Engineering Solid Mechanics: Structural Irregularity: The Analysis of Two Reinforced Concrete (R.C.) Buildings
Engineering Solid Mechanics: Structural Irregularity: The Analysis of Two Reinforced Concrete (R.C.) Buildings
Engineering Solid Mechanics: Structural Irregularity: The Analysis of Two Reinforced Concrete (R.C.) Buildings
Mario Lucio Puppioa*, Linda Giresinia, Fabio Doveria and Mauro Sassub
a
Department of Energy, Systems, Territory and Constructions Engineering, University of Pisa, Italy
b
Department of Civil, Environmental Engineering and Architecture, University of Cagliari, Italy
A R T I C L EI N F O ABSTRACT
Article history: Structural irregularity is a crucial issue in assessing seismic vulnerability of both new and existing
Received 10 October, 2018 buildings. European technical codes provide simple criteria to define irregularities in plan and in
Accepted 10 December 2018 elevation, amplifying the seismic actions and/or introducing torsional effects. Nevertheless, this
Available online
approach only considers geometrical irregularity. For existing buildings, another source of
10 December 2018
Keywords:
irregularity comes from the non-uniform distribution of the material strength. In particular, for
Multi-storey r.c. buildings existing reinforced concrete (r.c.) structures, it is possible to detect significant spread of the
Material variability concrete compressive strength not only from different structural elements but also from different
Diaphragm flexibility parts of the same member. In this work, non-linear static analysis is performed on two case-studies
Irregularity in strength of r.c. buildings characterized by geometrical and mechanical irregularity. The resistance of each
column is determined with an extensive experimental campaign with in situ and laboratory test
(about 600 in situ tests). The results are analyzed considering both local and global effects in terms
of resistance of the single elements and of the entire buildings. In this sense, shear and bending
failure mechanism are taken into account. The effect of storey flexibility is also considered in the
models. Fragility curves are calculated for the buildings with random distribution of the
compressive strength of the columns. The results are then compared with the approaches proposed
by the Eurocodes evaluating in the standard approach proposed by technical codes is conservative
or not.
© 2019 Growing Science Ltd. All rights reserved.
1. Introduction
From the ‘50s, reinforced concrete (r.c.) buildings became more numerous than masonry structures.
Nevertheless, most of r.c. buildings in Italy were built without following specific seismic codes, with
consequent high vulnerability under earthquake type actions (Bonannini et al. 2017). The seismic
response of r.c. building is strictly related to the mechanical behaviour of concrete, depending on its
resistance and stiffness. The standard approach of the technical codes (American Society of Civil
Engineers (ASCE), 2000; CMIT, 2009; EC8-3, 2005; Manfredi & Masi, 2011) indicates to carry out
analysis with the mean values of resistance derived from in situ tests. The dispersion of the strength of
material is taken into account only considering a reduction of the design resistance as for example in
(FEMA, 2006). In addition, technical codes do not take into account the material variability as source
of irregularity. The structural irregularity is only addressed to geometric issues (i.e. geometry and
disposition of structural elements, presence of infill walls, etc.).
* Corresponding author.
E-mail addresses: [email protected] (M. L. Puppio)
The seismic capacity of reinforced concrete buildings is usually investigated by means of non-linear
static analysis (Bosco et al. 2015a,b c; Fajfar, 2000; Fajfar et al. 2005; Maru & Fajfar, 2005). In
literature, several contributions deal with the material variability (Varadharajan et al. 2012) referring
to historical constructions made both on masonry and on r.c. buildings. In particular, De Stefano and
co-workers analysed the material irregularity considering it both from analytical and numerical point
of views ( De Stefano et al., 2014; De Stefano & Pintucchi, 2002, 2010; Humar & Kumar, 1999; Lavan
& de Stefano, 2013). The analytical approach considers the main variables that affect the problem. The
evaluation of the structural response allows to refer to torsionally stiff and torsionally flexible slabs.
Torsional effects are also introduced throughout geometrical eccentricity with the aim of taking into
account mechanical variability with a simplified method (Bosco et al. 2015 a,b; Mittal & Jain, 1995;
Sadek & Tso, 1989). Biondini and co-workers proposed some models that consider aging and decay of
the structural materials. In particular are considered the corrosion of the steel bars and the decreasing
of the concrete strength due to atmospheric agents (Biondini et al. 2006; 2015; Keskikuru et al. 2001;
Titi et al. 2016). The effects of differential decay of concrete and steel on the structural capacity of
structural elements are described in (Puppio et al., 2017).
In this paper, the material variability is considered with the aim to evaluate its effects both on local
and on global level and to establish whether the traditional approach provided by technical codes is in
favour of safety or not. With this purpose, 50 different material distributions are considered for two
reinforce concrete buildings, taking into account uniform and varied material distributions. This is done
both on local (single structural elements) and on global scale and considering the flexibility of the slabs.
The comparison is also made with a deterministic and a probabilistic approach. At a deterministic level
is made a comparison for the capacity of single columns in the case of real material distribution with
reference to the uniform one.
The current tendency of national codes is oriented towards a probabilistic approach to estimate the
seismic capacity of existing buildings (Carley et al., 2004; Consiglio Nazionale delle Ricerche, 2013;
Sudret et al., 2014). This requires a great effort in terms of number of analysis. In this case the effects
of material variability are considered towards some random material distributions artificially generated
from mean value and standard deviation of the real material distribution. All the structural, geometrical
and mechanical details of these buildings, part of complex built in the 60s-70s can be found in Sassu et
al. (2017).
