Shear-Compressive 2ecess

You might also like

You are on page 1of 14

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

net/publication/274719517

SHEAR-COMPRESSIVE EXPERIMENTAL BEHAVIOUR OF ONE-LEAF STONE


MASONRY WALLS IN NORTH OF PORTUGAL

Conference Paper · August 2014

CITATIONS READS

6 171

4 authors:

Celeste Almeida João Miranda Guedes


University of Porto University of Porto
19 PUBLICATIONS   85 CITATIONS    90 PUBLICATIONS   348 CITATIONS   

SEE PROFILE SEE PROFILE

António José Coelho Dias Arêde A. Costa


University of Porto University of Aveiro
293 PUBLICATIONS   1,410 CITATIONS    447 PUBLICATIONS   2,565 CITATIONS   

SEE PROFILE SEE PROFILE

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

Flatjack in-situ tests View project

Architectural heritage Portugal Brazil View project

All content following this page was uploaded by Celeste Almeida on 16 April 2015.

The user has requested enhancement of the downloaded file.


SHEAR-COMPRESSIVE EXPERIMENTAL BEHAVIOUR OF ONE-
LEAF STONE MASONRY WALLS IN NORTH OF PORTUGAL
Celeste ALMEIDA1, António ARÊDE2, João Paulo GUEDES 3, Aníbal COSTA4

ABSTRACT

The present work gives a better insight on the in-plane shear behaviour of one-leaf stone
masonry walls made of large stone blocks, a structural element that is present in many old buildings of
the Northern part of Portugal, including Porto historic city centre. Information on this type of walls is
still very scarce, in particular in the context of the seismic assessment of existing buildings, fully
justifying this investigation. For this purpose, twelve full-scale one leaf stone masonry walls were built
with regular and irregular stone patterns to represent walls commonly found on typical Porto old
buildings. The walls were tested in the Laboratory of Earthquake and Structural Engineering (LESE)
of the Faculty of Engineering of the University of Porto (FEUP) under cyclic horizontal in-plane loads
for different levels of vertical stress. The experimental results allowed evaluating the lateral resistance,
the ductility, the energy dissipation capacity, the equivalent damping coefficient, the drift and the
predominant failure modes of the walls, showing that, for the same level of vertical stress, the
response is quite influenced by the arrangement of the stones. The results provided a first and
important overview on the shear-compression behaviour of this type of one-leaf stone masonry walls.

INTRODUCTION

The characterization of the quality of stone masonry must pass a thorough investigation of its
constituents and of the construction techniques that were used. Identify and catalogue masonry walls
in terms of geometry, material constitution and expected mechanical behaviour is still a big challenge.
Stone masonry walls are present in a significant number of old buildings in many European
countries; most of these buildings are located in the heart of villages and towns. The rehabilitation of
old stone masonry buildings is a topic of great concern, given the growing interest in the rehabilitation
of city centres. Still now, and in spite of the increased knowledge in this area, stone masonry is a
complex composite structural material, for which the definition of realistic behaviour laws remains a
challenge. The majority of these structures have a reasonable behaviour under compression loads but
do not have enough resistance to tensile stress. During an earthquake, the walls are subjected to
vertical and horizontal forces and different types of damages can occur (diagonal cracks on surfaces,
out-of-plane bending of the walls, desegregation). In order to predict the seismic response of these
walls is essential the assessment of strength capacity and deformation of existing structural elements,
even in region of low to moderate seismic hazard.
The evaluation of the mechanical behaviour of a masonry wall must always include, as first
step, experimental tests set for different possible layouts, or loading conditions, performed either in-

1
Researcher, University of Porto – Faculty of Engineering, Porto-Portugal, celeste.almeida@fe.up.pt
2
Associate Professor, University of Porto – Faculty of Engineering, Porto-Portugal, aarede@fe.up.pt
3
Assistant Professor, University of Porto – Faculty of Engineering, Porto-Portugal, jguedes@fe.up.pt
4
Full Professor, U. Aveiro, Aveiro-Portugal, agc@ua.pt

1
situ, or in laboratory. In particular, performing shear-compression tests on prototypes constructed in
the laboratory is a common practice when it comes to evaluate the structural behaviour of stone
masonry panels under seismic type actions (Magenes and Calvi, 1997; Vaconcelos et al 2009; Costa et
al, 2010; Silva et al, 2014). However, it is not easy to build wall samples in laboratory to represent
historical stone masonry. Studies carried out in Portugal and Italy allowed also evaluating, through
experimental in-situ testing, the mechanical properties of multi-leaf masonry walls, some with
different strengthening conditions (Chiostrini, et al, 2003; Costa et al, 2006; Corradi et al, 2006;
Calderini et al, 2009).
The buildings constructed until the beginning of the 20th century at the centre of Porto are
mostly made of stone masonry walls that support timber floors and roof trusses. The main walls are
usually made of one leaf of large granite blocks and are 0.30 to 0.50 m thick, Figure 1. Not many
studies have been taken on walls with such characteristics.

