panels
and beams, so as to characterize masonry behavior under uniaxial
and biaxial compression, flexure and shear, determining not only
its elastic parameters and strength values, but also the fracture
energies in tension and compression. The collected experimental
data can be useful in the calibration of suitable numerical models;
as an example, in this work an anisotropic nonlinear constitutive
model, well-known in the technical literature for FE analysis of
ordinary masonry structures [32–34], is adapted to AAC masonry
elements. The so calibrated model is subsequently validated by
performing a nonlinear FE analysis on a full-scale AAC masonry
wall subjected to a pushover test [31].
2. Experimental program on AAC masonry elements
The performed experimental program consisted in 33 tests on AAC masonry
elements with thin bed joints. A general view of some of the assembled specimens
before testing is shown in Fig. 1.
In order to characterize the material behavior in compression, several mono-
tonic uniaxial and biaxial tests were performed on small-scale masonry panels,
by varying the bed joint orientation with respect to the horizontal axis. Some of
the uniaxial tests were carried out under displacement control to obtain the com-
plete stress–strain curve and the corresponding fracture energy in compression.
Due to the limited features of the universal testing machines at our disposal,
tensile behavior of AAC masonry was instead investigated indirectly, by performing
three-point bending tests on small-scale masonry beams. In this case, only two
angles of inclination between the bed joints and the horizontal axis were consid-
ered (namely 0? and 90?). For each examined typology, two beams were provided
of a central notch to guide crack formation and were tested under crack mouth
opening displacement (CMOD) control, so obtaining the complete load–deflection
response and the corresponding fracture energy in tension. Finally, two small
masonry panels were subjected to diagonal compression tests.
2.1. Description of test specimens
All the specimens were prepared by using masonry-type AAC units directly
provided by the Manufacturer. It is worth noticing that AAC units are commonly
produced in different sizes that may reach 625 ? 250 ? 200 mm; in this last case,
masonry panels including a representative number of head and bed joints would
be huge and some problems may arise to test them into the frame of a universal
testing machine. However, considering that the dimensions of the units available
on the market are variable, in the present work it has been preferred to employ
non-standard small size bricks, with nominal dimensions equal to
250 ? 50 ? 100 mm. In this way, it has been possible to keep the specimen size
small, while having at the same time an adequate number of head and bed joints,
so emphasizing their influence on masonry global behavior and increasing possible
anisotropic effects. Therefore, the behavior of structural elements realized with

結果 (中文) 1:

