8/10/2019 Onur SEREN (03.02.13) http://slidepdf.com/reader/full/onur-seren-030213 1/109 i SEISMIC RETROFITTING OF UNREINFORCED LOAD BEARING BRICK WALLS IN HISTORIC BUILDINGS USING FIBER-REINFORCED POLYMER STRINGS by Onur Seren B.S. , Civil Engineering, Boğaziçi University, 2009 Submitted to the Institute for Graduate Studies in Science and Engineering in partial fulfillment of the requirements for the degree of Master of Science Graduate Program in Civil Engineering Boğaziçi University 2013
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
I would like to express my grat itude to everyone who contributed the development of
this research. I would like to express my sincere gratitude to my advisor Assoc. Prof. Dr.
Cem Yalçın for his valuable help in instructing, guiding and supporting me throughout theduration of this thesis.
Also, I would like to thank the members of my Master’s thesis examinationcommittee; Assist. Prof. Ku tay Orakçal, and Assoc. Prof. Ercan Yüksel, for their in-depth
comments and advice.
I would like to thank Assist. Prof. Ahmet Anil Dindar for his important contribution
to the analysis stage of this study with the software he developed.
Civil Engineer Ali Bayraktar and Hafez Keypour from SGM Construction provided
required materials and contributed this research financially, which is highly appreciated.
Special thanks to my friends Tevfik Terzioğlu, Hasan Altun and Furkan Çelenli fortheir assistance and suggestions during the construction and testing of the specimens, and
to the technicians, Hasan Şenel, Hamdi Ayar, Ümit Melep and Mesut Kardaş for theirhelp
in the experimental phase of this research.
I would like to thank my supervisors Ramiz Soylu and Tayfun Bayramkaya from
SURYAPI End. Tic. A.Ş. for their patience and tolerance at work for the time required forthis thesis to be completed.
Finally, I would like to thank my family and my love Ece Uçar for their continuoussupport and encouragement.
Most of the historical buildings and monuments in the world are unreinforced
masonry (URM) type and they are vulnerable to seismic actions. Considering the seismic
activities in the regions where historical masonry structures are located, their structural
assessment and rehabilitation or retrofitting if necessary against seismic forces are needed
in order to preserve them for future generations. International organizations such as
UNESCO (United Nations educational, Scientific and Cultural Organization) and
ICOMOS (International Council on Monuments and Sites) try to increase the awareness in
need for preserving these world-heritage structures. However, historical masonry structures
still need retrofitting techniques that are much different than that of buildings that were
built using conventional construction practice since the architectural features of the historic
buildings must remain unchanged after the retrofitting process. Therefore, conventional
methods have been used in retrofitting works are not suitable for such purposes. In this
study, the use of carbon fiber-reinforced polymer (CFRP) strings placed in mortar joints
for strengthening of URM structures was investigated. Nearly full-scaled four URM brick
wall specimens with aspect ratio of 1.00 were tested under varying axial load and cyclic
lateral loading. Two of the specimens were tested as control specimens while other two
specimens were retrofitted with horizontally-oriented CFRP strings. Also, one of the tested
control specimens were repaired and retrofitted and re-tested in order to see the effect of
this retrofitting technique after substantial damage occurred. Test results showed that
energy dissipation capacity of the wall specimens were enhanced with the proposed
technique. In addition, the crack openings due to shear effects were minimized while
keeping the historic and aesthetical view of the structures intact, since they were directly
placed inside the mortar joints and debonding of strings were prevented. However, nosignificant increase in the lateral load carrying capacity of the specimens was observed.
Yeryüzündeki birçok tarihi bina ve anıtın kâgir oluşu bu yapıları sismik etkilere karşı
savunmasız kılmaktadır. Tarihi yapıların konumlandığı bölgelerdeki sismik aktiviteler gözönünde bulundurulduğunda; yapısal durum değerlendirme, iyileştirme ve gerekligörüldüğünde sismik etkilere karşı güçlendirme çalışmalarının yapılması, bu yapılarıngelecek nesillere aktarılabilmesi adına zorunluluk arz etmektedir. UNESCO ve ICOMOS
gibi uluslararası organizasyonlar , dünya mirası olarak nitelendirilen bu yapıların geleceknesillere aktarılabilmesi adına gerekli farkındalığın oluşturulması için çalışmalaryürütmektedirler.Ancak, tarihi kâgir yapılar için, güçlendirme sonrası tarihi doku ve
mimari özelliklerin korunması gerektiğinden, diğer yapılarda kullanılan tekniklerden farklıyöntemlere ihtiyaç duyulmaktadır. Bu nedenle, tarihi kâgir yapıların güçlendirmeçalışmalarında kullanılan konvansiyonel teknikler, bahsi geçen zorunlulukların sağlanması konusunda yetersiz kalmaktadır . Bu tez çalışmasında, fiber takviyelikarbon polimer iplerin
(CFRP) derz aralarında çekme elemanı olarak kullanılmasıyla donatısız tuğla-yığma yapıların depreme karşı güçlendirilmesi konusunda çalışılmıştır. Bu yöntemin etkinliği,gerçek ölçeğe yakın, 1.00 narinlik oranına sahip donatısız tuğla duvarnumuneleriyle,
değişen düşey ve periyodik yatay yük tesirleri altında test edilmiştir . Numunelerden ikisikontrol numunesi olar ak kullanılırken, diğer iki numune, derz aralarına yatay d oğrultuda yerleştirilen CFRP ipler ile güçlendirilmiştir. Ayrıca, kontrol numunelerinden biri,
güçlendirme yönteminin ağır hasarlı bir yapıda etkisini incelemek adına, test edildikten
sonra tamir ed ilip güçlendirilmiş ve yeniden test edilmiştir. Test sonuçları, numunelerinenerji sönümleme kabiliyetlerinin önerilen güçlendirme tekniği ile arttığını göstermiştir. Buna ek olarak, CFRP ipler in doğrudan derz aralarına uygulanması ve sıyrılmalarınınönlenmesi numunelerde kesme etkisiyle oluşan çatlakların azal masını sağlarken, yapınıntarihi ve estetik görünümünü korumuştur. Ancak, yapılan testlerde numunelerin yatay yüktaşıma kapasitelerinde belirgin bir artış tespit edilememiştir.
It is of great importance, due to the probability of strong seismic event occurrence
being high in the near future, to determine the seismic safety of historical masonry
structures and to improve and spread the technical knowledge for strengthening, especially
considering our country which experienced the devastating earthquakes in the year 1999.