(a) (b) (c)


Figure 1. Old buildings of the city of Porto: (a) perspective view; (b) texture of a typical stone masonry wall and
(c) cross-sections of two single leaf walls.

Recently, a one-leaf granite masonry wall was retrieved from a building located at the centre of
Porto to be studied. The wall was cut into panels and transported to LESE to be submitted to an
experimental campaign that included four compression tests (original state and after mortar injection)
and two shear-compression tests (Almeida et al, 2011; Almeida et al, 2012), Figure 2.

(a) (b) (c) (d)


Figure 2. Building subject rehabilitation works: (a), (b) cutting and extraction panels for laboratory tests, (c)
setup of compression tests and (d) setup of shear compression tests.

This first experimental study using specimens from a real construction allowed assessing the
main geometrical and mechanical characteristics of these walls. In particular, it allowed concluding
about the low walls’ strength, around 3 N/mm2, and the even lower stiffness, which gives compression
stiffness to strength ratios of about 100, thus much lower than the values referred in the general
bibliography. Moreover, the improvement of the walls’ stiffness and strength capacity after mortar
injection allowed concluding that the large amount of voids found inside the joints was partially
responsible for the low strength of these walls.
The results of shear tests showed that these walls have some ductility (around 2.5) and capacity
for dissipating energy. No significant degradation of load was observed and the more important
damage areas were located at the walls mid-height, being the failure mode controlled by the formation
of two “rigid blocks” separated by a diagonal cracking essentially following the joints. In particular,

2
C.Almeida, A. Arêde, J. Guedes and A.Costa 3

the understanding of this behaviour is fundamental to assess the performance of stone masonry walls
under seismic type actions. Nevertheless, the available information on this type of walls is still very
scarce, fully justifying further research involving campaigns of cyclic in-plane tests on one leaf stone
masonry walls.
The following sections present the results of an experimental campaign on stone masonry panels
constructed in laboratory and representing common bearing walls of the old buildings of Porto. The
prototypes were built following the in-situ textures and construction procedures to represent, as close
as possible, the actual conditions of real walls. The analysis focuses, in particular, the failure
mechanisms and evaluates the seismic performance of the walls, especially through the quantification
of the mechanical parameters: ductility, energy dissipation, equivalent damping and drift.

DESCRIPTION OF PROTOTYPES AND EXPERIMENTAL PROGRAM

The present experimental programme aims assessing new and valuable data on the shear
behaviour of one-leaf granite stone masonry walls, a structural element that is present on many of the
old buildings of the city of Porto. In addition, the programme will also analyse the influence of the
constructive process on the mechanical performance of this type of masonry. Therefore, the layout of
the samples, i.e. the wall panels, was based on the observation of real cases, focusing on single leaf
walls about 0.28 m thick. The textures replicate the predominant geometric characteristics found on
the stone masonry walls of the old buildings of the city of Porto. Four large specimens, 7.20 m wide,
1.80 m high and 0.28 m thick, representing four different textures, namely Regular (R), Partially
Regular (PR), IRregular (IR) and very IRegular (IR+), were constructed and each one was
subsequently cut into six individualized panels 1.20 m wide, approximately. The assembly sequence
followed the constructive practice applied to this type of walls and the construction was performed by
experienced stonemasons. Figure 3 shows a schematic elevation view of each wall typology and the
line (red) that limits the panels. This paper presents the results of the shear-compression tests
performed on three panels of each typology, referred to the numbers 4, 5 and 6.

R1 R2 R3 R4 R5 R6 PR1 PR2 PR3 PR4 PR5 PR6

1.80 1.80

7.20 7.20

(a) (b)
IR1 IR2 IR3 IR4 IR5 IR6 IR1++ IR2++ IR3++ IR4++ IR5++ IR6++

1.80 1.80

7.20 7.20

(c) (d)
Figure 3. Schematic representation of the arrangement of stones on the four typologies designed in the
laboratory: (a) Regular wall (R), (b) Partially regular wall (PR), (c) Irregular wall (IR), (d) Very irregular wall
(IR +).

