Time-History Analysis of Reinforced Concrete Frame ...iranarze.ir/wp-content/uploads/2018/05/E7201-IranArze.pdf · Time-History Analysis of Reinforced Concrete Frame Buildings ...

Arab J Sci EngDOI 10.1007/s13369-016-2366-1

RESEARCH ARTICLE - CIVIL ENGINEERING

Time-History Analysis of Reinforced Concrete Frame Buildingswith Soft Storeys

Sayed Mahmoud1,2 · Magdy Genidy2 · Hesham Tahoon2

Received: 21 March 2016 / Accepted: 17 November 2016© King Fahd University of Petroleum & Minerals 2016

Abstract This research study investigates the change indynamic characteristics of reinforced concrete moment-resisting frame buildings without and with fully infill walls.In addition, building models with partially infill walls havealso been investigated. A set of different building modelshave been developed to perform the analysis as (1) bareframe (without infill walls), (2) frame with fully infill walls,(3) frame models with infill panels and soft storey locatedat base level, 3rd storey level, 6th storey level, 9th storeylevel, and 12th storey level. The equivalent diagonal strutmethod has been utilized in order to account for the stiffnessand structural action of the masonry infill panels. Dynamictime history, using two ground motion records from nearand far-fault regions, has been used to perform the seismicanalysis of the consideredmodel configurations. The selectedtwo ground motion records have been scaled to meet theexpected peak ground acceleration in Cairo zone. The twoground excitations are applied separately in two orthogo-nal directions. The structural software package ETABS hasbeen used in developing the building models and performingthe simulation analysis. Some selected numerical simulationresults in terms of storey shear forces, lateral deflections,interstory drift ratios and overturningmoments at each storey

B Sayed [email protected]

Magdy [email protected]

Hesham [email protected]

1 Department of Construction Engineering, College ofEngineering, University of Dammam, Dammam, SaudiArabia

2 Faculty of Engineering at Mataria, Helwan University, Cairo,Egypt

level are obtained for all the considered configurations andpresented in comparative way. Based on the obtained time-history results, it has been found that the dynamic storeyresponses for bare frame model significantly differ from theresponses obtained for both fully infill and partially infillframe models.

Keywords Time-history analysis · Masonry infill walls ·Single diagonal strut · Soft storey

1 Introduction

A large number of moment-resisting frame buildings havebeen or are being constructed. In addition, more are beingplanned to be constructed all over the world. These typesof buildings have various social and functional uses such asparking garages, reception lobbies and any other open airspaces which have no infill masonry walls and called softor weak storey. Although multi-storey reinforced concretebuildings with open spaces are highly vulnerable to collapseunder the effect of lateral earthquake loads, they have becomean unavoidable feature for the most of the newly constructedreinforced concrete framed buildings. This may be due to theessential needs of such open spaces particularly in big citieswith limitations in land availability. Figure 1 shows a crosssection through a frame buildingwith soft storey. These typesof buildings are generally designed considering walls as non-structural elements without regard to the masonry infill wallaction.

Some of the modern seismic codes and the conventionalpractices as well neglect the effect of masonry infill wallbased on the assumption that itmay lead to some conservativeresults [1]. Some other codes (e.g. IS 1893–2002 [2], IBC-2003 [3]) provide a factor to magnify the induced straining

123

Arab J Sci Eng

Fig. 1 Cross section through aframe building with soft storey

actions in terms of bending moments and shear forces. Inorder to propose a magnification factor for the induced shearat base for a building with soft storey, Scarlet [4] performedan analysis based on two extreme situations inwhich uniformstructures as well as rigid structures with soft storey havebeen used.

Most of structural design practices in Egypt treat masonryinfill walls as non-structural elements, and consequently thecontribution from stiffness and strength of such elements tothe building is neglected during the analysis. Actually, thepresence of such infill walls significantly changes the frameaction behaviour and results in changing lateral load transfermechanism.

A building with an open storey or sometimes called softstorey is the one that has a stiffness discontinuity due to thesignificant flexibility of the open storey compared with theadjacent storeys. Several codes defined the stiffness disconti-nuity in a building storey as the one with lateral stiffness lessthan 70% of the lateral stiffness of the storey above or lessthan 80% of the average stiffness of the three storeys above[3,5,6].

