NONLINEAR DYNAMIC ANALYSIS OF BASE-ISOLATED R.C.

FRAMED STRUCTURES WITH IN-ELEVATION IRREGULAR

MASONRY INFILLS SUBJECTED TO NEAR-FAULT EARTHQUAKES

Fabio MAZZA1, Mirko MAZZA2, Alfonso VULCANO3

ABSTRACT

The effectiveness of the base-isolation in seismic retrofitting of existing reinforced concrete (r.c.) framed structures
needs more attention in the case of masonry infills (MIs) not uniformly distributed in elevation. Specifically, soft-
storeys may be responsible for the concentration of inelastic demand in a part of the superstructure. Moreover,
amplification of the inelastic demand is generally expected for base-isolated structures located in a near-fault area,
in the event of long-duration velocity pulses. In order to understand the nonlinear seismic behaviour of masonry-
infilled base-isolated r.c. framed structures, a six-storey r.c. framed building is primarily designed (as fixed-base)
in compliance with a former Italian seismic code, for a medium-risk zone. Then it is retrofitted by the insertion of
a base-isolation system with elastomeric and sliding bearings to meet the requirements of the current Italian code,
in a high-risk seismic zone. Failure mechanisms of totally and partly infilled structures are compared considering
three structural models: (i) bare structure with nonstructural MIs; (ii) infilled structure with in-elevation uniform
distribution of structural MIs; (iii) infilled structure with in-elevation uneven distribution of structural MIs.
Nonlinear dynamic analysis of the original (fixed-base) and retrofitted (base-isolated) structures is carried out
considering near-fault ground motions with significant horizontal pulses, selected from the Pacific Earthquake
Engineering Research center database. Numerical results highlight that in-elevation distribution (i.e. totally and
partly infilled structures) and modelling (i.e. nonstructural and structural) of MIs significantly influence the
nonlinear seismic response of the superstructure while only limited effects are observed for the base-isolation
systems.

Keywords: R.c. framed structures; Nonlinear modelling of masonry infills; Elastomeric and sliding bearings;
Near-fault earthquakes; Nonlinear dynamic analysis.

1. INTRODUCTION

Irregularities in elevation due to soft-storeys or unsymmetrical layout of the masonry infills can produce
significant variations in stiffness, strength and mass distribution of reinforced concrete (r.c.) framed
buildings, as well as producing unfavourable local failure mechanisms (Hak et al. 2012). During the
earthquake in L’Aquila (April 6th, 2009) a large number of buildings suffered severe damage or collapse
due to the influence of different failure modes of the MIs (Braga et al. 2011). Many models with different
degrees of discretization and accuracy have been proposed in the literature to reproduce the in-plane
nonlinear seismic behaviour of the MIs (Liberatore et al. 2017). Each model has its advantages and
limitations, but the infill panel with an equivalent diagonal strut model without tension allows obtaining
the key features of the inelastic response without detailed information about local phenomena.
To mitigate the detrimental effects of irregular MIs and retrofit the building, base-isolation systems can
be inserted in the framed structure. Different base-isolation strategies or their combinations can be used