In Section 2, the effects of concrete variation are examined both in terms of resistance and in terms
of stiffness. In Section 3 the two cases of study are presented considering the effective material
distribution. The great dispersion in compressive strength of the concrete is shown. In Section 4 the
different analyses carried out are considered particularly referred to described material distributions
and the results are discussed considering both deterministic and probabilistic approaches, at local and
global scale in Section 5.
The results deriving from the analysis of new and existing buildings can be influenced by different
irregularities. The current approach in earthquake engineering only considers geometric irregularities
(EN 1998-1, 2004). In particular, the irregularities can be defined considering in plan and in elevation
issues (EC8-1, 2004; Garcia et al., 2004; Varadharajan et al., 2012). Considering the planimetric
distribution is possible to highlight: (1) Symmetry in in-plan configuration - plan configuration shall
be compact and, regarding mass and stiffness distribution, approximately symmetric with respect to
two orthogonal axes; (2) Shape ratio - the ratio between plan dimensions of a rectangle in which the
building plan is inscribed shall be not greater than 4; (3) Presence of set-back or protrusion - No in-
plan set-back or protrusion shall exceed 25 % of the building total dimension in the same direction; (4)
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 15
Stiffness of slabs - floor slabs can be considered infinitely stiff in their own plan with respect to lateral
stiffness of vertical resisting elements and sufficiently strong.
Beyond in plan irregularities it is possible to found vertical irregularities, so categorized: (5) All
lateral load resisting systems (such as cores, frames, walls) shall run without interruption from their
foundations to the top of the building; (6) Both the lateral stiffness and the mass of individual storeys
shall remain constant or vary gradually, without abrupt changes, from the base to the top of the building;
(7) For buildings designed with so-called “low ductility”, the ratio of storey resistance to the resistance
required by the analysis should not vary disproportionately between adjacent floors; (8) When setbacks
are present, reductions in the horizontal building section shall occur gradually between adjacent floors.
No explicit attention is given to the effects of spatial variation of mechanical properties of the load-
bearing elements. Geometrical irregularities are usually introduced in structural analysis increasing the
effect towards: (a) The modification of the behaviour factor, a dimensionless parameter to reduce the
seismic action); (b) The introduction of the accidental eccentricity that produces torsional effects. (c)
The level of knowledge of the real material properties of the structure traduced by means of the
confidence factor FC that reduces the strength to be used in the analysis.
Non-linear static analyses are carried out considering the effects of material variability in terms of
resistance, maximum displacement and structural ductility. The effects of strength irregularity are
evaluated both on local on global scale evaluating whether the standard approach proposed by technical
code is in sake of safety or not. In addition, random material distribution are considered in order to
obtain fragility curves.
The seismic response of a r.c. building is strictly related to the mechanical behaviour of the concrete,
depending on its resistance and stiffness. The Italian code (EC8-1, 2004) indicates to use for the
analyses the average concrete compressive strength fc from in situ tests, together with the confidence
factor (CF) that is associated to the level of knowledge (LK) attained for the structure. Indeed, the
design compressive strength is obtained as ratio of the average strength to the confidence factor CF.
The higher the level of knowledge, the lower the confidence factor. The LK depends on the number of
in situ tests, while the effective distribution of in situ tests and their results do not influence the
confidence factor. This can generate some contradictory results, i.e. a confidence factor equal to 1
(CF=1) in case of large number of in situ tests characterized by a wide dispersion of results. FEMA 356
(‘Prestandard and Commentary for the Seismic Rehabilitation of Buildings - FEMA 356’, 2000) defines
the average compressive strength fc as a function of the coefficient of variation (cov) of the data obtained
from in situ test (American Society of Civil Engineers (ASCE), 2000). Conventionally, if cov > 14%,
a proper level of knowledge is not reached and the average compressive strength is reduced by the
standard deviation (Eq. (1)).
f c f c , m st.dv. (1)
Furthermore, if the covariance cov < 14% the compressive strength is assumed (Eq. (2)).
f c f c ,m . (2)
Nevertheless, a relevant dispersion of in situ tests is frequent in existing buildings and in general it
is not related to low accuracy in testing. In this sense, it is of interest to evaluate how this physical
variation could affect the seismic analysis both in local and global scale. This aspect influences also
the stiffness of the building. As well known, further source of irregularity for r.c. constructions are the
infill panels. Because of their asymmetrical disposition in the building and because of the shape and
the dimensions of the openings, the infill panels can significantly modify the structural response of
buildings in terms of stiffness and strength. (Tanganelli et al., 2013).
16
Considering in-plan configuration of a generic building with rigid slabs, the current procedure
consists in first identifying the centre of mass (CM) and centre of stiffness (CS) as relevant points for
the application of seismic forces. In order to take into account irregularity in strength of r.c. columns,
it is possible to define the centre of resistance (CR), as the point conventionally provided by the balance
of the strength of each column.
x K i y ,i
xs , j i 1
n (3)
K y ,i
i 1
n
y K i x ,i
ys , j i 1
n (4)
K x ,i
i 1
where:
xi , yi are the coordinates of the centroid of the i-th column;
Kxi , Kyi is the stiffness of the i-th column (or wall) in one of the main building directions.