Dolsek and Fajfar [7] attempted to explain the reasonbehind occurrence of soft storey effect in uniformly infilledframes as well as when this phenomenon occurs. Structuralmodels designed according to the Eurocode 2008 togetherwith structures designed with limited strength and ductilityaccording to previous codes have been utilized to performthe analysis. In 2002, Demir and Sivri [8] studied the seis-mic response of reinforced concrete structures with differentconfigurations of masonry infill in order to show the effectsof non-structural masonry infill walls on the induced build-ing’s response. The results of the conducted elastic analysisdemonstrated that the presence of non-structural masonry

infill significantly modifies the overall seismic response ofthe studied framed building structures.

Performance of a number of configurations ofmulti-storeyreinforced concrete frame models as bare frame, masonryinfill andmasonry infill with soft storey at ground floor underearthquake loads has been investigated [9]. Kabir and Shadan[10] developed a finite element model of a 3D-panel buildingsystem using the ABAQUS software package to investigatethe effect of presence of a soft storey on seismic performanceof such building systems. Results verified numerically thatthe 3D-panel system has considerable resistance under theapplied ground motion records.

Several methods of analysis in terms of linear and non-linear have been utilized to deeply understand the behaviourof building structures with masonry infill actions. Hirde andGanga [11] employed the pushover analysis to discuss theseismic performance of a twenty storey reinforced concretebuildingwith soft storey located at different levels alongwitha soft storey at ground level. The conducted study indicatedthe formation of plastic hinges in columns at ground openstorey. From safe design point of view, this is not acceptablecriterion. Karwar and Londhe [12] conducted a comparativestudy in order to investigate the seismic response behaviourof reinforced concrete framedbuildingmodelswith andwith-out masonry infill action. The nonlinear static analysis hasbeen used to perform the response analysis in terms of shearat base, displacement and the performance point. Setia andSharma [13] employed the equivalent static analysis to per-form response evaluation of reinforced concrete buildingswith soft storey. Five different models with shear wall in x-direction as well as in z-direction used in the analysis anddeveloped by the structural software package namely STAAPro.

123

Arab J Sci Eng

Fig. 2 Typical plan floor of sixbays four bays thirteen storeysframe building

The effect of seismic level on the response of masonryinfilled structures subjected to severe earthquake records hasbeen investigated [14]. In addition, the effect of existence ofa soft storey on the design strategy has also been consideredin the analysis. Agrawal [15] analysed the performance ofmasonry infilled building structures with and without open-ings. The effect of variation of opening percentage on thelateral stiffness of infilled building model has been analysedas well. A trial to investigate the seismic response of base-isolated building with soft storey has been carried out byPinarbasi and Konstantinidis [16]. The effect of soft storeyflexibility on the corresponding building with fixed-base hasalso been investigated and compared with the isolated-basebuilding case.

In this paper, the dynamic response time-history of rein-forced concrete moment-resisting frame buildings to near-fault records of El Centro (1940) and far-fault records ofLoma Prieta (1989) has been considered. Several buildingmodels, including fully infill frame model and frame mod-els with infill panels and soft storey at base level, 3rd storeylevel, 6th storey level, 9th storey level and 12th level, havebeen developed for analysis purposes. Since buildings with-out inclusion of masonry infill action can behave differentlythan buildings with inclusion of such action, a bare framebuilding model has also been considered in the analysis.

2 Building Models

In order to seismically investigate frame buildings with-out and with fully infill walls as well as frame buildingswith open soft storeys, a twelve storey reinforced concretemoment-resisting frame building is considered. The consid-ered building has a width of 16 m divided into 4 bays andlength of 36 m divided into six bays as well (see Fig. 2). Theassociated storey height considered is of 3 m.

Table 1 Dimensions and reinforcement of building elements

Structural element Dimensions (mm) Reinforcement

Beams 300 × 600 6 � 16

Columns 300 × 900 4 � 25 + 24 � 22

Different building models have been developed in orderto meet the cases considered in the study. Bare frame model,fully infill walls model and partially infill walls models dueto the existence of soft storeys at different levels have beencreated. Due to the symmetrical view of the considered framemodel, the effect of torsional response has been avoided. Thedesigned reinforced concrete horizontal elements in terms ofbeams have been set to be of 300mm× 600mm. The verticalelements in terms of columns have been found to be of crosssections 300mm× 900mmwithout reduction in dimensionsthroughout the building height. Table 1 presents dimensionsand reinforcement details of the designed building elements.The columns orientation as can be seen in the Fig. 2. ETABSsoftware package is used to perform the dynamic analysisfollowing the Egyptian Code for loads (Figs. 3, 4).