Since the aforementioned peculiarities make AAC different fromconventional masonry materials, it is of fundamental importanceto provide a detailed experimental characterization of itshttp://dx.doi.org/10.1016/j.conbuildmat.2015.08.1210950-0618/? 2015 Elsevier Ltd. All rights reserved.⇑ Corresponding author.E-mail addresses: [e-mail] (D. Ferretti), [e-mail](E. Michelini), [e-mail] (G. Rosati).Construction and Building Materials 98 (2015) 353–365Contents lists available at ScienceDirectConstruction and Building Materialsjournal homepage: www.elsevier.com/locate/conbuildmatmechanical behavior. Test results allow indeed to calibrate sophis-ticated numerical models to be used in structural analyses, both forthe design of new structures and for the assessment/retrofit ofexisting ones [12].As known, masonry assemblages such as shear walls, infillwalls in framed construction or walls supported on beams aregenerally subjected to a biaxial state of stress, due to the presenceof normal stresses parallel and perpendicular to the bed joints, aswell as of shear stresses along the joints themselves. Moreover,unreinforced conventional masonry generally exhibits anisotropicproperties due to its composite structure, with mortar joints act-ing as planes of weakness. Therefore, its failure cannot bedescribed solely in terms of the two principal stresses, but a third— — 相关的床联合定位 — — 变量必须也考虑-走进。出于这些原因，在过去的几个研究重点他们的注意力上可靠的实验测定参数的砌体强度和开发砌体构件，在平面的破坏准则加载 (例如 [13-16])。最完整的实验之一砌体相对运动受到比例双轴[17，18] 页进行加载。这些试验是每-形成于半规模方板由实心粘土单位调查对失效模式和强度形成的影响由床联合方向 (垂直校长应力方向)，以及应用的主应力比。基于这些实验数据，双轴失效面了首先导出了在两个主应力和他们容易-张力对床接头 [17，18] 以及随后在职权范围强调系统相关方向的关节 [19]，这是更适合有限元建模。此外，同一测试结果也被用在 [20] 来确定宏观弹性与非线性应力-应变关系。然而，值得注意的是-ing 强度包络线获得由页很有限对于其他类型的砌体，特点是由不同的适用性材料、 单元形状或几何。例如，影响-ence of joint orientation was found to be less significant forgrouted concrete masonry, whose experimental behavior underbiaxial stresses seems to be essentially isotropic [21]. Furtherexperimental investigations were also carried out on masonrypanels subjected to in-plane forces, by considering different unitgeometries and materials (concrete blocks, calcium-silicate blocksand clay bricks [22], or grouted unreinforced brick masonry [23]),so as to define suitable failure criteria.It should be also remarked that the results provided from theabovementioned experimental programs could be hardly extendedto AAC masonry, also because this latter belongs to ‘‘thin bedmasonry” typology. The units are indeed connected togetherthrough thin glue layers, with thickness usually ranging between0.5 and 3 mm. Researches carried out on thin bed masonry (amongothers, e.g. [24–27]) have shown that joint thickness significantlyaffects masonry behavior. As an example, the compressive strengthof thin bed masonry is higher with respect to conventionalmasonry, since it tends to approach the strength of the blocks.Moreover, its shear and flexural strengths are not significantlyaffected by the interface bond behavior and therefore thin bedmasonry performs more similarly to a continuum under loading,without excessive localization of the failure path along the joints.Biaxial compression tests carried out by Vermeltfoort [28] on thinbed masonry panels with different joint orientations showed thattheir failure mechanism was characterized by the three followingphenomena: spalling of the units with fragments of approximately20 mm, vertical splitting, and bending of the sample.It is not clear if AAC masonry displays the same behavior. Themost of the experimental programs carried out in recent years onthis specific type of masonry were indeed mainly focused on theassessment of its seismic performances and were devoted to thedevelopment of seismic design provisions to be included in DesignCodes (among others, e.g. [8–10,29–31]).Aim of this research work is to provide a complete descriptionof AAC masonry behavior under in-plane static loading, with par-ticular attention to the softening regime. To the purpose, severalexperimental tests are performed on both AAC masonry panelsand beams, so as to characterize masonry behavior under uniaxialand biaxial compression, flexure and shear, determining not onlyits elastic parameters and strength values, but also the fractureenergies in tension and compression. The collected experimentaldata can be useful in the calibration of suitable numerical models;as an example, in this work an anisotropic nonlinear constitutivemodel, well-known in the technical literature for FE analysis ofordinary masonry structures [32–34], is adapted to AAC masonryelements. The so calibrated model is subsequently validated byperforming a nonlinear FE analysis on a full-scale AAC masonrywall subjected to a pushover test [31].2. Experimental program on AAC masonry elementsThe performed experimental program consisted in 33 tests on AAC masonryelements with thin bed joints. A general view of some of the assembled specimensbefore testing is shown in Fig. 1.In order to characterize the material behavior in compression, several mono-tonic uniaxial and biaxial tests were performed on small-scale masonry panels,by varying the bed joint orientation with respect to the horizontal axis. Some ofthe uniaxial tests were carried out under displacement control to obtain the com-plete stress–strain curve and the corresponding fracture energy in compression.Due to the limited features of the universal testing machines at our disposal,tensile behavior of AAC masonry was instead investigated indirectly, by performingthree-point bending tests on small-scale masonry beams. In this case, only twoangles of inclination between the bed joints and the horizontal axis were consid-ered (namely 0? and 90?). For each examined typology, two beams were providedof a central notch to guide crack formation and were tested under crack mouthopening displacement (CMOD) control, so obtaining the complete load–deflectionresponse and the corresponding fracture energy in tension. Finally, two smallmasonry panels were subjected to diagonal compression tests.2.1. Description of test specimens
All the specimens were prepared by using masonry-type AAC units directly
provided by the Manufacturer. It is worth noticing that AAC units are commonly
produced in different sizes that may reach 625 ? 250 ? 200 mm; in this last case,
masonry panels including a representative number of head and bed joints would
be huge and some problems may arise to test them into the frame of a universal
testing machine. However, considering that the dimensions of the units available
on the market are variable, in the present work it has been preferred to employ
non-standard small size bricks, with nominal dimensions equal to
250 ? 50 ? 100 mm. In this way, it has been possible to keep the specimen size
small, while having at the same time an adequate number of head and bed joints,
so emphasizing their influence on masonry global behavior and increasing possible
anisotropic effects. Therefore, the behavior of structural elements realized with