Masonry structures are usually rigid, highly resistant against compressive stresses, but in
horizontal direction effect, especially in plane and out of plane forces induced by
earthquakes, are very weak and could get severe damages. Therefore, they are be classified
as brittle in nature and one of the most vulnerable among the different types of structural
buildings under seismic loads. Besides, masonry structures are one of the oldest types
among the historical buildings and it is necessary to preserve them for contributing
common heritage of mankind. For that reason, restoration of historical buildings in the
earthquake zones, and continuously strengthening them is a major necessity.
According to Calvi et al . (1996), lateral load resistance of masonry structures is
highly dependent on shear resistance of in-plane walls. In addition, shear resistance of in-
plane walls are directly related to its constituents; masonry brick, binding unit=mortar
ability and workmanship. Therefore, improvement techniques should target these
constituents’ bonding and adhesion capacities (Somerset al ., 1996). For the historical
building cases, the chosen method of seismic retrofitting must preserve the architecturaland historical features of the structure. A variety of techniques have been applied for
strengthening historical masonry structures. However, most of the methods do not take into
account the historical features of the building, which leads incompatible views of the
interior and exterior facades and thus, losing the entire historical features of the structure.
Use of lightweight materials, especially fiber-reinforced polymer (FRP) composites
in the form of strips or sheets, have a significant role in the development of repairing andstrengthening of civil engineering structures due to their superior properties such as cost
effectiveness, high tensile strength and ease of application. In this study, by taking the
advantages of using FRP and the necessity of preserving architectural and historical
features of the structures into account, placing carbon fiber reinforced polymer (CFRP)into the brick masonry load carrying structural wa lls’ joints may offer the optimum
strengthening technique for historic unreinforced masonry structures.
1.2. Problem Definition
Historic and older buildings are vulnerable against forces induced by earthquakes
due to their well-known brittle and inflexible behavior. Furthermore, solid mass and heavy
weight of the materials used in these buildings could increase the probability of
counteracting with high seismic forces. Therefore, severe damage followed by collapse
mechanism could be observed in structural supports, walls, floors, stairs and other
structural members.
Under earthquake effect, tension zones become critical for the load carrying walls in
historical buildings. Considering low tensile resistance capacity of bricks and mortar,
diagonal cracking could start developing and rapidly propagate within the member or wall
element. Therefore, reinforcement against tension is needed to be implemented within the
members during repairing and renovation of the structure. However, since the architectural
and historic condition of the existing structure is needed to be preserved, the retrofitting
method should not give any damage or alter its architectural and historic condition (Arun
G.,2005; Altın et al. , 2005).
There are conventional methods used for retrofitting and restoration of historical
buildings. The tension capacity of the critical or damaged sections on the structural
members are generally increased by means of additional reinforced concrete or steel
supporting members, steel clamping, jacketing with plaster (Bayraktar, 2006). Although
these methods are commonly used in Turkey, they have serious disadvantages such as
difficulty in application and causing damage on historical and architectural view of the
structures. Also, the integrity of the structures no longer exists after these applications.
As it is seen from the above figures, wrong applications have damaged or altered the
structures permanently in terms of historical and architectural point of view. Also, the
structural continuity and integrity of the structures are disturbed by these applications.Therefore, more practical and efficient techniques are needed in retrofitting of historical
structure.
In order to overtake the excessive tensile forces developed in load carrying walls,
CFRP strings could be used as tension elements. The application differs from other
conventional techniques since the CFRP strings are embedded into mortar joints
horizontally and vertically which is also called as structural repointing (SR). Infunctionality point of view, it does not affect the structure visually and at the same time it
provides high durability against tension due to the high strength capacity of CFRP
(Tumialan and Nanni, 2002).
Representative application of CFRP strings shown in Figure 1.5.
CFRP strings in horizontal direction
Reinforcing Plaster
Existing wall
CFRP strings in vertical direction
Figure 1.5. A representative sketch for a sample application of CFRP strings.
This research investigates and evaluates the performance of CFRP strings as a means
for seismic strengthening technique that are horizontally placed between the joints of
unreinforced masonry brick walls. The wall specimens represent the load carrying walls of
a historical building and they are constructed with blend brick and Horasan mortar which
are the common construction materials exist in historical buildings in Anatolia region of
Turkey (Arioglu and Acun, 2006).
1.3. Literature Review
1.3.1. Mechanical Properties of URM Structures and Their Components
Behavior of masonry assemblage is highly dependent on the characteristics and
interface of their constituents. Usually, masonry walls have high compressive strength
whereas their tensile strength capacity is low. Besides, non-homogeneity of masonry units
and complexity of the interaction between masonry unit and mortar make it hard to predict
the lateral load capacity of these structures. Therefore, understanding these properties is
important for going further in seismic in-plane behavior of URM structures.
There are various researches on identifying the mechanical properties of masonry
structures both in individual material case and their interaction as masonry member: brick
unit and mortar, and shear/tensile bond strength and interface friction (McNary and
Abrams, 1985; Binda et al. , 1994; Atkinson et al ., 1994).
McNary and Abrams (1985) studied on different types of mortars and brick units that
vary in strength. Compression tests were performed for indicating the effect of
confinement in increasing compressive strength and ductility of mortar. In addition, the
tensile strength of mortar found to be negligible compared to its compressive strength. Rad(1978) examined the variation of compressive strength of different types of brick units. In
these studies, it was found to be that compressive strength of brick units on average was 2
to 3 times larger than the tensile strength. It is found that typical compressive strengths of
clay bricks range between 8.60 MPa and 17.20 MPa (Rad, 1978).
The interface between mortar and brick unit is ruled by two mechanisms; bonding
due to chemical interaction and friction. Thus, depending on the mechanism, two main
types of failures are associated with brick-unit mortar interface which are tension (mainly
governed by chemical bond) and shear (mainly ruled by friction).
1.3.2. Seismic In Plane Behavior of URM Structures
According to Vas concelos and Lourenço (2009), with the condition of prevented outof plane failure, resistance of URM structures against seismic action was sustained by in
plane behavior of masonry walls. In an earthquake, in-plane walls could deform or fail due
to diagonal shear failure, rocking at toe sections of the walls, sliding shear deformation
along bed joint and compression failure (toe crushing) (Magenes and Calvi, 1997).