The materials used for the specimens, namely the mortar and the stones, were chosen to fit the
in-situ existing conditions. The yellow granite was taken from the North of Portugal and exhibited an
average compressive strength, measured on cylindrical cores (ø100 mm and 200 mm high), of about
55 N/mm2. The lime mortar is composed by a binder of air hardening lime and granitic sand with a
ratio of 1:3. The average compressive strength of the mortar, measured on 40x40x160 mm3 prisms
after 28 days curing, was equal to 1.58 N/mm2.
The constructive process used in the regular typology differed from that used in the remaining
walls. In the typology R, the stones were laid on a layer of mortar about 1.5cm thick, with complete
filling of the horizontal and vertical joints. Wood wedges were placed to avoid crushing of the mortar
after positioning the next layer, but were removed before carrying out the laboratory tests. Figure 4
shows the constructive phasing of the regular wall.

Figure 4. Construction phasing of the regular wall specimen (R).

In the typologies PR, IR and IR++ the stones were laid in layers using stone wedges to ensure
the stability of the stones, a common in-situ procedure on the construction of this type of walls, see
Figure 5. The mortar was placed at the end to close the vertical and horizontal joints, a procedure that
yields voids in the joints. Subsequently, the panels R, PR, IR and IR++ were separated to observe the
cross section and make the survey of the percentage of materials (stone, mortar and voids) before the
experimental test. It was found that the R panels had joints completely filled in, while the remaining
panels had voids in the cross section, as it is observed in-situ in real walls, Almeida et al (2011).

(a)

(b)

(c)
Figure 5. Elevations of the walls before and after placing mortar: (a) Partially Regular (PR), (b) Irregular
(IR) and (c) Very irregular (IR++).

The shear-compression tests were performed using the steel reaction frame of LESE and leaving
the specimen free to rotate at the top. The apparatus consisted on two hydraulic jacks: one to impose
the vertical load and the other to apply the horizontal displacements to the samples (see Figure 6a). The
horizontal actuator, with 200 kN capacity, was attached to a stiff steel reaction structure at one
extremity and to the top of the samples at the other. The vertical load was applied using an actuator
with 500 kN capacity connected to the top of the samples, and to a transversal steel beam fixed with
two Dywidag rods to the base of the steel frame, creating a self-equilibrated system: the compression
forces are imposed through the axial tension applied to the two rods that include load cells to record
and control the vertical forces applied to the samples during the test (see Figure 6b). The specimen was
blocked against out-of-plane motions through spherical hinges that can slide along two lateral beams
(see Figure 6c). Both actuators were linked to the samples through a steel beam that was casted at the
top of the samples to allow a more uniform distribution of the vertical and horizontal forces applied to.

4
C.Almeida, A. Arêde, J. Guedes and A.Costa 5

(b)

(a) (c)
Figure 6. Laboratory apparatus at LESE: (a) general setup; (b) system of vertical load an (c) out-of-plane
protective system.

In order to record vertical, horizontal and out-of-plane displacements, the specimens were
monitored with a complex and dense instrumentation system that included about 30 LVDTs, as
illustrated in Figure 7. Between samples, the instrumentation has undergone changes in the location
and number of sensors due to the variations on the size and arrangement of the stones. The samples
exhibit height to width ratios of about 1.5.
35 36 SUL SUL
NORTE NORTE SUL NORTE
5 34 60
26 18
LFio10 6 LFio8

21 20

LFio11
LFio15
LFio7 28 29 LFio6 31
33
32

LFio11 LFio15
16 15

LFio3 LFio4
23
22

24
37 LFio1 LFio2 39
12

42 43

Figure 7. Instrumentation scheme of the shear-compression tests.

The three panels selected from each typology were tested with different constant vertical forces;
thus, four panels were tested with a vertical stress of 0.4 N/mm2 (a value based on the estimated
vertical load at the basement of a 5 floors building, a situation that is found in many buildings at Porto
centre), four with 0.8 N/mm2 and the final four panels with 1.2 N/mm2, to evaluate the structural
response with much higher levels of compression load. The vertical force was applied before any
horizontal load, and this allowed evaluating the compression behaviour of the samples. Afterwards,
the vertical force was kept constant and the horizontal actuator applied to the samples increasing
cyclic top displacements (three cycles for each displacement level) until failure.