結果 (中文) 3:

由于上述特点使AAC不同于
传统砌体材料，它提供的HTTP
详细的实验表征是至关重要的
：/ / DX。DOI。org / 10.1016 / j.conbuildmat。2015.08.121
0950-0618 /？2015 Elsevier公司保留所有权利。
⇑通讯作者。
电子邮件地址：[e-mail]（D. Ferretti），[e-mail]
（如该），[e-mail]（G.罗萨蒂）。
施工和建筑材料，98（2015）353–365
目次在ScienceDirect
建筑建材
杂志主页：www.elsevier。COM /定位/ conbuildmat
力学行为。测试结果让的确对索菲亚-
表明数值模型进行结构分析，为
设计的新结构和新的评估和改造现有的
[ 12 ]。
众所周知，砌体剪力墙的组合，如，在框架结构或墙体支撑梁上填充墙一般

双轴应力状态下，由于正常应力平行和垂直于床缝
的存在，以及作为
剪沿关节应力。此外，
无筋砌体一般表现出各向异性
常规属性由于其复合结构，砂浆缝法-
为飞机的弱点。因此，它的失败不能
在两主应力仅仅描述，但三分之一
变–与床关节定位–也必须考虑到
。因为这些原因，在过去的几方面的研究集中在实验测定的可靠的砖砌体强度的参数，以及在平面内的砌体元素的破坏标准的发展（例如，[ 13（16）]）。一个最完整的实验
活动相对受到比例双轴
加载砌体是由页[17,18]进行。这些测试是由实心粘土单位制成的半规模的方形面板，调查的影响施加的故障模式和强度的床联合方向（相对于垂直的主方向），以及由所施加的主应力比。首先推导出的两个主应力及其方向
要床缝[17,18]和随后的
应激系统相关方向的关节[ 19 ]，这是
更适合有限元建模。此外，同样的测试结果也被用于在[ 20 ]，以确定宏观的弹性和非线性应力-应变关系。不过值得注意的
-，强度包络线得到的页面是有限的
适用其他类型的砌体，其不同的
材料，单元形状和/或几何。例如，影响
性关节的方向被认为是不重要的
灌浆混凝土砌块，其实验行为
双轴应力下似乎基本上是各向同性的[ 21 ]。进一步的实验研究也
砌体
板面内的力量进行，采用不同的单元
几何形状和材料（混凝土块，硅酸盐块
和粘土砖[ 22 ]，或注浆加固砖砌体[ 23 ]），
从而确定合适的破坏准则。
还应评论说，从
提供的结果上述实验程序可以扩展到几乎
AAC砌体，也因为后者属于“轻系列床
砌体类型。这些单位实际上是连接在一起，通过薄的胶合层，厚度通常为0.5和3毫米之间。研究进行了薄床砌体（其中），例如，[ 24（27）]）已经表明，联合厚度显着影响砌体行为。作为一个例子，薄床砌体的抗压强度的增加相对于传统的砖砌体，因为它往往接近的强度的块。此外，它的剪切和弯曲强度没有显着影响的界面粘结的行为，因此薄床层的砌体进行更类似于一个连续的负载下，没有沿接头的失效路径过度本土化。
双轴压缩试验vermeltfoort [ 28 ]薄
床砌体的不同关节方向进行显示，
失效机制主要为以下三
现象：剥落的碎片
单位约20 mm，垂直分裂，和样品
弯曲。