1
Researcher, Dip. Ingegneria Civile, Università della Calabria, Rende (CS), Italy, [email protected]
2
Research fellow, Dip. Ingegneria Civile, Università della Calabria, Rende (CS), Italy, [email protected]
3
Full professor, Dip. Ingegneria Civile, Università della Calabria, Rende (CS), Italy, [email protected]
(Mazza et al. 2017): i) increasing the fundamental vibration period of the structure, to shift it in the range
of low spectral accelerations; ii) limiting the maximum force transmitted by sliders to the superstructure,
depending on their friction coefficient. Therefore, the isolation systems are usually made with
elastomeric bearings (e.g. high-damping-laminated-rubber bearings, HDRBs), sometimes coupled in
hybrid combination with flat steel-PTFE sliding bearings (SBs), or curved friction pendulum bearings
(FPBs). In the case of an in-parallel combination of HDRBs and SBs (i.e. EFBI structure) or FPBs acting
alone (i.e. FPBI structure), the base-isolated structure behaves as a fixed-base structure in the horizontal
direction until the friction threshold of the SBs or FPBs is not exceeded. The features of earthquakes
recorded in proximity to the source can be significantly different from those observed for far-fault
ground motions (Chioccarelli and Iervolino 2010). In particular, the frequency content of the motion
transmitted by the isolators can become critical for the superstructure under the long-duration
(horizontal) velocity pulses recorded in the near-fault area (Mazza and Vulcano 2012).
The aim of the present work is to investigate the nonlinear seismic behaviour of masonry-infilled base-
isolated r.c. framed structures subjected to near-fault (NF) ground motions. For this purpose, a six-storey
r.c. framed building designed (as fixed-base) in compliance with a former Italian seismic code (D.M.
1996) for a medium-risk zone, is then retrofitted by the insertion of a base-isolation system to meet the
requirements of the current Italian code (NTC 2008) in a high-risk seismic zone. In detail, elastomeric
(i.e., HDRBs) and friction (i.e. steel-PTFE sliders, SBs, or pendulum bearings, FPBs) bearings are
considered. Besides the (fixed-base) primary structure, three cases of base-isolation are studied: HDRBs
acting alone (i.e. the EBI structure); in-parallel combination of HDRBs and SBs (i.e. the EFBI structure);
FPBs acting alone (i.e. the FPBI structure). In order to investigate the influence of MIs on the nonlinear
seismic response of the original and retrofitted structures, three layouts of MIs are considered for the
superstructure: (i) bare structure, considering only the mass contribution of nonstructural MIs; (ii)
regularly infilled structure, with in-elevation uniform distribution of structural MIs; (iii) irregularly
infilled structure, with soft-storey at the lowest levels of the superstructure. Then, nonlinear dynamic
analysis of the original (fixed-base) and retrofitted (base-isolated) structures is carried out considering
NF earthquakes with significant horizontal pulses, which are selected from the Pacific Earthquake
Engineering Research center database (PEER 2008) and scaled on the basis of the design hypotheses
adopted for the test structures.

2. FIXED-BASE ORIGINAL BUILDING

A six-storey residential building with r.c. framed structure, whose symmetric plan is shown in Figure 1,
is considered as primary test structure. Masonry infills are considered as nonstructural elements
regularly distributed in the corner bays of the perimeter frames (Figure 1a), with two in-elevation
configurations: i) totally infilled frames (TIFs), with a regular distribution along the building height
(Figure 1b); ii) partly infilled frames (PIFs), with soft-storey at the first (Figure 1c), the first two (Figure
1d) and the first three levels (Figure 1e) of the superstructure. Note that infill walls with openings are
not taken into account. The infill typology selected for this study consists of two leaves of clay horizontal
hollowed bricks, with a thickness of twe=12cm (exterior) and twi= 8cm (interior). The infill mechanical
properties are reported in Table 1: fwh and fwv, compression strengths in the horizontal and vertical
directions; fwu, sliding shear resistance of the mortar joints; fws, shear resistance under diagonal
compression; Ewh and Ewv, secant moduli of elasticity in the horizontal and vertical directions; G, shear
modulus; , coefficient of Poisson. A simulated design of the original framed building is carried out in
accordance with the former Italian code (D.M. 96), for a medium-risk seismic region and a typical
subsoil class. The gravity loads for the r.c. framed structure are represented by a dead load of 4.2 kN/m2
on the top floor and 5.0 kN/m2 on the other floors, and a live load of 2.0 kN/m2 on all the floors; an
average weight of about 2.7 kN/m2 is considered for the masonry infill walls (both leaves). Concrete
cylindrical compressive strength of 25 N/mm2 and steel reinforcement with yield strength of 375 N/mm2
are considered. The geometric dimensions of the lateral, interior and central frames are shown in Figure
1, while cross section of beams (i.e. deep and flat) and columns (i.e. corner, perimeter and central) are
reported in Table 2. Further details regarding the design of the fixed-base structure can be found in
Mazza et al. (2017).