The stiffness of a column or of a wall depends on the material elastic modulus. Due to the relation
usually adopted for the elastic modulus (CMIT, 2009):
0 .3
f
Ec 22' 000 c (5)
10
It is straightforward to observe that a variation of concrete compressive strength causes the
modification of the centre of stiffness. Then the eccentricity of the centre of stiffness with respect to
the centre of masses is:
xs ,k xM ,k
esx ,k (6)
Lx ,k
ys ,k yM ,k
esy ,k (7)
L y ,k
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 17
where:
xM, yM are the coordinates of the centre of mass;
Lx, Ly are the building dimension in the x and y coordinates.
To define the centre of resistance of the k-th floor composed by n columns, it is necessary to establish
which of the different mechanisms (axial or shear force, bending moment) has to be considered (Puppio
et al., 2017). Regarding the axial forces, the centre of resistance in x and y directions are:
n
N u ,i xi
xR , k i 1
, (8)
n
N
i 1 u ,i
n
Nu ,i yi
yR , k i 1
, (9)
n
N
i 1 u ,i
where:
Nu,i axial strength of the i-th columns;
xi, yi coordinates of the centre of the i-th colums;
Considering now the bending mechanism, it is analogously possible to calculate the centre of resistance
through the ultimate bending resistance Mux,i Mu,yi of the i-th column:.
n
M uy ,i ( Nu ,i ) xi
xR ,k i 1
(10)
n
i 1
M uy ,i ( Nu ,i )
n
M ux ,i ( Ni ) yi
yR , k i 1
. (11)
n
i 1
M ux ,i ( Ni )
Therefore, the eccentricity of the centre of resistance with respect to the centre of masses is:
x xM ,k
eR , xk R ,k , (12)
Lx
y yM ,k
eR, yk R,k , (13)
Ly
where:
Lx , Ly are the main dimensions of the buildings in the direction x and y.
An eccentricity between centre of mass and centre of stiffness (eS) and between centre of mass and
centre of resistance (eR) generally produces torsional effects in seismic response. In particular, in linear
elasticity it is possible to define torsional stiffness of the structure as in Eq. (12) and Eq. (13). The
mentioned parameters take into account the main effect of material variability that is the cause of torque
component. This leads to a modification of the stresses in elastic and inelastic response as the following
analysis shows, in terms of resistance of the bearing elements, ultimate displacements, ultimate shear
and ductility.
In this section, the values of mechanical parameters and their dispersion are accurately evaluated for
two cases of r.c. buildings affected by material irregularities by means of experimental in situ tests.
From them, the centre of resistance is obtained for each case according to the indications given in
Section 2.
18
A vulnerability analysis is carried out on two r.c. school buildings in Italy. Built in the early ‘70s, is
made of four substructures, from A to D (Fig. 2). The three-storey buildings, of a similar in plan total
area (about 1000 sqm), are characterized by a high number of slender columns (Fig. 3a). Apparently
regular, buildings A and C do not have axes of symmetry because of the position of the stairs and
irregularity in elevation, due to greater extension of the ground level. They respectively have a volume
of about 10’800 m3 and an average surface of 990 m2 at the ground floor and of 615 m2 at the upper
floors.
B A
(a) (b)
Fig. 2. Main views of the four buildings. Aerial views (a); identification of the Building A and C (b)
This case study was selected for its significant characteristics in terms of dispersion of mechanical
properties and due to its high seismic vulnerability. Such a dispersion is probably to foresee for many
other r.c. Italian structures dated back to the sixties and the seventies. The dispersion of concrete
compressive strength is mainly due to the lack of standardized procedure of control and automation
during the production stage. The concrete was directly produced on site in small batches, so its quality
sensitively varied during construction. This aspect is amplified in the case study due to inadequate mix
design and procedure of casting and vibration. The large dispersion in concrete compressive strength
was noted not only for single structural elements (i.e. beams and columns casted in different time) but
also within the same structural member (i.e. at the top and at the bottom of the same column). These
structures are generally composed by a series of plane frames. The design approach took into account
only vertical and wind loads: this implied weak columns unable to withstand seismic loads and to
respect the concept of capacity design. Other elements of weakness, common for r.c. structures of this
age, are also un-confined structural nodes and improper anchoring of the bars. In addition, architectural
choices were addressed to show the r.c. structures on the facades (Fig.3a), creating further elements of
vulnerability as direct exposure to climate events, together with stocky columns (Fig.3b). The main
features of the building, in terms of inter-storey height and number of columns, are summarized in
Table 1.
(a) (b)
Fig. 3. External view of the building A. Longitudinal façade (a); Main entrance (b).
3.2. Experimental tests
A large number of tests (Montgomery, n.d.) were made on the structural members of the building
A and C of the following type:
1. Tensile strength of reinforcing bars;
2. Compressive strength of coring samples;
3. Schmidt hammer test of the columns.
The results are shown in Table 2. A great dispersion both in concrete and in steel resistance is
observed. In this work the variation of steel resistances are neglected. The mean values and the relative
standard deviation for building A and C are summarized in Table 5.