Analytical studies were performed in the scope of experimental researches for
identifying the failure mechanisms due to in-plane forces that are subjected to URM walls
in terms of material properties, geometry and boundary conditions of the structures as seen
in Figure 1.6.
In general, rocking failure tends to prevail among other mechanism for masonry
walls that have slender geometry whereas sliding failure tends to occur in squat walls
contrary, masonry specimens under quasi-static loading exhibit more damage and lower
strengths compared to dynamic tests which could be defined as conservative. Therefore,
both testing techniques could result in differences in the evaluation of stiffness andstrength parameters of URM brick walls (Calvi et al. , 1996).
1.3.4. Conventional Retrofitting Techniques for Historical URM Structures Against
Seismicity
The need for preserving, restoring and strengthening of historical structures againstseismicity has been noticed for years. Development in interventions for strengthening
techniques has been followed by international collaborations since the Athens Charter for
the Historic Monuments 1931 (ICOMOS 1931). The general requirement for the
strengthening techniques are; being reversible in means of application and preserving the
character and features of the structures. In this section, by taking the expectations into
account, frequently used conventional strengthening and retrofitting techniques will be
reviewed.
1.3.4.1. Filling of Cracks Using Grout and Epoxy Injections. This technique has been
commonly used for filling the cracks and voids within the multi-wythe masonry structures
for maintaining the integrity. As a grout material, both epoxy resin and cement based
grouts could be used for injection. The methodology for this technique could be defined as
the following steps (Hamid et al .,1999; Calvi and Magenes,1994; Schuller et al .,1994):
As a first step, injection docks should be anchored in the determined sections. Then,
the openings around the docks and other cracks should be sealed.
Cracks and other openings should be cleaned with water by injecting water from the
docks.
Finally, grout should be injected with low injection pressure.
A case study was carried by Perret et al ., (2002) which evaluates the performance ofhigh strength cement grout in a 130-year old masonry bridge pier. According to in-situ
pullout bond and sonic tomographic tests, grouting technique managed to increase the
performance of the tested pier. Furthermore, studies of Schuller (1984) showed that this
technique could increase the compressive strength of URM walls up to 0.8 of the un-retrofitted masonry compressive strength. A real application of this technique was studied
by Valluzzi et al ., (2002) in Modena, Italy for the Bell Tower of Cathedral of Monza.
Although different studies on this technique proved that a significant increase in
ultimate load capacity is acquired, correct application requires a comprehensive pre-study
on composition of the type of grout and its penetration into the structure. In addition, sonic
tests should be conducted in order to evaluate the effectiveness of the technique during theapplication (Valluzzi, 2007).
1.3.4.2. External Jacketing by Shotcreting. The principle of jacketing technique for
strengthening reinforced concrete (RC) structures is also valid for masonry structures as
well. Masonry members under excessive compression are confined by either reinforced
concrete units or steel plates. Since the treatment is practiced on the surface, the historical
and architectural features of the structure are highly affected by this technique.
Other than RC and steel members, ferrocement, reinforced plasters, and shotcrete
could be used.
Ferrocement could be defined as mesh of fine rods placed in a high-strength mortar
matrix. Abrams and Lynch (2001) found that lateral resistance of masonry walls that had
been retrofitted by ferrocement technique increased by a factor of 1.5 (Abrams and Lynch,2001).
In contrast to ferrocement intervention, high strength reinforcing steel is covered by a
thin plaster layer in reinforced plaster surface treatment technique. Shepperd and Tercelj
(1980) studied on this technique and in the reference of diagonal compression and static
cyclic loading experiment results, it was found that in-plane resistance of masonry walls is
linearly proportional with the thickness of the application, mortar strength and
reinforcement quality whereas, it is inversely proportional with the damage condition of
1.3.5. Evaluation of the Performance FRP Retrofitted Historical URM Structures
with FRP
The use of FRP has been increasing in various industries with discovering the
effectiveness of these materials in strength, durability, and workability. FRP materials have
been already introduced to structural engineering by means of strengthening reinforced
concrete structures which extensive research could be found about in literature. For
retrofitting and strengthening of URM historical structures with FRP, studies mainly
focused on the structural walls and in-plane panels.
Use of FRP in reinforced concrete is followed by strengthening of masonry
structures both in out-of-plane and in-plane under cyclic loading, in the form of carbon
laminates (Schwegler, 1994; Abrams and Lynch, 2001). Schwegler studied on retrofitting
configuration with FRP laminates and compared single-side retrofitting with double side
retrofitting of squad specimens. The studies of Schwegler concluded by the finding that
full surface coverage and inclined plates were the best configuration (Schwegler, 1994).
Another retrofitting technique with FRP laminates were studied on cracked specimen
with diagonal configuration for seismic retrofitting of URM historical structures. The
studies showed that proposed technique with the diagonal configuration was unsuccessful
(ElGawady et al , 2005a). Similar study on both damaged and non-damaged specimens
showed that retrofitting with diagonal configuration was effective only in non-damaged
specimen case (Zhao et al , 2003).
In-plane static cyclic loading performance of URM walls that had been retrofitted
with FRP were evaluated before and after retrofitting procedure (ElGawady et al , 2007).
As URM wall specimen, one-half scale single-wythe walls that had been constructed using
half-scale hollow clay brick and weak mortar. Three specimens were tested as reference
specimens. Later on, the damaged specimens were retrofitted by FRP on the surface and
tested again. One specimen was retrofitted directly after the construction stage and tested.
In total, seven specimens were tested. Experiment results proved that for particularspecimens, lateral load capacity were increased after retrofitting by a factor of 1.4 to 5.9.
Besides, it was found that the severity of the existing damage in the specimen before
retrofitting had an influence in the ultimate lateral strength of the specimen after
retrofitting. Furthermore, it was observed that cracking load and pattern were effectivelycontrolled after FRP retrofitting.
Performance of in-plane shear behavior of URM that strengthened by near surface
mounted (NSM) CFRP strings was experimentally investigated (Petersen et al ., 2010). In
this study, different orientation of CFRP strings was tested including effect of
nonsymmetrical reinforcement. Dimension of all solid clay brick panel specimens was
1.2m x 1.2m (aspect ratio of 1.00) whereas unidirectional pultruded CFRP strings were15mm wide and 2.8mm thick. CFRP strings were glued using epoxy, into rectangular
grooves which were 20mm deep and 6mm wide and had been cut into surface of the
masonry panels by a circular saw. Seven URM panels with and four URM panels without
FRP strengthening were tested under diagonal shear compression test. The test results
proved that vertical orientation of CFRP strings prevented sliding failure effectively. In
addition, it was found to be nonsymmetrical reinforcement didn’t cause any change in in-
plane behavior of URM panels. Furthermore, it was observed that diagonal cracking was
prevented by horizontal oriented CFRP strings.