3 Modelling of Masonry Infill Walls

Two methods have been proposed in order to properly sim-ulate the behaviour of masonry infill walls, namely themicro-model method (see for example, Ref. [17]) and themacro-model method which has been introduced in 1960 byPolyakov [18]. Although the micro-model method is pro-ducing the better results and can be used for understandinglocal and global response, it is rarely used due to its com-plexity in generating the model and the computational costs.Themacro-modelmethod, also called the equivalent diagonalstrutmethod, has been developed to study the global response

123

Arab J Sci Eng

Fig. 3 Frame building models

123

Arab J Sci Eng

Fig. 4 Equivalent diagonal compressive strut action

of masonry infill frame buildings. The main disadvantageof the equivalent diagonal strut method is the deficiency inmodelling the openings accurately. However, there are someadvances in considering openings in walls where some num-ber of struts can be used in order to accommodate the effectof openings [19]. In the current study, walls are modelledas panel elements without any opening. Requirements ofFEMA 356 [20] will be followed to model the masonry infillwalls. According to FEMA 356, masonry infill walls prior tocracking is modelled with an equivalent diagonal compres-sion strut of width a. The thickness and modulus of elasticityof the strut are same as those of the represented infill panel.

The thickness of the strut can be written in terms of thecolumn height hcol between centrelines of beams and thelength of panel L as:

a = 0.175(λ1 hcol)−0.4rinf (1)

where the value of diagonal length of infill panel rinf can becalculated according to Eq. (2)

rinf√

(L inf)2 + (hinf)2 (2)

The Coefficient λ1 which is used to determine equivalentwidth of infill strut can be calculated as a function of the infillpanel height hinf , moduli of elasticity of both framematerialsEfe andmaterial of infill panel Eme, columnsmoment of iner-tia Icol, infill panel length L inf and thickness tinf , accordingto Eq. (3):

λ1 =[Eme tinf sin 2φ◦

4Efe Icol hinf

] 14

(3)

4 Time-History Analysis Method

Equivalent static force method, as a representative to linearstatic analysis, is the simplest technique for performing lineardynamic analysis. This simple method requires less compu-tational efforts and follows formulations given in the codesof practice. However, it is applicable for specific types ofbuilding structures with regular shapes and limited heightsas well (see Ref. [21]). In addition, response spectrum analy-sis as a linear dynamic method is quite accurate than theequivalent static one [22]. The time-history analysis, as anonlinear dynamic analysis, is the best technique to evaluatestructural response under earthquake excitations describedby ground acceleration records. Dynamic earthquake loadsincrementally affect the structure with time intervals�t , andthe governing equations of motion are solved using a step-by-step integration procedure which is the most powerfultechnique for nonlinear analysis. The response is evaluatedfor a series of short time increment. The general equation ofmotion is:

MU (t) + CU (t) + KU (t) = −MUg(t) (4)

WhereM, C andK are themass, damping and stiffnessmatri-ces, respectively. The symbols U, U and U , respectively,denote displacement, velocity and accelerations vectors. Ug

is the ground acceleration vector. The aforementioned vec-tors and matrices can be calculated for one dimensionalelement by defining the proper interpolation function [23].The equation of motion stated in Eq. (4) can be expressed inthe incremental form as:

M�U (t) + C(t)�U (t) + K (t)�U (t) = �P(t) (5)

where �U (t),�U (t) and �U (t), respectively, denote theincremental vectors of displacements, velocities and acceler-ations. The vector�P(t) is the incremental vector of externalearthquake load.