2
 Table 1. Mechanical properties of the masonry infills (units in MPa, apart from ).

fwh fwv fwu fws Ewh Ewv G 

1.11 1.50 0.25 0.31 991 1873 1089 0.20

(a) Plan

(b) In-elevation totally infilled frames (c) In-elevation partly infilled frames:
(TIFs) soft-storey at the first level (PIFs.1)

(d) In-elevation partly infilled frames: (e) In-elevation partly infilled frames:
soft-storey at the first two levels (PIFs.1_2) soft-storey at the first three levels (PIFs.1_2_3)

Figure 1. Original (fixed-base) structure (units in cm)

The fundamental vibration periods and effective masses, expressed as a percentage of the total mass
(mtot) of the building, along the in-plan X and Y principal directions are reported in Table 3, with
reference to three structural models: i) bare frame (BF); ii) totally infilled frame (TIF); iii) partly infilled
frame (PIF). The expression proposed by Bertoldi et al. (1993) is used to evaluate the width (bw) of the
diagonal strut, with length dw and total thickness tw (=twe+twi), equivalent to the masonry infill panel:

bw dw =K1 (λh )+K2 (1)

where the parameters K1 and K2 are defined as functions of λh (see Table 4), h being the centerline
height of a frame storey and λ a dimensionless relative stiffness parameter (Stafford Smith 1966).

3
 Table 2. Cross section of r.c. frame members (units in cm).
Column Beam
Storey Corner Perimeter Interior Deep Flat
6 30x30 30x30 30x30 30x45 40x25
5 30x35 30x40 40x40 30x45 40x25
4 30x40 30x50 50x50 30x50 50x25
3 30x40 30x50 50x50 30x55 50x25
2 30x50 30x60 60x60 40x60 60x25
1 30x50 30x60 60x60 40x70 60x25

Table 3. Dynamical properties of the fixed-base structure (mtot=1465 ton).

Case T1,X (s) T1,Y (s) me1,X me1,Y
BF 0.763 0.624 77.2% mtot 76.0% mtot
TIF 0.645 0.557 78.9% mtot 77.5% mtot
PIF.1 0.664 0.568 81.5% mtot 79.7% mtot
PIF.1_2 0.685 0.575 82.4% mtot 80.2% mtot
PIF.1_2_3 0.706 0.586 81.6% mtot 79.7% mtot

Table 4. Numerical values of the K1 and K2 parameters (Stafford Smith 1966).

λh<3.14 3.14<λh<7.85 λh>7.85
K1 1.3 0.707 0.47
K2 0.178 0.01 0.04

3. BASE-ISOLATED RETROFITTED BUILDINGS

To attain performance levels imposed by the current Italian code (NTC 2008), in a high-risk seismic
zone (peak ground acceleration on rock, ag=0.262g at the life-safety limit state) with medium subsoil
class (class C, site amplification factor, S=1.319), three in-plan configurations of elastomeric and sliding
(flat or curved) bearings are employed to retrofit the original (fixed-base) framed building: (a) EBI
structure, with elastomeric bearings acting alone (i.e. HDRBs type 1 in Figure 2a); (b) EFBI structure,
with an in-parallel combination of elastomeric (i.e. HDRBs type 2) and friction (i.e. steel-PTFE sliding
bearings, SBs) bearings shown in Figure 2b; (c) FPBI structure, with friction pendulum bearings (i.e.
FPBs in Figure 2c) acting alone. An additional mass of 511 ton, placed above the isolation level, is
assumed at the level of the rigid beams, with a cross section of 50x100 cm2. The base-isolation systems
are designed on the same values of the fundamental vibration period (i.e. T 1H=2.5s) and equivalent
viscous damping ratio (i.e. H=18%).
The design of the EBI structure shown in Figure 2a is carried out considering the same dimensions of
the HDRBs type 1. A value of the stiffness ratio of the isolation system, defined as the ratio between the
horizontal (KH0) and vertical (KV0) nominal stiffnesses of the HDRBs, K0=800 is considered. A shear
modulus G=0.4 MPa and a volumetric compression modulus Eb=2000 MPa are assumed for the
elastomer. The isolation system properties are reported in Table 5: i.e. the horizontal (KH0) and vertical
(KV0) nominal stiffnesses and the corresponding equivalent damping coefficients (CH and CV), assuming
an equivalent viscous damping ratio in the vertical direction, V, equal to 5%.
The design of the EFBI structure includes an in-parallel combination of HDRBs type 2 and four interior
and two perimeter steel-PTFE SBs (Figure 2b). This solution corresponds to a value 0.44 of the nominal
sliding ratio S0(=FS0/FS0,max) of the SBs under gravity loads, defined as the global sliding force (F S0)
divided by the maximum sliding force (FS0,max); for the latter force, sliding bearings are supposed placed
under each column of the test structure. The same nominal stiffness ratio adopted for the EBI structure
(i.e. K0=800) is assumed for the HDRBs type 2. Equivalent viscous damping ratios H,HDRBs=11.6% and
H,SBs=6.4% are evaluated for the elastomeric and sliding bearings, respectively, considering a dynamic-