Table 2. Experimental test
Concrete coring Steel failure test-Steel reinforcing
N (Number of test) fc (Compressive strengh) fy (Yield steel strength) ft
[MPa] [MPa] [MPa]
1 18.7 388 574
2 49.6 506 605
3 25.0 366 529
4 53.2 398 584
5 34.6 373 544
6 32.4 401 537
7 16.2 388 559
8 11.6 356 469
9 17.4 362 473
10 48.1 385 539
11 36.7 375 510
12 28.7 379 512
13 42.3
14 36.9
15 30.0
16 30.0
17 21.4
18 46.5
Average 32.2 389.8
Standard deviation 12.4 39.1
The hummer tests are made on 70% of the 294 columns of Building A and on 90% of the 309
columns of the Building C. Experimental tests on Building C are carried out during the phase work of
seismic rehabilitation of the structure, obtaining about 600 sclerometeric data as shown in Appendix
A.
The models of the Buildings A and C is made by continuous beam frames (Fig. 4). Building A is made
by 823 elements and 508 joint nodes (3048 DOF) while Building B is made by 833 elements and 546
nodes (3276 DOF). Both the models are fixed to the ground.
20
The eccentricity between beams and columns is modelled throughout rigid links. This happens in
correspondence of squat beam and columns.
(a) (b)
Fig. 4. Main view of the structural models of the Buildings (a) and (b)
The two buildings are modelled separately by technical joints at the first level of the structure. Two
limit scenarios are considered modelling both rigid and flexible slabs. The slabs are modelled with rigid
slabs and with flexible diaphragms, by using plate elements (about 500) with proper thickness and
stiffness. The flexible diaphragms are noticeable in the evaluation of the effects of material variability
as shown in the introduction. The flexibility of the slab has a relevant role (Sassu et al., 2017),
modifying the actions on elements characterized by different resistance and stiffness: i.e., in most cases
rigid slab can maximise the actions on the bearing elements far from the centre of stiffness. The beams
are modelled as elastic frames, neglecting their ductile capacity. The columns are modelled with
mechanical non linearities introducing plastic hinges for shear and bending mechanism. A more proper
model would consider a discretization of the ends of the columns by means fibres subjected to uniaxial
stress (Bosco et al. 2015). The shear failure mechanism is also introduced for squat elements (Puppio
et al., 2017), due to their slenderness or by the effect of infill walls or other non-structural elements
(Çelebi et al., 2010). Non-linear static analyses are carried out by means of FEM Software Midas Gen
v.1.1. The slabs are modelled with rigid slabs and with flexible diaphragms, by using plate elements.
The goal of the analysis is to evaluate whether the approach of Technical Code (Eurocode and NTC) is
conservative or not in seismic vulnerability. With this purpose, several configurations of column
strength, with the same mean value, are analysed:
1. Uniform - (Distribution 1), given by technical codes (fca for all the columns);
2. Actual - (Distribution 2), with the effective compressive strength for each column as measured from
experimental tests. Where was not possible to measure fci the mean value fca is considered;
3. Extreme - (Distributions 3 - 10), it is assumed the maximum (fc,i+Δfci and fc,i-Δfc,i) and the minimum
strength values as shown in Fig. 5;
4. Random - (Distributions 11 - 60), generating 50 random distributions with same mean value.
Distributions 1 and 2 are used to compare the two structures with the code assumptions and with actual
conditions. Distributions 3-11 evaluate the effect of extreme material distribution as shown in Fig. 5,
to obtain corresponding resistance domains. The approach used here considers only the spatial
variability of strength distribution and neglects possible effect in non-symmetric introduction of the
seismic input (Andreini et al., 2014). The random distributions (Case 4) are used to obtain fragility
curves related to different limit states.
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 21
1. Generation of the random vectors (kj). - The superscript j is the number of the vectors and
consequently of the structural analysis; the vectors kj have average value equal to zero and
standard deviation equal to one;
2. Determination of the cylindric compressive strength for the i-th column. This is generated
according to:
f c ,i f c , avg k i , j f c (14)
in which ki,j is the i-th component of the j-th vector. This ensures that each random distribution j has
the same average value (fc,avg) and the same standard deviation (Δfc) as the uniform one.
3. Determination of the elastic modulus for the j-th columns:
0.3
f
Ec ,i 22.000 c ,i (15)
10
4. Non-linear static analyses are carried out for each distribution of (fc,i; Ec,i);
5. Calculation of the PGA related to the considered limit states;
6. Determination of the fragility curves.
22
The data used here follow a Log-Normal distribution. For this reason, the probability of reaching
the analysed limit state is calculated with:
Considering the average value in Table 3 is possible to obtain α=-5%; β=-8% and γ=-4%. The data in
Table 3 are referred to the PO cases here described and the comparison is made both for flexible and
rigid slabs. The comparison, made for real and uniform material distribution, highlighted that there is
a dispersion in the results both in terms of ductility (γ) and in terms of ultimate displacement (β). In
several cases the adoption of the real material distribution produced results more severe with respect to
the uniform one. This means that the uniform material distribution is not safe. In fact, from the average
values of the analysis is evident a reduction of the 5% of α and of the 8% of β. The more representative
P.O. curves are reported in Fig. 6. “T” and “A” are respectively referred to the distribution of forces
from modal dynamic analysis and to that proportional to masses (uniform acceleration).