EXPERIMENTAL RESULTS

The present experimental campaign of shear-compression tests aim evaluating the behaviour of
a set of stone masonry walls under horizontal earthquake type actions, namely assessing the in-plane
lateral resistance, displacement and energy dissipation capacity and predominant failure modes. The
campaign involved testing twelve wall panels (three for each geometric typology) and considering
three levels of vertical compression force. The next sections analyse these aspects in detail.
Failure mechanisms

Masonry walls subjected to horizontal in-plane forces may present three different failure modes,
which can appear alone or combined: shear sliding, diagonal shear cracking and flexural. The
occurrence of one mode over another depends on several parameters: the panel geometry (texture,
cross-section and height to width ratio), the boundary conditions, the vertical load and the mechanical
characteristics of their constituents (mortar, blocks and joints), Tomaževič (1999). In the case of the
twelve panels, the behaviour was particularly dependent on the texture and vertical load; in general,
the first openings appeared in the bed joint mortar at the base of the panels. The R samples exhibited
predominant flexural behaviour, accompanied by sliding of stones. Moreover, higher pre-compression
forces implied less damage. The PR and IR samples are more sensitive to the arrangements of the
stones; damage occurred with the formation of approximately diagonal cracks and the test ended with
a rocking mechanism. The IR++ samples also responded with the formation of diagonal cracks, but
also with the crushing of stones located at the lower corners, especially for higher levels of vertical
stress. Figure 8 shows the final state of four specimens (one per typology) under a compressive stress
of 0.4N/mm2.

(a) (b) (c) (d)


Figure 8. Final state of the panels under shear-compression load (pre-compression of 0.4 N/mm2): (a) R4; (b)
PR4; (c) IR4 and (d) IR4++.

Lateral resistance and displacement capacity

The shear compression tests are typically described by hysteretic horizontal force-displacement
diagrams, accompanied by the evaluation of mechanical parameters. The hysterical behaviour of a
masonry wall under cyclic loading can be represented by an idealized envelope and, to do so, different
procedures are reported in the literature (Magenes and Calvi, 1997; Tomaževič, 1999). In this study, a
four branches curve was adopted, defining four different limit states, namely: the first limit state that
occurs until the opening of the first flexural crack (Hf, df); the second limit state that corresponds to the
stage up to the opening of the first significant crack (Hcr, dcr); the third limit state that goes from the
last state up to the point of the maximum horizontal load (Hmax, dHmax), and the final limit state that
ends at the point with the maximum horizontal displacement (Hdmax, dmax). The four transition points
between stages had to be identified during the tests.
In this study, the overall behaviour was analysed by comparing the envelopes of the hysteretic
loops of the response of each sample. Figure 9 shows the horizontal force vs horizontal displacement
response curves for the different typologies and 0.4 N/mm2 of top compression stress (the whole series
of hysteric curves are found in literature, Almeida (2013). Figure 10 shows the idealised envelope
curves of all tested panels and Figure 11 summarizes the values of transition horizontal forces and
horizontal displacement obtained for the four limit states of the different specimens.

6
C.Almeida, A. Arêde, J. Guedes and A.Costa 7

140.0 140.0
+ - + -
R4 PR4
100.0 100.0

60.0 60.0

H (kN)
20.0 20.0
H (kN)

-20.0 -20.0

-60.0 -60.0

-100.0 -100.0
s0=0.4 N/mm2
s0=0.4 N/mm2
-140.0 -140.0
-40.0 -30.0 -20.0 -10.0 0.0 10.0 20.0 30.0 40.0 -40.0 -30.0 -20.0 -10.0 0.0 10.0 20.0 30.0 40.0
d (mm) d (mm)

(a) (b)
140.0 140.0
+ - + -
IR4 IR4++
100.0 100.0

60.0 60.0

20.0 20.0
H (kN)

H (kN)
-20.0 -20.0

-60.0 -60.0

-100.0 -100.0
s0=0.4 N/mm2 s0=0.4 N/mm2
-140.0 -140.0
-40.0 -30.0 -20.0 -10.0 0.0 10.0 20.0 30.0 40.0 -40.0 -30.0 -20.0 -10.0 0.0 10.0 20.0 30.0 40.0

d (mm) d (mm)

(c) (d)
Figure 9. Horizontal force vs horizontal displacement curves for 0.4 N/mm2 compression force: (a) R4, (b) PR4,
(c) IR4 and (d) IR4++.