More recent study that examine the effectiveness of FRP systems as a seismic retrofit
intervention for in-plane loaded URM walls under seismic effects was done by Mahmood
and Ingham (2011). Seventeen URM wallettes were retrofitted with externally bonded
The intervention in FRP retrofitting of URM structures has great potential and a wide
research area. Common understanding of the mentioned researches is FRP retrofitting of
URM walls have influence in lateral load capacity of these structures. Besides, diagonalcracking and other in-plane failure mechanisms due to forces generated by seismic actions
could be prevented or controlled. However, nearly all applications require great surface
area on the structure which could result loss in aesthetical view and features in the
historical structures case. Therefore, studies on this issue should continue.
1.4. Research Significance and Rationale
The aim of this current research is to strengthen unreinforced masonry brick, load
carrying structural walls of older and historical structures against seismicity without
changing or affecting their architectural and historical features. The proposed technique in
this research is the retrofitting of load carrying masonry brick structural walls’ joints withfiber reinforced polymer, FRP strings. Strengthening methods that could positively affect
the tensile capacity of masonry members such as proposing use of high bonding and tensile
strength capacity mortar, use of GFRP, having different orientation of FRP strings other
than horizontal position or change in typology of wall formation are out of the scope of this
study. In addition, out-of-plane strengthening of the URM walls was not investigated.
An experimental research was carried on near full-scaled unreinforced load carrying
masonry brick walls in same typology. It aimed to evaluate and validate the performance
of proposed strengthening technique under in-plane cyclic lateral force action whether such
a technique is suitable for seismic strengthening of older and historical masonry structures
located in seismic regions.
1.5. Objective and Scope
This experimental study was mainly focused on the development and evaluation of
an applicable FRP retrofitting technique that increase the lateral load (shear) capacity of
The objectives of this study could be summarized as follows:
To evaluate the performance of FRP strings in increasing the lateral load capacity ofURM structural walls under the effect of seismic loading by comparing the test
results of retrofitted and post-strengthened specimen with control specimen,
To evaluate the performance of FRP strings in in-plane shear behavior of URM brick
walls,
To increase the resistance of URM brick walls against diagonal cracking,
To identify the effects of varying vertical load on unreinforced masonry structural
walls in lateral load capacity, To identify different failure mechanisms on masonry structural walls under cyclic
loading case,
To show FRP bands maintain the integrity in between masonry load carrying walls in
the structure properly.
In this study, unreinforced masonry structural wall specimens were tested under
varying vertical load and reversed cyclic loading. At the first phase of the study, a controlspecimen was tested in order to determine the behavior and natural lateral load capacity of
the member. As a second step, the first specimen had been repaired and strengthened, and
then was tested in order to compare the performance of the technique in post-strengthening
with control specimen. As a third step, second specimen that had been retrofitted by
mentioned technique was tested in order to evaluate and compare its performance with
control specimen. Fourth and final step was to test final specimen that had been retrofitted
by the same technique in addition to enhanced bonding of CFRP strings.
1.6. Methodology
This study investigates the actual behavior of unreinforced masonry brick structural
walls and evaluates the performance of CFRP string retrofitting that is developed for
improving the seismic behavior of historical masonry structures without damaging their
First of all, literature related with strengthening of historical masonry structures was
reviewed. Then, four specimens with the same typology, in other terms same formation of
bricks and use of same material, were produced. In scope of literature review, availablefacilities at the Structural Laboratory of Bogazici University were determined and a
preliminary test was conducted. According to the results of this test, the most suitable
testing conditions were determined. First three specimens were tested with the established
testing conditions and the results were compared. Fourth and the final specimen was
improved in order to prevent debonding of CFRP strings, and with this specimen the final
test was conducted.
1.7. Report Outline
This thesis presents the experimental research on the seismic behavior of
unreinforced load carrying URM structural walls and evaluation of retrofitting of these
members by using CFRP strings.
Brief information about the mechanical behavior of unreinforced load carrying
masonry structural walls under seismic action, literature review and previous studies
aiming strengthening historical masonry structures, and the objectives of the study with the
methodology are given in Chapter 1. In Chapter 2, experimental setup with the details of
construction, instrumentation and testing procedure is provided. Analysis on experimental
results, discussions and comparisons according to these results are presented in Chapter 3.
Finally, Chapter 4 gives a summary of the study, indicates the final outcomes, and
such that they could be successfully tested with the available facilities in the Structural
Laboratory of Bogazici University. A special type of mortar, Horasan that was used in the
Anatolian region for centuries, is used to build the walls.
Specimens were named in terms of their masonry unit constituent name (brick), test
number and initial letter of their condition. Accordingly, BW1-C refers to first tested brick
wall which was used as control specimen whereas BW2-RR stands for second tested brick
wall which is repaired and retrofitted. Similarly, BW3-R1 is the third tested brick wall with
retrofitting. And, BW4-R2 is the fourth tested brick wall with retrofitting.
One specimen (BW0) was tested for optimization of test setup. Second specimen
(BW1-C) was tested as it is, and later it was repaired, retrofitted (BW2-RR), and tested
again, while the other two (BW3-R1, BW4-R2) were only retrofitted and then tested.
CFRP strings were used as main retrofitting technique. They were horizontally inserted
between the brick elements and inside the mortar. A total of three lines at the top and three
lines at the bottom of the specimen with 20 cm offset of the strings were placed. The
middle portion of the walls is left non-retrofitted since the formation of shear cracks are
first expected to propagate at the corners of the top and the bottom sections (Figure 2.2).
Figure 2.2. FRP band layout.
Specimens were constructed on previously-built reinforced concrete foundations that
were designed for the experimental evaluation of the lateral load behavior of squat
structural walls by Terzioğlu T. (2011). The reinforced concrete structural wall had beenremoved from its foundation while the vertical reinforcement steel bars at 20-30 cm height
was kept in order to maintain a fixed support mechanism and prevent any premature
On the other side, CFRP strings with 5 meters of length each, were fully covered
with epoxy binder, BASF MBT-MBRACE, and left to dry for approximately 5 minutes.