4
fast sliding friction coefficient =4.2%. The equivalent viscous damping ratio of the HDRBs type 2 in
the vertical direction (V) is assumed equal to 5%. Finally, main properties of the HDRBs type 2 are
reported in Table 6.
The alternative of using only twenty FPBs, with the same effective radius of curvature (R) of the sliding
interface, is lastly taken into account in the design of the FPBI structure shown in Figure 2c. In detail,
the in-plan distribution of maximum axial load capacity (PEd) is assumed so as to obtain six types of FP
bearings, characterized by a same value of the dynamic-fast sliding friction coefficient (i.e. =4.2%) for
different values of the maximum compression axial load of the FP bearing (Psd). Main properties of the
FPBs are reported in Table 7. Further details regarding the design of the base-isolation systems for
retrofitting the original test structure can be found in Mazza et al. (2017).

(a) EBI structure (b) EFBI structure (c) FPBI structure

Figure 2. Base-isolated retrofitted structures

Table 5. Properties for the elastomeric bearings of the EBI structure (units in kN, cm and s).

KH0 KV0 CH CV
7.69 6149 0.79 3.43

Table 6. Properties for the elastomeric bearings of the EFBI structure (units in kN, cm and s).

KH0 KV0 CH CV
8.49 6789 0.81 3.24

Table 7. Properties for the friction pendulum bearings of the FPBI structure (units in kN and cm).

FPB R Psd PEd

1,5,16,20 216 515 970
2,4,17,19 216 974 1839
3,18 216 1029 1941
6,10,11,15 216 1020 1926
7,9,12,14 216 1458 2753
8,13 216 1547 2920

4. NONLINEAR MODELLING OF MASONRY INFILLS

Failure mechanisms of totally and partly infilled structures are compared by substituting the MIs of the
lower storeys of the test structure with glass windows and following a change in the use of these floors
from residential to office. A simplified diagonal pin-jointed strut model reacting only in compression
(Figure 3a), having inclination θ with respect to the horizontal direction, takes into account the in-plane
failure modes that can occur in the infill panels when subjected to seismic loading. Thus, four failure
modes are considered, with the corresponding equivalent compressive strengths (Bertoldi et al. 1993)
for diagonal compression (w1), crushing in the corners in contact with the frame (w2), sliding shear
along horizontal joints (w3) and diagonal tension (w4):
5
 1.16fwv tanθ 1.12fwv sinθcosθ (1.2sinθ+0.45cosθ)fwv +0.3σv ,
σw1 = , σw2 = , σw3 =
K1 +K2 λh K1( λh ) +K2 (λh )
-0.12 0.88 bw d w
0.6sinθfws +0.3σv
σw4 = (2a,b,c,d)
bw d w
where v is the vertical compression stress due to gravity loads. Then, the maximum lateral strength of
the strut is evaluated as
Fw =σw,min bw t w cosθ, σ w,min =minσ w1 ,σ w2 ,σw3 ,σw4 (3a,b)

As shown in Figure 3b, the skeleton curve of the lateral force-interstorey drift (F-) law considers three
linear branches, depending on parameters α, β and  (Cavaleri et al. 2014). In detail, the first ascending
branch corresponds to the uncracked stage, until the attainment of point C characterized by

FwC =αFw , α=0.4, ΔwC =FwC k w1 , k w1 =Ewθ bw t w cos θ dw ,

2

-1
 cos4θ sin4θ  1 2ν 
Ewθ =  + +cos2θsin2θ −  (4a,b,c,d,e)
 Ewh Ewv  G Ewv 
The second ascending branch represents the post-cracking phase up to the attainment of point FC,
corresponding to the full development of the cracking

FwFC =Fw , ΔwFC =ΔwC +(FwFC −FwC ) kw2 , kw2 =βkw1, β=0.15 (5a,b,c,d)

The third descending branch describes the post-peak strength deterioration of the infill up to the
attainment of the residual values of strength and displacement (point RS)
1 F   F −F 
FwRS =0.7FwFC , ΔwRS = ln  wFC eξΔwFC , ξ=0.02, k w3 =tan  wFC wRS  (6a,b,c,d)
ξ  FwRS   ΔwRS −ΔwFC 
A pivot hysteretic model is adopted to predict the nonlinear force-displacement law of the equivalent
diagonal strut (Figure 3b), based on geometrical rules that define loading and unloading branches
corresponding to the unsymmetrical tension-compression behaviour of the MIs.