140.0 140.0
s0=0.4 N/mm2 s0=0.8 N/mm2
120.0 120.0

100.0 100.0
H (kN)

80.0 80.0
H (kN)

60.0 60.0

40.0 40.0
R4 R5
PR4 PR5
20.0 20.0
IR4 IR5
IR4++ IR5++
0.0 0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0
d (mm) d (mm)

(a) (b)
140.0
s0=1.2 N/mm2
120.0

100.0

80.0
H (kN)

60.0

40.0
R6
PR6
20.0
IR6
IR6++
0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0
d (mm)

(c)
Figure 10. Four limit state envelope curves for: (a) 0.4 N/mm2, (b) 0.8 N/mm2 and (c) 1.2 N/mm2.
140.0 40.0
Hf s0=1.2 N/mm2 df s0=1.2 N/mm2
Hcr dcr
120.0 35.0
Hmáx dmáx
s0 =0.8 N/mm2
Hdmáx d (Hmáx)
30.0
100.0
s0 =0.8 N/mm2
25.0
80.0 s0=0.4 N/mm2

d (mm)
H(kN)

s0=0.4 N/mm2 20.0


60.0
15.0

40.0
10.0

20.0 5.0

0.0 0.0
R4 PR4 IR4 IR4++ R5 PR5 IR5 IR5++ R6 PR6 IR6 IR6++ R4 PR4 IR4 IR4++ R5 PR5 IR5 IR5++ R6 PR6 IR6 IR6++
Hf 36.05 20.95 20.38 17.31 42.99 47.78 28.62 28.61 54.71 57.76 34.15 50.80 df 1.51 1.49 1.40 1.54 1.23 2.79 1.47 1.54 1.51 2.39 1.22 1.76
Hcr 47.15 37.13 34.17 29.54 73.31 66.96 53.16 48.04 100.42 101.05 68.07 79.93 dcr 3.21 4.59 4.19 4.42 5.08 9.57 4.38 4.66 9.67 7.85 3.79 6.01
Hmáx 64.81 50.28 44.17 44.68 87.12 87.85 73.03 57.58 125.42 119.66 112.97 92.82 dmáx 18.29 12.84 22.53 21.00 17.86 29.20 21.73 14.98 37.99 30.11 15.96 18.69
Hdmáx 64.81 50.28 42.39 44.68 86.01 87.85 70.54 57.58 124.85 118.32 108.07 92.82 d (Hmáx) 18.29 12.84 16.70 21.00 15.40 29.20 14.15 14.98 34.08 21.36 13.43 18.69

(a) (b)
Figure 11. (a) Horizontal transition loads for the four limit states and the (b) corresponding displacements for all
the samples.

The analysis of the horizontal load vs horizontal displacement curves shows that the shape of
the diagrams, although variable, does not seem to depend on the level of vertical load. On the other
hand, the maximum horizontal force is directly related to the vertical stress imposed. For all typologies
it was found that the horizontal load capacity increases with the increase of the pre-compression force
Generally, and regardless of the vertical stress, the R typology exhibited the highest load
capacity and the IR++ panels the lowest load capacity (see Figure 11a). Thus, it can be concluded that
the lateral strength capacity increases with the increase of the geometric regularity of the masonry. The
regular contact surfaces between stones of the R panels explain these results. In some panels (R4, PR4,
IR4++, PR5, IR5+, R6 and IR6++) the maximum force recorded (Hmax) corresponded to the force for
the maximum displacement (Hmax=Hdmax). In the remaining panels there was a slight decrease at the
end of the test, generally of around 3%. Note that in some tests it wasn’t possible to conduct the wall
to final failure, in particular to reach the softening phase of the global behaviour curve. Safety reasons
contributed to stop the tests before, in particular the high level of damage observed associated to the
fact that the collapse of these structures is very likely to be sudden and fragile. The relationship
between Hcr/Hmax was similar in all typologies; the crack limit state occurred between 64% to 75% of
Hmax. These values fit the proposal by Tomazevic (Hcr=60 to 80% of Hmax) (Tomaževič, 1999).
The displacement capacity results show, in most cases, a different tendency when compared to
the lateral force capacity (see Figure 11b). In particular, it was found that the samples with larger
displacements were not those that responded with higher horizontal forces. However, in some cases
this result may be a consequence of the fact that some tests stopped before reaching a strength
decrease. Nevertheless, these cases corresponded to tests where the samples exhibited already very
high damage and, therefore, where a sudden collapse was anticipatable not changing much the
displacement capacity.