Drying process is necessary for hardening of the CFRP strings which enables them to getthe desired shape. After the drying process was done, the CFRP strings were placed
horizontally in the specified sections where the Horasan mortar had been removed as
shown in Figure 2.6. The CFRP strings should be as tight as possible and should fully
cover the section. As a final step, Horasan mortar was applied on the sections that had been
reinforced with CFRP strings.
BW4-R2 was retrofitted during the construction stage since it was experienced fromtests of BW3-R1 and BW2-RR that the lateral load capacity of the specimen did not
decrease due to removal process of mortar for CFRP retrofitting. Two steel re-bars were
placed at two ends of the wall without anchoring to the foundation and CFRP strings were
tied around these re-bars in order to provide proper confinement and also minimize
possible slip of the CFRP strings from the mortar joint during testing as seen in Figure 2.7.
Therefore, the continuity of FRP strings was maintained.
Figure 2.7. Preparation of BW4-R2.
2.2.3. Test Setup and Instrumentation
Varying pre-compression loads were applied using a servo-controlled verticalactuator with a maximum capacity of 1000 kN, with reaction on the strong floor by means
of vertical pre-stressing cables, which kept the vertical load approximately constant. The
lateral load was applied based on the displacement-control criteria via the horizontal
actuator which had a capacity of 250 kN and connected to the reaction wall. Both thevertical and horizontal loads were transmitted by means of a reinforced concrete beam that
had been designed in a way that it could be easily placed on and removed from the top of
the wall (Figure 2.9). A set of steel rollers were used between the vertical actuator and the
reinforced concrete beam in order to allow relative displacement between the vertical
actuator and the beam (Figure 2.8).
Figure 2.8. Vertical actuator, RC beam and test specimen joint detail.
Figure 2.9. Brick wall and RC beam joint detail.
In any quasi-static cyclic loading, the specimens were subjected to predetermined
numbers of displacement-controlled loading cycles. Under a certain vertical load set, three
cycles of the same amplitude in story drift were repeated and then displacement amplitude
was increased (Figure 2.10). Loading cycles were applied until the specimens reached their
yield strength under a certain vertical load case. Depending on the vertical load,approximately 15-21 reversed cycles were applied throughout the test for a specific vertical
load set. All data were recorded by using data acquisition system. Crack propagation,
rocking mechanism, de-bonding and other failures were also recorded.
Figure 2.10. The displacement based loading protocol used in the tests.
Critical sections where displacement was expected to be observed had been
instrumented by Linear Variable Differential Transformers (LVDTs) as it is seen in Figure
2.11.
Six LVDTs’ were placed at two sides of the specimens in order to measure
deformations due to rocking, One LVDTs was placed to control and measure any relative displacement between
top beam and the specimens,
Two LVDTs were mounted for measuring the top displacement of beam,
Two LVDTs were placed on the beam diagonally for measuring the diagonal
displacements between the beam and the foundation,
Two LVDTs were placed on the specimens diagonally for measuring the diagonal
displacements between the top and bottom parts of the specimens,
One LVDT was placed on the strong floor for measuring any relative displacement
of foundation and the strong floor.
All of the LVDTs were connected to the data acquisition system.
Figure 2.11. Sensor layout.
In order to investigate the contribution of applied vertical load in the lateral load
capacity of the specimens, five different vertical load sets were determined (from 50 kN to250 kN). At each vertical load set, lateral load cycles were applied until yielding was
observed in the lateral load capacity in order to prevent any damage that could occur on the
specimens. Here, yielding is referred as lateral load capacity to stay constant for increasing
target displacement where rocking mechanism is observed.
After the tests were completed for each vertical load set, the test procedure was
followed until the specimens failed due to shear cracking under 300 kN vertical load.
Figure 3.1. Setup and deformations of BW0 at first test.
At the second test, a hydraulic jack was used for inducing vertical load which was
increased from 100 kN to 250 kN in four sets. However, during the cyclic loading tests, it
was observed that the hydraulic jack’s position waschanging at push and pull cycles which
led varying vertical load instead of desired constant load. Under 250 kN vertical load,110.79 kN lateral load capacity was recorded at 1.75 % drift level. Rocking was observed
at all drift levels during the test (Figure 3.2). Furthermore, V-shaped crack which followed
the mortar and brick joints, and crushing of the toe sections of BW0 at the same drift and
lateral load level were observed. The second test was ended at 1.75 % drift level.
Figure 3.2. Deformations on BW0 at the second test.
The third test was aimed to stabilize the vertical load on BW0 by means of spherical
steel rollers which would allow the hydraulic jack to move freely at the displacement
cycles.
The objective of this test was to determine whether the spherical steel rollers would
properly function. Hence, only one vertical load set was applied which was 200 kN. In this
set, 89.90 kN in the push direction was recorded as the highest lateral load capacity at 1.00
% drift level. Rocking mechanism at 0.60% drift level was observed at the sections of
BW0 where the V-shaped cracks had been formed in the second test (Figure 3.3). Besides,
V-shaped cracks became longer whereas shear cracks were started to propagate at the bottom corners of BW0 at 0.75 %. The test was ended at 1.25 % drift level since the lateral
load capacity of BW0 was decreased by 50%. Throughout the test, variation of vertical
load was prevented which figured out that use of steel rollers provided the desired loading
conditions.
Figures 3.3. Setup and deformations on BW0 at the third test.
3.2.2. Specimen BW1-C
As it was mentioned previously, this specimen belonged to control specimen.
Incremental vertical load started with 50 kN and ended with 300 kN. The pre-determined
target displacement cycles were followed until yielding is observed in lateral load vs.
The lateral load versus displacement relationship of specimen BW1-C under 300 kN
vertical load is provided in Figure 3.5. Pinching effect could be clearly observed in the
hysteresis curves. In addition, stiffness degradation was occurred during the repeatedcycles of drift levels. Decrease in strength started after 20 mm lateral displacement.
Figure 3.5. Lateral load versus top displacement for specimen BW1-C.
According to Equation 1.1, maximum rocking load, V r , is found to be 139 kN.
Average compressive strength was taken as 12.9 MPa according to studies of Rad (1978).
In addition, K was taken as 0.85 whereas boundary condition ψ was taken as 1.00 since the
specimen was fixed only at one end. In the tests, it was observed that rocking mechanism
was started at 0.075% drift level under 43.42 kN lateral force action.
Crack patterns and load history are provided in Table A.1. at Section A.1.