(b)
(a)
Figure 3. Nonlinear modelling of a masonry infill panel

The cyclic behaviour can be described by primary and pinching pivot points, which are governed by
four parameters (1, 2,  1 and 2) when the strength envelope, without tension resistance, is defined.
Simplifications of the pivot model can be adopted when applied to masonry infills (Cavaleri and Di
Trapani 2014). Specifically, the hysteretic law is governed by parameter 2(=0.25) describing a
fundamental pivot (FP) point as function of the cracking resistance (FwC). Monotonic constitutive laws
of the perimeter masonry infills (MIs), for the six levels of the test structure, are plotted in Figure 4. It
should be noted that the three fundamental points (i.e. C, FC and RS shown in Figure 3b) of each curve
are characterized by values of the horizontal strength and displacement in the X direction (Figure 4a)
lower than those in the Y direction (Figure 4b).
6
 (a) (b)

Figure 4. Monotonic curves for the masonry infills (MIs) of the test structure

5. NUMERICAL RESULTS

To investigate the influence of an in-elevation irregular distribution of MIs, the nonlinear seismic
analysis is carried out on three configurations of the superstructure: (i) bare structure (BS), but
accounting for the mass contribution of nonstructural MIs; (ii) infilled structure with in-elevation regular
structural MIs (IS_R); (iii) infilled structure with in-elevation irregular structural MIs (IS_IR). The
original (i.e. fixed-base: FB) and retrofitted (i.e. base-isolated: EBI, EFBI and FPBI) buildings are
modelled and analyzed by the structural analysis program SAP 2000 (CSI 2017). In order to take into
account the nonlinear behaviour of the framed structure, a one-component lumped plasticity model with
an elastic (central) beam and two inelastic rotational end-hinges represents the flexural behaviour of
beams and columns. The plastic hinges are evaluated on the basis of the material and r.c. cross-section
properties, assuming a five-point moment-plastic rotation law to represent: no deformation before
yielding, yielding, ultimate capacity, residual strength and failure. Specifically, plastic hinge for major
axis of bending only or for axial load and bending moment around the two principal axes are considered
for beams or columns, respectively. According to the Rayleigh hypothesis, the damping matrix of the
structure (superstructure) is assumed as a linear combination of the mass and stiffness matrices,
assuming an equivalent viscous damping ratio equal to 5% (or 2%) with reference to the vibration
periods of the fixed-base structure (or base-isolated structures) corresponding to the high-participation
modes with components prevailing in the X (i.e. T1X) and Y (i.e. T1Y) directions. The simplified diagonal
pin-jointed strut model, previously described for an infill panel, takes into account the in-plane failure
modes that can occur. Link elements are adopted to describe the base-isolation systems, assuming: i) an
in-parallel combination of linear spring and viscous dashpot, both in the horizontal and vertical
directions, for the elastomeric bearings; ii) velocity-dependent friction properties for the flat (i.e. R→)
and curved (i.e. R=216 cm) sliding bearings, with post-slip stiffness in the shear directions related to the
pendulum radius of the slipping surfaces and gap behaviour in the axial direction for tensile loads. In
detail, the friction coefficient is based on an exponential law, which attains the value max or min,
respectively, at very high or very low sliding velocity. The nonlinear seismic analysis is carried out
assuming max/min=2.5 and a rate parameter  equal to 5.5 s/m, depending on experimental results.
Then, near-fault (NF) earthquakes with pulse-like horizontal ground velocity are selected from the
Pacific Earthquake Engineering Research center database (PEER 2008). To this end, the In-Spector
software is adopted for a computer aided selection of seven real spectrum-compatible (Acunzo et al.
2014), considering ground motions whose spectrum matches the NTC 2008 design spectrum at the life-
safety limit state to a certain value of the root mean square difference between a real spectrum and the
target spectrum. Then, the selected earthquakes are scaled on the basis of the design hypotheses adopted
for the test structures. To this end, the modified acceleration spectrum intensity is evaluated by
integration of the spectral values of the acceleration in a suitable range of vibration periods (Mazza and
Labernarda 2017), where the lower limit accounts for contribution of the higher modes to structural
response while the upper limit considers the lengthening of period due to nonlinear structural behaviour.
In Table 8, the main data of the NF earthquakes are reported: earthquake location, date, recording station,
magnitude, peak ground acceleration in the horizontal directions (PGAH1 and PGAH2), scale factor, ratio
between the pulse period (TP) and the fundamental vibration period of the base-isolated structures (TI).
7
 Table 8. Main data of the selected near-fault earthquakes.