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 23
Building A Building C
Vb [kN] PO T +X 5500
Vb [kN] PO T -X
5500
Case 1 - Diaphragm Case 1 -
5000 5000 Diaphragm
Case 2 - Diaphragm
4500 4500
Case 1 - Flexible slab Case 2 -
4000 4000
Case 2 - Flexible slab Diaphragm
3500 3500
3000 3000
2500 2500
2000 2000
1500 1500
1000 1000
500 500
d [cm] d [cm]
0 0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
5.50
6.00
6.50
7.00
7.50
8.00
8.50
9.00
9.50
0.00
5500
Vb [kN] PO T -Y 5500
Vb [kN] PO T -Y
Case 1 - Diaphragm Case 1 - Diaphragm
5000 5000 Case 2 - Diaphragm
Case 2 - Diaphragm
4500 4500 Case 1 - Flexible slab
Case 1 - Flexible slab
4000 4000 Case 2 - Flexible slab
Case 2 - Flexible slab
3500 3500
3000 3000
2500 2500
2000 2000
1500 1500
1000 1000
500 500
d [cm] d [cm]
0 0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
5500
Vb [kN] PO A +Y 5500
Vb [kN] PO A +Y
Case 1 - Case 1 -
5000 5000
Diaphragm Diaphrag
4500 Case 2 - 4500 m
4000 Diaphragm 4000 Case 2 -
3500 Diaphrag
3500
m
3000 3000
2500 2500
2000 2000
1500 1500
1000 1000
500
500 d [cm]
d [cm] 0
0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
5.50
6.00
6.50
7.00
7.50
8.00
8.50
9.00
9.50
0.00
Fig. 6. P.O. curves for Building A and C in case of Uniform and real material (considering rigid and flexible slab)
Actual and uniform material distributions do not show great differences in terms of ultimate shear
(due to the same mean value of compressive strength of the two distributions) but exhibit significant
differences in terms of ultimate displacements. Furthermore, the diagrams in Fig. 7 highlight that the
main differences between capacity curves are due to the stiffness of the slab.
24
Table 4. Percentage of Elements with Under-estimated Capacity (P.E.U.C.) for different failure
mechanism.
Pushover Failure
Δmax Δave Δmin P.E.U.C.
direction mechanism
Building A X My 10% 21% 52% 21%
X Fz 10% 20% 50% 15%
Y Mz 17% 67% 89% 29%
Y Fy 12% 66% 98% 29%
(a) (b)
Fig. 7. - Local effects of P.O. analysis. Delta values for the ground floor of Building A. In blue bending mechanism and in
red brittle mechanism. Push over in X dir. (a) and in Y dir. (b)
(a) (b)
Fig. 8. Local effects of P.O. analysis. Delta values for the ground floor of Building C. In blue bending
mechanism and in red brittle mechanism. Push over in X dir. (a) and in Y dir. (b)
The limit states are defined with the method proposed by Vona (2014) and the indications of EMS-98
(Grünthal, 1998) summarized in Table 6.
Table 6. Limit states for structural damages in non-linear static analysis - EMS 98 (Grünthal, 1998)
Limit state Structural Damage Repairability Interstory Drift
LS0 No damage - 0.0010 ∙
LS1 Weak – No structural damage Total 0.0010 ∙ 0.0025 ∙
Moderate – limitated structural
LS2 Easily repairable 0.0025 ∙ 0.050 ∙
damage
Significant – extensive structural
LS3 High restoring cost 0.0050 ∙ 0.0100 ∙
damage
LS4 – LS5 High – Structural Collapse Difficult to repair 0.0100 ∙
26
The procedure to determine fragility curves is described in Section 0. The main parameters to generate
them are summarized in Table A1.
Standard
Building A - Dir. Y Min Max Average
Dev.
Normal [%g] 0.044 0.070 0.059 0.007
ULS
Log-Normal [-] -3.124 -2.659 -2.845 0.118
Normal [%g] 0.029 0.063 0.043 0.008
DLS
Log-Normal [-] -3.540 -2.765 -3.152 0.175
Standard
Building C - Dir. X Min Max Average
Dev.
Normal [%g] 0.049 0.088 0.060 0.007
ULS
Log-Normal [-] -3.016 -2.430 -2.816 0.123
Normal [%g] 0.029 0.072 0.044 0.007
DLS
Log-Normal [-] -3.540 -2.631 -3.131 0.177
Standard
Building C - Dir. Y Min Max Average
Dev.
Normal [%g] 0.044 0.082 0.063 0.008
ULS
Log-Normal [-] -3.124 -2.501 -2.768 0.123
Normal [%g] 0.029 0.068 0.047 0.008
DLS
Log-Normal [-] -3.540 -2.688 -3.082 0.177
In Fig. 9 the fragility curves related to buildings A and C are displayed. The fragility curves are both
for damage limit state (DLS) and for ultimate limit state (ULS). With a dashed vertical line are also
shown the values of PGAd related to the Operating Limit State and to Damages Limit State related to
the building site.
1.00
PGAc - ULS - Buiding A - Dir. X
0.055, 0.924
0.90
PGAc - ULS - Buiding A - Dir. Y
0.80
PGAc - DLS - Buiding A - Dir. X
0.70 PGAc - DLS - Bulding A - Dir. Y
0.60 PGAd - OLS 0.060, 0.593
CDF [%]
0.045, 0.533
0.50 PGAd - DLS 0.061, 0.504
0.40
0.30
0.20
0.10
0.00
0.000
0.005
0.010
0.015
0.020
0.025
0.030
0.035
0.040
0.045
0.050
0.055
0.060
0.065
0.070
0.075
0.080
0.085
0.090
0.095
0.100
PGA [%g]
Fig. 9. Fragility curves for DLS and ULS for building A. The Square and the circle in the graph represent the value of
PGA related to the uniform material distribution (Case 1). The numbers into the rectangles complete are ...