As it was indicated in S ection 2.2.3, four diagonal LVDT’s were located forrecording the diagonal di splacements. Readings from these LVDT’s on BW1-C are
provided in Figures 3.6 and 3.7. Measurements obtained from diagonal LVDTs on wall
panel are classified as DG1-2 whereas other two measurements are presented in the name
of DG3-4.
From Figure 3.7, in every push and pull cycles, almost symmetric diagonaldisplacements were recorded. In contrast to Figure 3.7, shear displacement hysteresis
cycles were predominantly remained in the tension zone as indicated in Figure 3.6. This
could be related with lack of reinforcement which would have provided resistance against
the movement. In addition, in the elastic region, displacement values corresponded tomaximum lateral load values tend to be close to y-axis of the graph. This could show that
the specimen exhibited a rigid body motion in this region which resulted in rocking
mechanism. As the lateral force-shear displacement curves shifted away from y-axis,
excessive damage on the specimen was expected.
-30 -20 -10 0 10 20 30-150
-100
-50
0
50
100
150
Shear Displacement [mm]
Lateral Force [kN]
Shear 2
Shear 1
Figure 3.6. Lateral force-shear displacement relationship for BW1-C (DG1-2).
-30 -20 -10 0 10 20 30-150
-100
-50
0
50
100
150
Shear Displacement [mm]
Lateral Force [kN]
Shear 4
Shear 3
Figure 3.7. Lateral force-shear displacement relationship for BW1-C (DG3-4).
From Figure 3.10, symmetric lateral force-shear displacement curves were obtained
which was a result of free movement of the specimen in push and pull cycles. For the pull
cycle case, it was observed that maximum diagonal displacement values tended to be closeto vertical axis whereas they were away from vertical axis in the push cycles. That could
be related with the existing un-repaired deformations on the wall which could alter the
behavior of the specimen in push and pull cycles. Furthermore, distance of the maximum
values from vertical axis is relatively larger than the ones in BW1-C. That could be related
with the excessive deformation occurred on this specimen.
-30 -20 -10 0 10 20 30-150
-100
-50
0
50
100
150
Shear Displacement [mm]
Lateral Force [kN]
Shear 2Shear 1
Figure 3.10. Lateral force-shear displacement relationship for BW2-RR (DG1-2).
-30 -20 -10 0 10 20 30-150
-100
-50
0
50
100
150
Shear Displacement [mm]
Lateral Force [kN]
Shear 4
Shear 3
Figure 3.11. Lateral force-shear displacement relationship for BW2-RR (DG3-4).
Crack patterns and load history are provided in Table A.5 at Section A.2.
The variation of maximum lateral load and the target displacement cycle for each
incremental vertical load case are given in Table 3.4.
Table 3.4. Max. lateral load and drift levels of under incremental vertical load sets
for BW4-R2.
VerticalLoad [kN]
Max. LateralLoad [kN]
Max. TopDisplacement
[mm] Drift Level [%]
50 25.90 4.5122 0.25
100 42.62 5.0979 0.25
150 48.68 4.1052 0.20
200 68.78 5.2647 0.25
250 72.64 5.5787 0.30
The first horizontal cracks due to rocking mechanism were observed at 0.30% drift
level under 50 kN vertical load set. Similarly, first crushing cracks at toe sections wereexamined again in this vertical set at 0.075% drift level. However, this crack was
developed due to stress concentration around the screw that was attached to the specimen
for holding at that section. Therefore, the type of this crack was differed from other crush
cracks that had been observed in previous specimens.
In 300 kN vertical load set, maximum lateral load capacity was reached at 0.60%
drift level which was 90.16 kN at push. First local diagonal hair line cracks were
developed at 0.15% drift level. Moreover, diagonal shear cracks at upper-sides of BW4-R2
were examined at 0.60% drift level. Furthermore, rupture of CFRP strings at the bottom
line was observed at that drift level which was followed by a sudden decrease in lateral
load capacity. Crushing at toe sections on the right from front of BW4-R2 was observed
after the rupture of CFRP strings. In addition, out-of-plane deformation was recorded
towards to the surface where the ruptured CFRP strings were located.
The test was ended at 0.75% drift level due to the decrease in the lateral load
The lateral load versus displacement relationship of specimen BW4-R2 under 300
kN vertical load is provided in Figure 3.18. Compared to BW1-C and BW3-R1 specimens,
the load deformation hysteresis cycles became fat and specimen exhibited a rigid behavior.Stiffness and strength degradation could be clearly observed at 13.93 mm that corresponds
to rupture of CFRP strings.
In contrast to previous specimens, BW4-R2 remained its integrity in terms of
deformation. But, the maximum lateral load capacity in this specimen couldn’t overwhelmthe others. That could be explained by the difference in batches which specimens had been
prepared (BW0, BW1-C, BW3-R1 were constructed in the first batch and BW4-R2 prepared in the second batch).
-30 -20 -10 0 10 20 30-150
-100
-50
0
50
100
150
Shear Displacement [mm]
Lateral Force [kN]
Shear 2
Shear 1
Figure 3.19. Lateral force-shear displacement relationship for BW4-R2 (DG1-2).
-30 -20 -10 0 10 20 30-150
-100
-50
0
50
100
150
Shear Displacement [mm]
Lateral Force [kN]
Shear 4
Shear 3
Figure 3.20. Lateral force-shear displacement relationship for BW4-R2 (DG3-4).
In Figures 3.19 and 3.20, it was observed that shear deformations were recorded
mostly in compression zone for push and pull cycles in contrast to control specimen. This
could be explained by the contribution of CFRP strings in restricting the movementtowards tension zone. In addition, it could be stated that restricted movement at tension
zone resulted in diagonal compression.
Crack patterns and load history are provided in Table A.13 at Section A.4.
3.3. Analysis of Test Results
In this section, analysis of test results will be provided. Normalized lateral load
versus drift level envelopes, lateral force-shear deformation, and moment-base rotation
relationships of the wall specimens will be presented. In addition, lateral load versus drift
level backbone curve, energy dissipation and change in rigidity graphs for all specimens
will be compared.
3.3.1. Normalized Lateral Load versus Drift Level Relationship
Normalized lateral load versus drift level graphs are provided for each specimen in
Figures 3.21 to 3.24. Hysteresis loops that stayed under 80% of the maximum lateral load
value at both push and pull cycle regions were neglected.
From Figure 3.21, it could be observed that yielding of specimen BW1-C started
after 0.25% drift level which corresponded to 85% of the maximum lateral load capacity.