Evaluation of seismic performance

The evaluation of the seismic performance of masonry walls involves the analysis and
quantification of mechanical parameters, namely the displacement ductility, the stiffness degradation,
the energy dissipation capacity and the equivalent viscous damping. Theses parameters were
computed from the results given in the previous section.
The displacement ductility was computed using the displacement attained for the maximum
lateral load (Hmax=dHmax/dcr) and the drift was calculated using both the displacement for the
maximum lateral load (drifHmax=dHmax/L) and the maximum displacement (drifdmax=dmax/L), where L is
the height of the panels (equal to 1.8 m). Figure 12 presents the graphs with the values of theses
parameters. The variability of the results prevents systemic findings. However, in almost all typologies
there was a trend towards reduction in ductility with the increase of the level of pre-compression (with
some exceptions). Moreover, for the same level of vertical stress, and analysing the effect of the
irregularity, it was found that the change in ductility between typologies doesn’t follow a clear trend.

8
C.Almeida, A. Arêde, J. Guedes and A.Costa 9

In some cases, more irregular samples exhibited higher ductility than samples with more regular
geometry. On the other hand, the values of drift exhibited no tendency to increase or decrease with the
level of vertical stress and (or) of irregularity of the panels (see Figure 12b).

6.0 2.50
Hmax
5.5 2.25 dmax
5.0
2.00
4.5
1.75
4.0

Drift (%)
3.5 1.50
Hmax

3.0 1.25

2.5 1.00
2.0
0.75
1.5
0.50
1.0

0.5 0.25

0.0 0.00
R4 PR4 IR4 IR4++ R5 PR5 IR5 IR5++ R6 PR6 IR6 IR6++ R4 PR4 IR4 IR4++ R5 PR5 IR5 IR5++ R6 PR6 IR6 IR6++
5.70 2.80 3.99 4.75 3.03 3.05 3.23 3.21 3.53 2.72 3.55 3.11 Hmax 1.02 0.71 0.93 1.17 0.86 1.62 0.79 0.83 1.89 1.19 0.75 1.04
dmax 1.02 0.71 1.25 1.17 0.99 1.62 1.21 0.83 2.11 1.67 0.89 1.04

(a) (b)
Figure 12. (a) Ductility for the maximum lateral load and (b) Drift for the maximum lateral load and maximum
displacement.

Figure 13 shows the evolution of the secant stiffness degradation K, divided by K cr (the secant
stiffness for the point of significant cracking, Kcr=Hcr/dcr), with the drift for the different samples and
pre-compression levels. The secant stiffness (K) was determined for each loading cycle as the slope of
the straight line that connects the maximum and minimum lateral strength obtained from the force-
displacement diagrams, Tomaževič (1999).
3.0 5.0
s0=0.4 N/mm2 R4 s0=0.8 N/mm2 R5
PR4 4.5 PR5
2.5 IR4 IR5
4.0
IR4++ IR5++
3.5
2.0
K/Kcr

3.0
K/Kcr

1.5 2.5
2.0
1.0
1.5
1.0
0.5
0.5
0.0 0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
Drift (%) Drift (%)
(a) (b)
7.0
s0=1.2 N/mm2 R6
6.0 PR6
IR6
5.0 IR6++

4.0
K/Kcr

3.0

2.0

1.0

0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Drift (%)
(c)
Figure 13. Evolution of the secant stiffness K/Kcr: (a) R4, PR4, IR4 and IR4++ (s0=0.4 N/mm2); (b) R5, PR5,
IR5 and IR5++ (s0=0.8 N/mm2) and (c) R6, PR6, IR6 and IR6++ (s0=1.2 N/mm2).