Although the time intervals should be short enough to givean accurate representation of such a rapid varying functionof time at the conventional methods, the method proposed byChen and Robinson [24] removed the limitation on the size

123

Arab J Sci Eng

Fig. 5 a El Centro and b LomaPrieta time history records

0 5 10 15 20 25 30 35 40 45 50-0.4

-0.2

0

0.2

0.4

Time (sec)A

ccel

erat

ion

(g)

0 5 10 15 20 25 30 35 40-0.6

-0.4

-0.2

0

0.2

0.4

Time (sec)

Acc

eler

atio

n (g

)

(a)

(b)

of time intervals and allows longer time intervals. Anothertechniques for solving incrementally the equations of motionare based on the explicit and implicit Runge–Kutta meth-ods and can be found in Refs. [25,26]. Two different groundexcitations have been selected to perform the dynamic analy-sis of the current study. One of these two records has beentaken from the near-fault region of El Centro (1940) with sitesource distance of about 8km. The second has been takenfrom the far-fault regions of Loma Prieta (1989) with sitesource distance of about 22km. Figure 5 provides the accel-eration time histories for the two earthquake ground motionsused in the current analysis. The ground motions records areobtained from the PEER Strong Motion Database (http://peer.berkeley.edu/smcat/).

4.1 Nonlinear Analysis

Unfortunately, the linear elastic analysis cannot providethe structural designer with the complete picture about theperformance of the structure when subjected to a groundmotion record. Nonlinear analysis in terms of material andgeometrical nonlinearities is considered as a reliable struc-tural analysis capable of simulating the proper behaviourof the material and the deformation of structural elementsunder the applied dynamic loads. When the materials movewithin the yield strength limits, then the behaviour of suchmaterials follows a linear trend. However, for the case thematerials exceed the elastic limit or the yield strength, per-

manent deformations, cracks, beam rotations and energydissipations in the form of inelastic and strain energy occur.Geometric nonlinearities refer to nonlinearities in kinematicquantities such as the strain-displacement relations in solids.Large deformations usually result in nonlinear strain- andcurvature-displacement relations. All equilibrium equationsare written in the deformed configuration of the structure.This may require a large amount of iteration. Althoughlarge displacement and large rotation effects are modelled,all strains are assumed to be small. The lateral deforma-tions are more pronounced under dynamic loads. In thegeometric nonlinearity, as the deflection of structural ele-ment gets increase the element starts to lose its stability.Due to the P − � effect, the applied force follows thedeformed member and creates further more instability veryquickly. In the structural computer program, some infor-mation in terms of stress–strain curve for concrete andsteel and the limit states has to be added to the computermodel in order to perform nonlinear analysis. In additionthe P − � effect for the vertical elements has to be definedas well.

The performed nonlinear TH analyses herein employedTakeda hysteretic model. This model is considered as oneof the best models that follow hysteretic rules for describ-ing the nonlinear relation between the applied force and thecorresponding deformation of the structural members. Theschematic representation of the Force–displacement relation-ships of Takeda hysteretic model is presented in Fig. 6a.

123

Arab J Sci Eng

Fig. 6 Force–deformationrelationship, a schematicrepresentation, b developed inETABS

Table 2 Shows the fundamental natural periods calculated by the generalized stiffness method and the corresponding ones calculated by dynamicanalysis using ETABS

Model no ETABS dynamic analysis Generalized stiffness method

Longitudinal direction Transverse direction Longitudinal direction Transverse direction

Fundamental natural period (s)

I 2.064 2.17 2.34 2.21

II 0.852 0.88 1.00 1.05

III 0.946 1.02 0.89 0.94

IV 0.98 0.98 1.05 1.09

V 0.939 0.944 0.99 1.03

VI 0.889 0.91 0.94 0.98

VII 0.85 0.89 0.90 0.96

By the way, in ETABS structural package, this model is thedefault one for the structural elements of the building model(see Fig. 1b). The computer model starts with the initial stiff-ness k0 of the building (see Fig. 6), and then the model isloaded incrementally till reaching the linearity limits. As thebuildingmodel exceeds the elastic limits and reaches the postyield stiffness kp, it hits the nonlinear zone and the iterationprocess starts to calculate strains, deflections and stiffness.Umax and Uy define the peak and yield displacements of theconcrete elements. Here fy is the yield force.