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 27
1.00
PGAc - ULS - Bulding C - Dir. X
0.90
PGAc - ULS - Bulding C - Dir. Y 0.053, 0.793
0.80 0.064, 0.718
PGAc - DLS - Bulding C - Dir. X 0.067, 0.710
0.70
0.049, 0.743
0.60
CDF [%] PGAc - DLS - Bulding C - Dir. Y
0.40
0.30
0.20
0.10
0.00
0.000
0.005
0.010
0.015
0.020
0.025
0.030
0.035
0.040
0.045
0.050
0.055
0.060
0.065
0.070
0.075
0.080
0.085
0.090
0.095
0.100
PGA [%g]
Fig. 10. Fragility curves for DLS and ULS for building C. The square and the circle in the graph represent the value of
PGA related to the uniform material distribution (Case 1). The numbers into the rectangles indicat ....
In addition, Risk Indexes are calculated with the following three different formulations, by means
shear force, PGA and return period T_r.
V
IRVb b ,C (23)
Vb , D
PGAC
IRPGA (24)
PGAD
0.41
T
IRTr R ,C (25)
T
R,D
where:
PGA is the peak ground acceleration;
Tr is the return period.
while in Eq. 23-25 the subscript C is the capacity and D is the demand;
The risk indexes assume values lower than 1 in all the analyses, i.e. the structure has not the proper
capacity to withstand seismic actions. The histograms in Fig. 11 and Fig. 12 represent the risk indexes
of the structure evaluated with Equations 23-25 relative to the cases from 1 to 10 (Fig. 5). The extreme
material distribution shows the main difference with respect to the uniform one (1). The real material
distribution (2) present risk indexes, indicative of global effects, that are nearest to the uniform one (1).
35% 35%
30% 30%
25% 25%
20% 20%
15% 15%
10% 10%
5% 5%
0% 0%
Case 10
Case 10
Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
Case 8
Case 9
Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
Case 8
Case 9
(a) (b)
Fig. 11. Risk indexes for Building A in X dir. (a) and in Y dir. (b). Eq. (23)-(25).
28
Corpo C - Direzione X IRVb IRPGA IRTR Corpo C - Direzione Y IRVb IRPGA IRTR
35% 35%
30% 30%
25% 25%
20% 20%
15% 15%
10% 10%
5% 5%
0% 0%
Case 10
Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
Case 8
Case 9
Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
Case 8
Case 9
Case 10
(c) (d)
Fig. 12. Risk indexes for Building C in X dir. (c) and in Y dir. (d). Eq. (23)-(25).
5. Discussion of results
The comparison between different models is made in terms of local and local verification.
Regarding the local verification of the single structural elements it was verified that to consider the
real material distribution (Case 2) is not in sake of safety. In this case the capacity is over-estimated for
about a maximum of the 31.0% of the numbers of the structural elements with respect to the uniform
material distributions (Case 1). This comparison is made only for the models with rigid slab. Taking
into account the effective resistance of the bearing elements produces effects both on ductile and on
brittle failure mechanisms as shown in (Puppio et al., 2017). The effects of the real material
distributions are shown in Table 8.
Considering the global analysis of the structures the comparison can be made in terms of: (1)
defined parameters ( α, β, γ ) (Eqs. (18-20)); (2) risk indexes; (3) fragility curves. As for (1), the
hypothesis of flexible slabs makes extreme the value of the parameters (Eqs. (18-20)). Considering the
average values for α, β, γ one has some variation always lower than zero. This means that considering
the effective resistance of the structural elements produces a reduction of the performances of structural
elements in terms of global shear, in terms of displacements and, consequently, in terms of ductility.
Therefore this approach that is of course more realistic should be recommended in the seismic analysis.
Risk indexes vary in the range from 15% to 30% according to whether they are calculated in
function of the global shear (IRVb), of the peak ground acceleration (IRPGA) or of the relative return
period (IRTr). Considering the real strength distribution usually does not lead to a decrease of the global
safety. The extreme material distributions lead to a major difference highlighting that some distribution
of materials can affects the vulnerability of the buildings. In particular the extreme material
distributions (Cases from 3 to 10 in Fig. 5) have values less than the 3.0% if referred to the uniform
one. This difference is very small in absolute terms (about 10% in relative terms). Considering the
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 29
small variation of the risk indexes towards the reference distribution (Case 1) the extreme material
distribution can be considered as lines of level of the function risk index (Puppio et al., 2017).
The fragility curves express the cumulate distribution of the PGA values for ULS and DLS. These
are obtained for both the buildings considering some random strength distributions (Case from 11 to
50 in Fig. 5). In Fig. 9 the fragility curves related to building A are displayed. The square and the circle
represent the point - Case 1. Although the random material distributions are determined with the same
means of the uniform one, the points representing the latter are placed beyond the half of the fragility
curves. This means that most of the analysed random cases presents PGA of collapse that is lower than
the PGA related to Case1.
Is it also possible to observe that, also considering a probabilistic approach, the uniform material
distribution is not in sake of safety.
In the current structural modelling of the structure Technical Codes consider the elastic and the
inelastic approach as alternative methods. The action derived from elastic approach produced a more
severe design of the structural elements. In case of inelastic analysis, it is possible to adopt all the
actions that act indifferently on structural elements. This approach reduces the concentration of the
actions producing a less expensive design. It is possible to consider these two approaches referring to
the centre of stiffness and to the centre of resistance of the structure. In particular: (1) in linear elastic
analysis the material variability is taken into account considering a variation in the position of the centre
of stiffness and the more severe effects of slab torsions; (2) in non-linear analysis it is possible to
considers the variation in the positions between centre of mass and centre of stiffness evaluating the
effects on the results.