Yielding plateau continued up to 1.00% drift level where lateral load capacity started to
decrease. At 1.25% drift level, lateral load capacity decreased up to 80% of the maximum
value. Furthermore, slope of the curves drastically decreases after 0.50% drift level which
Figure 3.21. Normalized lateral load vs. drift level for BW1-C.
Figure 3.22. Normalized lateral load vs. drift level for BW2-RR.
In BW2-RR, increase in lateral load capacity continued linearly up to 0.60% drift
level where yielding started to be observed after this point (Figure 3.22). Maximum lateral
load capacity was reached at 0.75% drift level. After this drift level, load capacity startedto decrease. The behavior of BW2-RR could be defined as rigid by taking small change in
the slope of the load – drift level curves into account. In other terms, loss in rigidity was
controlled in BW2-RR.
From Figure 3.23, yielding of BW3-R1 could be seen at 0.30% drift level where the
slope of load-drift level curves started to decrease slightly. In addition, 90% of maximum
load capacity was reached at that drift level. At 0.50% drift level, maximum load capacity
was reached. Prior to that level; de-bonding of CFRP strings had been recorded. After that
level, decrease in slope of the load-drift level curves was increased. At 1.00% drift level,
Six LVDT’s were instrumented on both sides of the specimens for measuringdisplacements due to rocking mechanism. These LVDT’s could be grouped into threelevels which each level could refer to three distinct plane of base rotation due to rocking
mechanism. Rotation calculation for each rotation level can be done using Equation 3.1
Figure 3.27. Vertical displacement readings and base rotation measurement.
Sub-indices i in Equation 3.1 refers to the number of the level of base rotation
described previously. Δ is the displacement measured from corresponding LVDT wheresub-indices R and L are representing the position of LV DT’s (right or left). Similarly,h is
the initial length of corresponding LVDT located on the right or left side of the specimens.
ε is the calculated strain value for each base rotation level. Moment was calculated with
respect to the toe section of the specimens. The moment-base rotation relationships for all
specimens are presented in Figures 3.28 to 3.39.
Figure 3.28. Moment-base rotation relationship for BW1-C at first level.
From Figures 3.31 to 3.33, rotations in the limits of -0.3 x 10 -4 and 0.0 radians were
obtained in BW2-RR. In addition, compared to other two levels, rotations, and hence
deformations, in the third level were relatively higher, which pointed out rockingmechanism at that level. In push cycles, base rotations of -0.05 x 10 -4, -0.06 x 10 -4, and -
0.13 x 10 -4 radians were obtained in first, second, and third levels respectively at maximum
moment. For the pull cycle case, rotations were calculated as -0.06 x 10 -4, -0.07 x 10 -4, and
-0.13 x 10 -4 radians in first, second, and third level respectively.
Figure 3.34. Moment-base rotation relationship for BW3-R1 at first level.
Figure 3.35. Moment-base rotation relationship for BW3-R1 at second level.
As it was indicated in section 2.2.3., four diagonal LVDT’s were located forrecording diagonal displacements. The readings from these LVDT’s could be used incalculations for shear deformations by using Equation 3.2 with reference to Figure 3.40.
D2
D1
Y
Y
θ
Defromed Shape
Undefromed Shape
Figure 3.40. Shear deformation measurement.
2 2 2 21 2 = tan'
2
D Y D Y
Y
(3.2)
D1 and D2 refers to the deformed length of diagonal LVDT’s.
According to Equation 3.2, two types of deformation angles for each specimen were
calculated: Shear deformation angle of wall panel that obtained from readings of diagonal
LVDTs on the specimen (DG1-2), and deformation which was calculated by using
readings from diagonal LVDTs on reinforced concrete beam and foundation (DG3-4).
Lateral force versus shear deformation relationships are presented in Figures 3.41 to 3.48.
From Figure 3.41, it could be seen that, -2.00 x 10 -3 and 2.25 x 10 -3 radians of shear
deformations in push and pull cycles respectively was observed at 110 kN lateral force
level. This level of lateral force was obtained around 0.25% drift level (Figure 3.5). After
In general, it could be summarized that maximum shear deformation of wall panels
stayed within the range of -0.01 and 0.01 radians which could be classified as small shear
deformations. Comparison of normalized lateral force-shear deformation relationship backbone curves is provided in Figure 3.49. From Figure 3.49, in push cycles, at 80% of
the maximum lateral force level, BW1-C and BW4-R2 have almost zero shear deformation
whereas -1.5 x 10 -3 radians of shear deformation was reached for BW3-R1 at the same
level. Especially at the push direction and at the pull direction in general, compared to
control specimen, lower shear deformations were observed in retrofitted specimens for the
same lateral force level.
3.3.5. Rigidity – Drift Level Relationship
Rigidity of the specimens could be calculated from the slope of lateral force versus
displacement and drift level curves in the elastic region which are presented in Section 3.2.
A representative sketch for definition of rigidity is given in Figure 3.50. Rigidity
calculations were performed until the lateral load capacity of the specimens decreased
down to 80% of the maximum value.
Figure 3.50. Stiffness and energy calculations.
Change in rigidity with increasing drift level for all specimens is presented from
3.3.6. Energy Dissipation – Drift Level Relationship
Cumulative dissipated energy could be defined as the total area under the hysteresis
loops in lateral force and displacement response as it is illustrated in Figure 3.50.
Accordingly, the hysteresis loops in lateral force and displacement response are discretized
and the area under each loop is calculated. From Figure 3.56 to Figure 3.59, cumulative
energy dissipation versus displacement and drift level response is provided. In the analysis,
calculations were performed until the lateral load capacity decreases down to 80% of the
maximum recorded value. Therefore, ultimate cumulative energy dissipation levels vary
for all specimens due to varying number of cycles that remained within specified the
limits.
It is observed from Figure 3. 56 that, up to 0.50% drift level, 3,291 kN·mm energywas dissipated. Cumulative energy dissipation reached to 16,000 kN·mm at the final driftlevel.
Figure 3.56. Cumulative energy dissipation-drift level relationship for BW1-C.
From Figure 3.57, a cumulative energy of 2,339 kN·mm up to 0.50% drift level was
recorded. Compared to control specimen, cumulative energy dissipation was decreased by
a factor of 0.72 at this drift level. At the final drift level, 8,41 9 kN·mm cumulative energylevel was reached.
Figure 3.57. Cumulative energy dissipation-drift level relationship for BW2-RR.