Earthquake Date Station Mw PGAH1 PGAH2 SF TP/TI

Chi-Chi 20/09/99 TCU068 7.3 0.512 g 0.371 g 0.60 4.92
Northridge 17/01/94 Newhall P.C. 6.7 0.419 g 0.357 g 0.90 1.19
Sup. Hills 24/11/87 Parachute 6.4 0.432 g 0.384 g 0.39 0.96
Cape Mend. 25/04/92 Petrolia 7.1 0.585 g 0.662 g 0.70 1.20
Kobe 16/01/95 Takatori 6.9 0.618 g 0.671 g 0.31 0.62
Tabas 16/09/78 Tabas 7.7 0.854 g 0.862 g 0.50 2.48
Erzincan 13/03/92 Erzincan 6.7 0.496 g 0.387 g 0.50 1.08

Firstly, storey damage of the superstructure for the original (i.e. fixed-base, FB) and retrofitted (i.e. EBI,
EFBI and FPBI) structures is plotted in Figures 5 and 6, for irregular (i.e. partly infilled frames with
soft-storey at the first level, PIFs.1) and regular (i.e. totally infilled frames, TIFs) in-elevation
distribution of the MIs, respectively. The results discussed below are obtained as maximum of those
obtained for each of the seven pairs of accelerograms. In detail, the maximum interstorey drift ratio,
defined as maximum drift () divided by the storey height (h), is plotted for all storeys and in both
principal directions of the building plan. The drift ratio thresholds related to different damage levels of
r.c. elements, in the case of ductile structural systems, are also checked (Ghobarah 2004): i.e. light to
moderate damage, /h<1%; severe damage up to partial collapse, 1%≤/h≤3%; collapse, /h>3%. As
expected, the original (fixed-base) structure suffers moderate-to-severe damage, highlighting an
irregular vertical distribution of drift ratio, characterized by higher values at the lower storeys.
Interestingly, the pulse-type nature of the selected NF ground motions can induce unexpected irreparable
damage also at the lower storeys of the retrofitted (base-isolated) test structures. Moreover, NF effects
are more evident in the in-plan X direction (Figures 5a,c and Figures 6a,c), characterized by flat beams
in the interior frames, than in the Y one (Figures 5b,d and Figures 6b,d). Note that MIs are protective in
the infilled base-isolated structures (i.e. the IS_IR and IS_R models in Figures 5c,d and 6c,d,
respectively), while the bare base-isolated structures (i.e. the BS_IR and BS_R models in Figures 5a,b
and 6a,b, respectively), where the MIs are assumed as nonstructural elements, are the most vulnerable.

(a) (b)

(c) (d)
Figure 5. Storey drift ratio for the original (fixed-base) and retrofitted (base-isolated) structures with partly infilled
frames: (a,b) bare structures; (c,d) infilled structures
8
 (a) (b)

(c) (d)

Figure 6. Storey drift ratio for the original (fixed-base) and retrofitted (base-isolated) structures with totally infilled
frames: (a,b) bare structures; (c,d) infilled structures