6. Conclusions
The introduction of the material variability produces changes in structural response both in local
and global terms. The evaluation of the structural response of existing buildings has therefore to
consider the variability of the material strength. This can be done when, considering a limited number
of experimental tests, it is possible to find a relevant dispersion of the strength values. In r.c. buildings,
the dispersion of concrete compressive strength can affect the structural response and the verification
of the elements. In particular the variation of compressive strength can have some effects both in
ductility and in the failure mechanism. In order to obtain some simplified methods to take it into
account it should be recommended to:
introduce some extreme distributions of resistance referred to the minimum and the maximum
values (cases 3-10);
evaluate how to the considered building is prone to torsional effects (particularly for linear
elastic analysis).
These methods are highlighted in the paper referring to two cases of study. Non-linear static
analyses are carried out considering rigid and flexible slabs. In general, the results show that the flexible
slab produces more significant variation in structural response.
In the verification of local elements it was highlighted that for some elements, to take into account the
effective resistance of the material produces an effect that is not in sake of safety in structural
30
verification. The global indicator of the structural response (index of risk) exhibits a reduced variation
from the comparison between Case 1 and Case 2. The information related to how the material variability
affects the structural response is also useful to better evaluate a possible strategy of intervention.
Interventions that allow the reduced the material eccentricity thanks to the construction of a new bracing
system have to be preferred.
Acknowledgment
The Authors would like to sincerely thank the Province of Lucca for technical data on material provided
and used for the analyses.
References
Alecci, V., Briccoli Bati, S., & Ranocchiai, G. (2009). Study of Brick Masonry Columns Confined with
CFRP Composite. Journal of Composites for Construction , 13, 179–187.
American Society of Civil Engineers (ASCE). (2000). FEMA 356 Prestandard and Commentary for
the Seismic Rehabilitation of Building.
Andreini, M., Caliò, I., Cannizzaro, F., De Falco, A., Giresini, L., Pantò, B., & Sassu, M. (2014).
Seismic assessment of the historical mixed masonry-reinforced concrete government Palace in La
Spezia. In 9th International Conference on Structural Analysis of Historical Constructions.
Biondini, F., Bontempi, F., Frangopol, D. M., & Malerba, P. G. (2006). Probabilistic service life
assessment and maintenance planning of concrete structures. JOURNAL OF STRUCTURAL
ENGINEERING-ASCE, 132(5), 810–825.
Biondini, F., Camnasio, E., & Titi, A. (2015). Seismic resilience of concrete structures under corrosion.
Earthquake Engineering and Structural Dynamics, 44(14), 2445–2466.
Bonannini, E., Cinotti, M., Puppio, M. L., & Sassu, M. (2017). Seismic response of a stock of social
housings in Italy with R . C . and masonry materials . (Vol. 102, pp. 285–291). Advances in
Engineering Research (AER), volume 102 285 Second International Conference on Mechanics,
Materials and Structural Engineering (ICMMSE 2017).
Bosco, M., Ferrara, G. A. F., Ghersi, A., Marino, E. M., & Rossi, P. P. (2015). Predicting displacement
demand of multi-storey asymmetric buildings by nonlinear static analysis and corrective
eccentricities. Engineering Structures, 99, 373–387.
Bosco, M., Ferrara, G. A. F., Ghersi, A., Marino, E. M., & Rossid, P. P. (2015). Seismic assessment of
existing r.c. framed structures with in-plan irregularity by nonlinear static methods. Earthquake and
Structures, 8(2), 401–422.
Carley, K. M., Kamneva, N. Y., & Reminga, J. (2004). Response Surface Methodology 1 CASOS
Technical Report, (October).
Çelebi, M., Bazzurro, P., Chiaraluce, L., Clemente, P., Decanini, L., Desortis, A., … Stephens, C.
(2010). Recorded motions of the 6 April 2009 Mw 6.3 L’Aquila, Italy, earthquake and implications
for building structural damage: Overview. Earthquake Spectra, 26(3), 651–684.
CMIT. (2009). Circolare del Ministro delle Infrastrutture e dei Trasporti 2 febbraio 2009, n. 617,
contenente le Istruzioni per l’applicazione delle “Nuove norme tecniche per le costruzioni” di cui al
DM 14 gennaio 2008. Gazzetta Ufficiale della Repubblica Italiana n. 47 del 26 febbraio 2009,
Supplemento Ordinario n. 27.
Consiglio Nazionale delle Ricerche. (2013). CNR-DT 212/2013. Istruzioni per la Valutazione
Affidabilistica della Sicurezza Sismica di Edifici Esistenti.
De Stefano, M., & Pintucchi, B. (2002). A model for analyzing inelastic seismic response of plan-
irregular building structures. In 15th ASCE Engineering Mechanics Conference. New York.
De Stefano, M., & Pintuchi, B. (2010). Predicting torsion-induced lateral displacements for pushover
analysis: Influence of torsional system characteristics. Earthquake Engineering and Structural
Dynamics, 39(March), 1369–1394. https://doi.org/10.1002/eqe
De Stefano, M., Tanganelli, M., & Viti, S. (2014). Variability in concrete mechanical properties as a
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 31
source of in-plan irregularity for existing RC framed structures. Engineering Structures, 59, 161–
172.