In all typologies there was a decrease of stiffness with the increase of drift. The variation of this
parameter was more pronounced for the first drifts (up to around 0.20%), being more stable for higher
drifts. Regardless of the vertical load, the R typology had a higher initial stiffness and a more rapid
degradation. The greater initial stiffness is related to the more regular joints completely filled in with
mortar. On the other hand, the faster degradation of R samples may be explained by the rocking
mechanism that dominated the behaviour of these walls, which gave the samples a higher flexibility.
The results of the other typologies were identical, not showing major differences, most probably
because the samples passed through the same type of failure mechanisms: formation of diagonal
cracks first, followed by rocking movements. It was also observed a tendency for stiffness to decrease
with the increase of the irregularity.
Such as ductility, energy dissipation and equivalent viscous damping allowed to obtain relevant
information about the global behaviour of structures under seismic actions. A dissipative structure can
lead to a reduction of the seismic response and, hence, to the reduction of the ductility demands [1].
The dissipated energy (Ediss) is defined for each loading cycle and is obtained by numerical integration
of the hysteretic loop (inner area of the force-displacement diagrams). Moreover, the energy required
to deform the structure to an imposed displacement, i.e. the energy introduced in the system (Einp), can
be obtained from the sum of the areas of the positive and negative cycles of the force-displacement
diagrams.
Figure 14 shows the energy dissipation of the tested walls. Regardless of the level of the vertical
stress and the geometry of the panels, and with few exceptions, there has been a tendency to reduce the
ratio Ediss/Einp until cracking (Hcr), increasing slightly, or stabilizing, up to the maximum force (H max).
This phenomenon is associated with the progression of the damage after the formation of the main
rupture lines. Regular walls exhibited values of dissipated energy (Ediss/Einp) lower than the other
typologies, especially when compared to the most irregular walls (IR and IR++) for the higher levels
of vertical load. In this case, and considering the values obtained at the end of the test, the differences
between R6 and IR6 were about 4 times, being about 1.5 times for the drift correspondent to the H cr.
The low energy dissipation capacity of the R walls when compared to the other typologies is related to
the rocking mechanism developed.
90.0 90.0
R4 R5
PR4 PR5
75.0 IR4 75.0 IR5
IR4++ IR5++
60.0 60.0
Ediss/Einp (%)
Ediss/Einp (%)

45.0 45.0

30.0 30.0

15.0 15.0

0.0 0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2
Drift (%) Drift (%)
(a) (b)
90.0
R6
PR6
75.0 IR6
IR6++

60.0
Ediss/Einp (%)

45.0

30.0

15.0

0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2
Drift (%)
(c)
Figure 14. Energy dissipation: (a) R4, PR4, IR4 and IR4++ (s0=0.4 N/mm2); (b) R5, PR5, IR5 and IR5++
(s0=0.8 N/mm2) and (c) R6, PR6, IR6 and IR6++ (s0=1.2 N/mm2).

10
C.Almeida, A. Arêde, J. Guedes and A.Costa 11

Analyzing the equivalent viscous damping (), the conclusions are similar to those attained for
the energy dissipation (see Figure 15). There was a tendency of  to increase with the increase of the
irregularity of the panels, especially for the displacements corresponding to the maximum horizontal
capacity and for the final displacement. The IR and IR++ samples showed higher damage and higher
equivalent viscous damping.
Table 1 summarizes the main values of the mechanical parameters of each sample, namely the
drift, the ductility, the energy dissipation capacity and the equivalent viscous damping computed for
the maximum horizontal load drift.
40.0 40.0
R4 R5
35.0 PR4 35.0 PR5
IR4 IR5
30.0 IR4++ 30.0 IR5++

25.0 25.0
 (%)

 (%)
20.0 20.0

15.0 15.0

10.0 10.0

5.0 5.0

0.0 0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2
Drift (%) Drift (%)
(a) (b)
40.0
R6
PR6
35.0
IR6
IR6++
30.0

25.0
 (%)

20.0

15.0

10.0

5.0

0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2
Drift (%)
(c)
Figure 15. Equivalent viscous damping: (a) R4, PR4, IR4 and IR4++ (s0=0.4 N/mm2); (b) R5, PR5, IR5 and
IR5++ (s0=0.8 N/mm2) and (c) R6, PR6, IR6 and IR6++ (s0=1.2 N/mm2).

Table 1. Values of the mechanical parameters of masonry typologies for the maximum horizontal
capacity
Wall typology σ0 (N/mm2) Drift (%)  Ediss/Einp (%)  (%)
R 0.4 1.02 5.70 50.93 14.09
0.8 0.86 3.03 40.96 10.14
1.2 1.89 3.53 34.14 8.02
PR 0.4 0.71 2.80 59.22 15.63
0.8 1.62 3.05 43.06 12.04
1.2 1.19 2.72 31.56 10.85
IR 0.4 0.93 3.99 59.19 14.82
0.8 0.79 3.23 44.56 10.45
1.2 0.75 3.55 66.05 19.52
IR++ 0.4 1.17 4.75 47.71 15.44
0.8 0.83 3.21 64.13 13.79
1.2 1.04 3.11 55.39 13.58
CONCLUSIONS