This behaviour emphasizes that the increase of lateral stiffness due to the insertion of the MIs produces
only a slight increase of seismic loads in the base-isolated structures, due to the decrease of the
fundamental vibration period, accompanied by a significant enhancement of the energy dissipation
capacity. The irregular distribution of masonry infills (i.e. the PIFs.1 cases) is responsible for the
concentration of inelastic demand in the soft-storey of the original and retrofitted structures, especially
for the IS_IR models in Figures 5c,d. On the other hand, a regular in-elevation distribution of the MIs
(i.e. the TIFs cases) induces drift ratio values in the bare structures (i.e. the BS_R models in Figures 6a
and 6b) greater than those in the infilled structures (Figures 6c and 6d).
Afterwards, similar graphs are plotted in Figure 7, where the storey damage of the infilled structures is
reported for three different in-elevation irregular distributions of MIs, assuming their inclusion in the
structural model. Specifically, detrimental effects of soft-storey at the first (i.e. PIF.1), the first two (i.e.
PIF.1_2) and the first three (i.e. PIF.1_2_3) storeys of the superstructure are investigated. As can be
observed, the behaviour of the fixed-base (original) structures is typically governed by the soft-storey
developing in the bottom storeys (Figures 7a,b), which significantly increases the inter-storey drift
values leading to the formation of local failure mechanisms at the soft-storeys. In the base-isolated
structures, irregular arrangement of the MIs induces an increase of structural damage for an increasing
number of storeys where the formation of a soft-storey mechanism occurs (Figures 7c-7h). Moreover,
the pulse-type nature of the NF ground motions is also responsible for the amplification and
concentration of inelastic demand in the lower storeys of the superstructure while the upper storeys
remain substantially elastic. Finally, base-isolation by elastomeric bearings (EBI structures in Figures
7c,d) is more effective than seismic retrofitting of the fixed-base structure obtained by the other base-
isolations systems (EFBI and FPBI structures in Figures 7e-7h).
Next, to compare the response of the base-isolation systems in the cases considered above (EBI, EFBI,
FPBI structures), the maximum horizontal deformation (uH,I) of the isolation systems is plotted in Figure
8. Note that the in-elevation distribution (PIFs.1 and TIFs cases in Figures 8a,b and 8c,d, respectively)
and modelling of MIs (BS and IS models) of MIs do not significantly influence the deformation of the
isolators. Moreover, the FPBI structures show higher values of horizontal displacement than those
observed for the EBI and EFBI ones, even exceeding the NTC 2008 design threshold.

9
 (a) (b)

(c) (d)

(e) (f)

(g) (h)

Figure 7. Storey drift ratio for the original (fixed-base) and retrofitted (base-isolated) structures with a soft-storey
at the first (PIF.1), the first two (PIF.1_2) and the first three (PIF.1_2_3) levels of the superstructure

Similar results are also reported in Figure 9 where the horizontal displacement of the base-isolation
systems is plotted as a function of an increasing number of soft storeys for the infilled structures with
in-elevation irregular structural MIs (i.e. PIF.1, PIF.1_2 and PIF.1_2_3).
Further results, omitted for the sake of brevity, confirm that during the nonlinear dynamic analyses a
significant different drift response is observed between the fixed-base and base-isolated structures. In
particular, all the examined base-isolation systems are found to be effective for the vibration control in
the case of regular in-elevation distribution of MIs, while moderate to severe damage is highlighted in
the case of soft-storeys in the superstructure.
10
 (a) (b)

(c) (d)

Figure 8. Horizontal displacement of the base isolation systems for the retrofitted (base-isolated) bare and infilled
structures: partly (a,b) and totally (c,d) infilled frames.

(a) (b) (c)

Figure 9. Horizontal displacement of the base isolation systems for the retrofitted (base-isolated) infilled structures
with a soft-storey at the first (PIF.1), the first two (PIF.1_2) and the first three (PIF.1_2_3) storeys of the
superstructure.

6. CONCLUSIONS

The nonlinear dynamic response of fixed-base structures, with and without soft-storeys, retrofitted with
a base-isolation system in order to attain the performance levels imposed by the current Italian code in
a high-risk seismic zone, has been studied under NF earthquakes. Three structural configurations are
compared: bare structure with nonstructural MIs; totally infilled structure, with regular distribution of
MIs; partly infilled structure, with irregular in-elevation distribution of MIs. Besides the original FB
existing structure, three base-isolation systems are designed: HDRBs acting alone (EBI structure); an
in-parallel combination of HDRBs and SBs (EFBI structure); FPBs acting alone (FPBI structure). The
original FB structure suffers moderate-to-severe damage, highlighting an irregular vertical distribution
characterized by higher values at the lower storeys. The pulse-type nature of the selected NF ground
motions can induce unexpected irreparable damage also in the lower storeys of the retrofitted (base-
isolated) structures. Specifically, the EBI, EFBI and FPBI isolation systems are effective for the
vibration control in the case of regular in-elevation distribution of the MIs, while irreparable damage
occurs in the case of a soft first-storey of the superstructure. Irregular in-elevation distribution of MIs
(PIFs cases) is responsible for concentration of inelastic demand in the soft-storey of the base-isolated
structures, especially when the IS model is considered. The assumption of an open storey at the second
and third storeys of the fixed-base structure produces a value of drift at the first storey lower than that
11
observed for the case of first soft-storey, while a slight increase of drift at the first storey of the base-
isolated structures generally occurs for an increasing number of soft-storeys. The response of the base-
isolation systems shows that in-elevation distribution (PIFs and TIFs cases) and modelling (BS and IS
models) of MIs do not significantly influence their horizontal deformation. Besides, the deformation of
the FPBI isolation system is greater than that observed for the EBI and EFBI structures, even exceeding
the NTC 2008 threshold.