EC8-1. (2004). Eurocode 8: Design of structures for earthquake resistance—Part 1: General rules,
seismic actions and rules for buildings. Brussels: European Committee for Standardization.
EC8-3. (2005). Eurocode 8: Design of structures for earthquake resistance—Part 3: Assessment and
retrofitting of buildings. Brussels: European Committee for Standardization.
EN 1998-1. (2004). EN 1998-1: Eurocode 8 - Design of structures for earthquake resistance. Part 1:
General rules, seismic actions and rules for buildings. CEN, 1, 1–229.
Fajfar, P. (2000). A Nonlinear Analysis Method for Performance-Based Seismic Design. Earthquake
Spectra, 16, 573–592. https://doi.org/10.1193/1.1586128
Fajfar, P., Marusic, D., & Perus, I. (2005). Torsional Effects in the Pushover-Based Seismic Analysis
of Buildings. Journal of Earthquake Engineering, 9(6), 831–854.
FEMA. (2006). Techniques for the Seismic Rehabilitation of Existing Buildings FEMA 547. Federal
Emergency Management Agency, Washington, D.C., 571.
Garcia, O., Islas, A., & Ayala, A. G. (2004). Effect of the In-Plan Distribution of Strength on the Non-
Linear Seismic Response of Torsionally Coupled Buildings. 13th World Conference on Earthquake
Engineering , (1891), Paper No. 1891.
Grünthal, G. (1998). European Macroseismic Scale 1998. European Center of Geodynamics and …
(Vol. 15). Retrieved from
http://scholar.google.com/scholar?hl=en&btnG=Search&q=intitle:European+Macroseismic+Scale
+1998#0
Humar, J. L., & Kumar, P. (1999). Effect of orthogonal inplane structural elements on inelastic torsional
response. Earthquake Engineering and Structural Dynamics, 28(10), 1071–1097.
Keskikuru, T., Kokotti, H., Lammi, S., & Kalliokoski, P. (2001). Effect of strength deterioration on
inelastic seismic torsional behaviour of asymmetric RC buildings. Building and Environment,
36(10), 1109–1118.
Lavan, O., & de Stefano, M. (2013). Seismic behaviour and design of irregular and complex civil
structures. Geotechnical, Geological and Earthquake Engineering, 24.
Manfredi, V., & Masi, A. (2011). Seismic assessment of existing rc buildings based on different hazard
maps. Bollettino Di Geofisica Teorica Ed Applicata, 52(2), 1–20.
https://doi.org/10.1061/(ASCE)ST.1943-541X.209; (2008) Nuove Norme Tecniche Per Le
Costruzioni NTC2008, , NTC2008 - Decreto Ministeriale D.M. G. U. n. 29, 04.02.2008 -
Supplemento Ordinario n. 30 (in Italian); (2003) Primi Elementi In Materia Di Criteri Generali Per
La Classificazione Sismica Del Territorio Nazionale E Di Normative Tecniche Per Le Costruzioni
In Zona Sismica, , OPCM3274-Ordinanza del Presidente del Consiglio dei Ministri n.3274/2003,
Supplemento ordinario G. U. n. 105, 8 maggio 2003 (in Italian
Maru, D., & Fajfar, P. (2005). On the inelastic seismic response of asymmetric buildings under bi-axial
excitation. Earthquake Engineering and Structural Dynamics, 34(8), 943–963.
Mittal, A. K., & Jain, A. K. (1995). Effective strength eccentricity concept for inelastic analysis of
asymmetric structures. Earthquake Engineering & Structural Dynamics, 24(1), 69–84.
Montgomery, C. (n.d.). Progettazione e analisi degli esperimenti, 1–18.
Muller, D., Forster, V., & Graunbner, C.-A. (2017). Influence of Material Spatial Variability on the
Reliability of Masonry Walls in Compression. In 13th Canadian Masonry Symposium, At Halifax.
Prestandard and Commentary for the Seismic Rehabilitation of Buildings - FEMA 356. (2000).
Puppio, M., Pellegrino, M., Giresini, L., & Sassu, M. (2017). Effect of Material Variability and
Mechanical Eccentricity on the Seismic Vulnerability Assessment of Reinforced Concrete
Buildings. Buildings, 7(4), 66.
Sadek, A. W., & Tso, W. K. (1989). Strength eccentricity concept for inelastic analysis of asymmetrical
structures. Engineering Structures, 11(3), 189–194. https://doi.org/10.1016/0141-0296(89)90006-0
Sassu, M., Puppio, M. L., & Mannari, E. (2017). Seismic Reinforcement of a R.C. School Structure
with Strength Irregularities throughout External Bracing Walls. Buildings, 7(3), 58.
Shinozuka, M., Feng, M. Q., Lee, J., & Naganuma, T. (2000). Statistical Analysis of Fragility Curves.
32
(appendix A)
(a)
(b)
Fig. 13. Columns at the ground floor of the Building A (a) and C (b).
M. L. Puppio et al. / Engineering Solid Mechanics 7 (2019) 33
The hummer test is not done for the underground floor because of the lack accessibility of this level
during the phase of survey.
© 2018 by the authors; licensee Growing Science, Canada. This is an open access article
distributed under the terms and conditions of the Creative Commons Attribution (CC-BY)
license (http://creativecommons.org/licenses/by/4.0/).