From Figure 3.58, up to 0.50% drift level, 4,455 kN·mm cumulative energydissipation was reached. Compared to control specimen, cumulative energy dissipation was
increased by a factor of 1.36 at this drift level. At the final drift level, 12,462 kN·mmenergy was recorded.
Figure 3.58. Cumulative energy dissipation-drift level relationship for BW3-R1.
For specimen BW4-R2, it c ould be stated that, 5,498 kN·mm cumulative energy wasdissipated up to 0.50% drift level (Figure 3.59). Compared to control specimen, cumulative
energy dissipation was increased by a factor of 1.68 at this drift level. However, ultimate
cumulative dissipa ted energy was recorded as 7,206 kN·mm at the final drift.
Normalized energy was simply evaluated by taking the ratio between the area of each
force-displacement curve and the area of the rectangle that enclosed the maximum
boundaries of the corresponding cycle (Figure 3.50). In addition, same ratio for cumulativeenergy dissipation was also evaluated. Accordingly, loop-wise and cumulative normalized
energy dissipation ratio versus displacement and drift level relationship of all specimens is
compared.
Figure 3.61. Loop-wise normalized energy dissipation ratio vs. drift level
relationship.
Figure 3.62. Cumulative normalized energy dissipation ratio vs. drift level
relationship.
From Figures 3.61 and 3.62, it was observed that retrofitted specimens BW3-R1 andBW4-R2 performed better in terms of energy dissipation at each loading cycle.
This study investigated use of CFRP strings in the joints of unreinforced masonry
brick walls in historic structures as a near surface retrofitting technique which would
preserve the historic and architectural features of the structures. An experimental study was
carried out on four nearly real-scaled unreinforced masonry brick walls subjected to cyclic
lateral loads at Structural Engineering Laboratory of Boğazici University. The effect of proposed strengthening technique on seismic behavior and lateral load capacity of the
specimens was evaluated. In addition, effect of vertical load on lateral load capacity was
examined.
4.2. Conclusions
The main objectives of the research were to improve seismic behavior of
unreinforced masonry brick walls while preservation of historic and aesthetic features of
the structures are maintained. The following conclusions could be derived from the results
and analysis of the experimental study.
Pinching effect in the lateral force versus top displacement relationship was
eliminated in the last retrofitted specimen BW4-R2, which provided larger
dissipation of energy under cyclic loading.
Cumulative energy dissipation was increased in the retrofitted specimens compared
to control specimen by a factor of 1.68, and 1.36 for BW4-R2 and BW3-R1,
respectively, at 0.50 drift level. This factor was found to be 0.72 for the repaired
specimen BW2-RR. However, loop-wise and cumulative normalized energy
dissipation curves showed that, BW2-RR performed better in dissipating the
encountered energy compared to BW1-C.
Compared to control specimen, crack propagation and deformation resistance were
enhanced in the last retrofitted specimen. Diagonal displacements were restrained
which compelled compressive reaction at the retrofitted specimen. In other terms,
confinement effect was provided by CFRP strings. Rupture of CFRP strings proved
that excessive tension was encountered by the proposed technique. In addition,crushing at toe sections were controlled due to the confinement effect of CFRP
strings.
Rotations were minimized by the proposed technique which also means rocking
mechanism was also controlled.
No significant difference in rigidity of retrofitted specimens was observed compared
to control specimen. Due to poor repairing workmanship, BW2-RR had the lowest
rigidity value at each drift level compared to other specimens. Due to the aspect ratio, failure modes of rocking, diagonal cracking (shear failure)
and toe crushing were simultaneously observed in all specimens except the last
retrofitted specimen, BW4-R2. In addition, rocking mechanism prevailed among the
other failure mechanisms in control and first retrofitted specimens.
Effect of vertical load on the lateral load capacity was verified by tests carried out
under incremental vertical load sets. It was proved that, increase in vertical load up to
65% of its ultimate, increased the lateral load capacity of all specimens. According to Eq (1.1), maximum rocking strength of the specimens was estimated as
139.20 kN. In all specimens, maximum lateral load capacity stayed below this value.
Therefore, by taking the deformation patterns and estimated rocking force into
account, it could be stated that failure mode of the specimens were not pure rocking.
No significant increase in lateral load capacity due to proposed technique was
obtained.
Due to de-bonding of CFRP strings in BW3-R1, a solid comparison with controlspecimen couldn’t be done in terms of lateral load capacity and rotation.
As it was aimed, application of the proposed technique didn’t damage the existingwall specimens in aesthetical and functionality point of view.
4.3. Recommendations
In this study, single orientation of CFRP strings was examined in one typology and
one aspect ratio of brick walls. Further experimental research with different
Abrams, D. P., and J. M. Lynch, 2001, “Flexural Behavior Of Retrofitted Masonry Piers .” Proceedings Of KEERC-MAE Joint Seminar On Risk Mitigation For Regions Of
Moderate Seismicity , University of Illinois, August 5-8, Urbana III. Mid-America
Earthquake Center.
Abrams, D. P., 1992, “Strength And Behaviour Of Unreinforced Masonry Elements ”, 10th
World Conference on Earthquake Engineering , Madrid, July 19-24, Taylor &Francis, Spain.
Al-Manaseer, A., and V. Neis, 1987, “Load Tests On Post-Tensioned Masonry Wall
Bati, S. B., L. Rovero, and U. Tonietti, 2007, "Strengthening Masonry Arches With
Composite Materials." Journal Of Composites For Construction , Vol. 11, No. 1, pp.
33-41.
Binda, L., G. M. Roberti, C. Tiraboschi, and S. Abbaneo, 1994, “Measuring MasonryMaterial Properties,”US-Italian Workshop On Guidelines For Seismic Evaluation
and Rehabilitation Of Unreinforced Masonry Buildings , NCEER Technical Report,
No.21, State University of New York, Buffalo, July.
Bosiljkov V, A. Page, V. Bokan-Bosiljkov, R. Zarnic, 2003, “Performance Based StudiesOn In-Plane Loaded Unreinforced Masonry Walls ”, Masonry International , Vol. 16,
No. 2, pp. 39 – 50.
Bosiljkov, V., A. W. Page, V. Bokan-Bosiljkov, and R. Žarnić, 2010, “Evaluation Of the
Seismic Performance Of Brick Masonry Walls ”. Structural Control And Health
Monitoring , Vol. 17, No. 1, pp. 100 – 118.
Calvi G. M., G. R. Kingsley, and G. Magenes, 1996, “Testing Of Masonry Structures For