7. ACKNOWLEDGMENTS

This work was financed by Re.L.U.I.S. (Italian network of university laboratories of earthquake
engineering), in line with “Convenzione D.P.C.–Re.L.U.I.S. 2017, PR6 line, Isolation and Dissipation”.

8. REFERENCES

Acunzo G, Pagliaroli A, Scasserra G (2014). In-Spector: un software di supporto alla selezione di accelerogrammi
naturali spettrocompatibili per analisi geotecniche e strutturali. 33° Convegno Nazionale GNGTS, Bologna, Italy,
2:107-114.
Bertoldi SH, Decanini LD, Gavarini C (1993). Telai tamponati soggetti ad azione sismica, un modello
semplificato: confronto sperimentale e numerico. 6° Convegno Nazionale di Ingegneria Sismica, ANIDIS, Perugia,
Italy, 815-824.
Braga F, Manfredi V, Masi A, Salvatori A, Vona M (2011). Performance of non-structural elements in RC
buildings during the L’Aquila, 2009 earthquake. Bulletin of Earthquake Engineering, 9(1):307-324.
Cavaleri L, Di Trapani F (2014). Cyclic response of masonry infilled RC frames: experimental results and
simplified modelling. Soil Dynamics and Earthquake Engineering, 65:224-242.
Cavaleri L, Papia M, Macaluso G, Di Trapani F, Colajanni P (2014). Definition of diagonal Poisson’s ratio and
elastic modulus for infill masonry walls. Materials and Structures, 47:239-262.
Chioccarelli E, Iervolino I (2010). Near-source seismic demand and pulse-like records: a discussion for L’Aquila
earthquake. Earthquake Engineering and Structural Dynamics, 39(9):1039-1062.
CSI Computers and Structures (2017). SAP2000 v19 Integrated Finite Element Analysis and Design of Structures.
DM96 (1996). Norme tecniche per le costruzioni in zone sismiche e relative istruzioni. Italian Ministry of Public
Works, D.M. 16-01-1996 and C.M. 10-04-1997.
Ghobarah A (2004). On drift limits associated with different damage levels. Performance-Based Seismic Design:
Concepts and Implementation. Proceeding of the International Workshop, Bled, Slovenia.
Hak S, Morandi P, Magenes G, Sullivan TJ (2012). Damage control for clay masonry infills in the design of rc
frame structures. Journal of Earthquake Engineering, 16(S1):1-35.
Liberatore L, Noto F, Mollaioli F, Franchin P (2017). A comparative assessment of strut models for the modelling
of in-plane seismic response of infill walls. COMPDYN 2017, 6th ECCOMAS Thematic Conference on
Computational Methods in Structural Dynamics and Earthquake Engineering, Rhodes, Greece, 2:3255-3268.
Mazza F, Labernarda R (2017). Structural and non-structural intensity measures for the assessment of base-isolated
structures subjected to pulse-like near-fault earthquakes. Soil Dynamics and Earthquake Engineering, 96:115-127.
Mazza F, Mazza M, Vulcano A (2017). Nonlinear response of r.c. framed buildings retrofitted by different base-
isolation systems under horizontal and vertical components of near-fault earthquakes. Earthquakes and Structures,
12:135-144.
Mazza F, Vulcano A (2012). Effects of near-fault ground motions on the nonlinear dynamic response of base-
isolated r.c. framed buildings. Earthquake Engineering and Structural Dynamics, 41:211-232.
NTC08 (2008). Technical Regulations for the Constructions. Italian Ministry of Infrastructures, D.M. 14-01-2008.
PEER (2008). Pacific Earthquake Engineering Research Center database. http://ngawest2.berkeley.edu.
Stafford Smith B (1966). Behaviour of square infilled frames, Journal of the Structural Division, 92(1):381-403.

12