In order to evaluate the shear behaviour of typical walls of Porto city centre, experimental tests
were performed in laboratory models built at real scale. Twelve one-leaf panels representing four
different textures of stone masonry walls were subjected to in-plane horizontal cyclic forces for three
levels of fixed pre-compression. With reference to the procedures applied in similar studies, several
parameters were computed and analysed from the experimental force-displacement response curves,
namely: load capacity, stiffness, ductility, energy dissipation and equivalent damping coefficients. The
results showed that the behaviour of these walls, in particular the failure mechanisms are clearly
influenced by the arrangement of the stones and the level of vertical stress; in most cases, the tests
ended without loss of strength. In the regular walls the rocking mechanisms were predominant, being
followed by sliding of the stones; in the more irregular walls damage passed through the formation of
approximately diagonal rupture lines leading to rocking mechanisms at the end of the tests. Regular
walls showed the highest load capacity due to greater regularity of the contact surfaces between
stones; the more irregular walls showed higher capacity for energy dissipation and damping,
particularly for higher levels of vertical stress. Generically, and for each typology, the ductility
decreased with the increase of the vertical stress and there was a tendency of the ductility to increase
with the increase of the irregularity of the samples.
The results are a starting point in identifying the behaviour and the performance of the different
textures of one leaf masonry walls when acted by horizontal in-plane forces. There is a need to
conduct more tests by typology, in order to validate the results obtained in this study.

ACKNOWLEDGEMENTS

The authors acknowledge the collaboration of “SRU Porto Vivo” (Society of Urban
Rehabilitation of Porto), the architect Adriana Floret and the construction companies 3M2P, ERI and
Lucios. The efforts of LESE staff, Mr. Valdemar Luis and André Martins, are also gratefully
acknowledged. This work includes research carried out with financial support of the FCT (Portuguese
Foundation for Science and Technology) through the founding of the research unit CEC (Construction
Study Centre of FEUP) and the Ph.D. grant given to the first author (SFRH/DB/45500/2008).

REFERENCES
Almeida C., Guedes J.P., Arêde A., Costa C.Q., Costa A. (2011), “Physical characterization and compression
tests of one leaf stone masonry walls”, Construction and Building Materials, 30: 188-197.
Almeida, C., Guedes, J., Arêde, A. e Costa, A. (2012), “Shear and Compression Experimental Behaviour of One
Leaf Stone Masonry Walls”, 15th World Conference Earthquake Engineering, Lisbon.
Almeida, C. (2013), Paredes de Alvenaria do Porto, Tipificação e Caraterização Experimental. PhD at FEUP,
Porto.
Binda L. and Saisi A., (2002) “State of the art of research on historic structures in Italy state of the art of research
on historic structures in Italy”, Proceedings of 11th Advanced Research Initiation Assisting and
Developing Networks in Europe (ARIADNE ) Workshop, Italy.
Borri A. (2006), Valutazione e Riduzione della Vulnerabilità di Edifici in Muratura. Progetto di ricerca N.1,
Rendicontazione Scientifica 1º anno, Allegato 3b.1_UR06_1, Progetto esecutivo 2005 – 2008, RELUIS,
Italy.
Calderini C., Cattari S., Lagomarsino S. (2009), In-plane strength of unreinforced masonry piers. Earthquake
Engineering and Structural Dynamics, 38 (2): 243-267.
Corradi M., Borri A., Vignoli A. (2003), “Experimental study on the determination of strength of masonry
walls”, Construction and Building Materials, 17 (5): 325-337.
Costa A., Silva B., Costa A., Guedes J., Arêde A. (2006), “Structural behaviour of a masonry wall under cyclic
load: experimental and numerical study”, Proceedings of the V International Conference on Structural
Analysis of Historical Constructions, New Delhi, India.
Costa A. A., Arêde A., Costa A., Oliveira C. (2010), “In situ cyclic tests on existing stone masonry walls and
strengthening solutions”, Earthquake Engineering and Structural Dynamics, 40 (4): 449–471.

12
C.Almeida, A. Arêde, J. Guedes and A.Costa 13

Chiostrini S., Galano L., Vignoli A. (2003), “In situ shear and compression tests in ancient stone masonry walls
of Tuscany, Italy”, Journal of Testing and Evaluation, 31 (4): 289-303.
Magenes G. and Calvi G.M. (1997), “In-plane seismic response of brick masonry walls”, Earthquake
Engineering and Structural Dynamics, 26 (121): 1091- 1112.
Silva B., Benetta M., da Porto F. (2014), “Experimental assessment of in-plane behaviour of three-leaf stone
masonry walls”, Construction and Building Materials, 53: 149-161.
Tomaževič M. (1999), Earthquake-resistant design of masonry buildings, Imperial College Press.
Vasconcelos G., Lourenço P.B. (2009), “Experimental characterization of stone masonry in shear and
compression”, Construction and Building Materials, 23: 3337-3345.

View publication stats

You might also like