5 Model Validation

In order to validate the results of the developed models, thegeneralized stiffness method has been utilized [27]. The gen-eralized stiffness method mainly depends on the equivalentlamped mass model. The method has been employed to cal-culate the fundamental natural periods and then comparewiththose values calculated by developed models using ETABS.Due to space limitations, only Table 2 with the calculated

results from both dynamic analysis and the generalized stiff-ness method are presented. Amore detailed review on how toapply the generalized stiffnessmethod to calculate the funda-mental natural period can be found in Ref. [27]. As it can beseen from the presented results in Table 2, the calculated val-ues employing the generalized stiffness method show slightdifference in comparable with those obtained employing theETABS dynamic analysis.

6 Numerical Results and Discussion

Time-history analysis provides the structural response of theconsidered building models over time during and after theapplication of the seismic load. A twelve storey reinforcedconcrete framed buildingmodelled as (i) bare frame (i.e. con-sidering masonry walls as non-structural elements), (ii) fullyinfill frame building with masonry wall considered as struc-tural elements and (iii) partially infill frame building withsoft or open storey located separately at base, 3rd, 6th, 9th,and 12th floor levels is considered for time-history analysis

123

Arab J Sci Eng

under two real ground motion records. El Centro as a near-fault motion has been selected to perform the analysis. Thefar-fault records namely LomaPrieta have been also selected.In order to model the masonry infill action, the most widelyused single-strut model is employed where it is simple andevidently most suitable for large structures. Properties of theused infill materials in terms of modulus of elasticity, unitweight and poisons ratio are 5500 MPa, 20.0kN/m3, 0.15,respectively. It is worth noting that masses of infill wallshave been considered during the dynamic analysis of all thedeveloped models. The dynamic time-history analysis of thebuilding models considered is performed using the dynamic

analysis software ETABS. The seismic loads produced by thestructural package ETABS correspond to the data records ofthe El Centro and Loma Prieta earthquakes with peak groundaccelerations of 0.34 and 0.48g, respectively. A dampingratio of 5% has been associated for all the models during theanalysis. The dynamic analysis software ETABS enables theuser to apply the ground excitations separately in two orthog-onal directions. Storey shear forces and storey momentswhich are considered as the most useful responses used forearthquake resistant design strategy are obtained along theheight of the building models and presented in a compara-tive way for all the developed models. Storey displacements

Fig. 7 Induced storey shearforces under the El Centroearthquake records for ax-direction loading, by-direction loading

123

Arab J Sci Eng

which is a measure of the building deflection are also pre-sented following the samemanner. The predicted storey driftswhich can be defined as the measured displacement betweentwo consecutive stories normalized by storey height are pre-sented as well. Figures 6 through 13 present the obtainedresults for the aforementioned responses under the El Centroand Loma Prieta records for both x and y-directions.

Distribution of storey shear forces due to the appliedlateral load patterns is presented in Figs. 6 and 7 for theconsidered building models under El Centro and Loma Pri-eta groundmotion records applied in both x and y-directions,respectively.Theplotted curves clearly showa significant dif-ference between the cases of consideringmasonry infill wallsand the case of bare frame in which modelling of masonryinfill is ignored. As shown in these figures, storey shearresults of bare framemodel show the lowest values among all

othermodels considered in this study.Thedifferencebetweenstorey shear values for the bare frame case and the other casesof masonry infill is highly pronounced at lower storeys underthe El Centro and Loma Prieta earthquake records. However,for themodelswithmasonry infill and soft storeys assigned atdifferent levels, the captured storey shear forces show slightdifferences along the height of themodels under theElCentroearthquake for both x and y-directions of loading (see Fig. 7).However, under the Loma Prieta records, the soft storey levelaffects the induced storey shear values especially at lowerstoreys. Regardless the direction of loading and under thenear-fault motion El Centro, it has also been noticed that themaximum shear at base is associated with the masonry infillmodel with soft storey at bottom level (see Fig. 7). For thefar-fault motion Loma Prieta, the induced maximum shear atbase has been assigned for the masonry infill model with soft

Fig. 8 Induced storey shearforces under the Loma Prietaearthquake records for ax-direction loading, by-direction loading

123

Arab J Sci Eng

storey at top (see Fig. 7). Since earthquake resistant designconsiders the shear at base as a governing parameter, theignorance of masonry infill action underestimates the valuesof shear at bases andmay lead to unsafe design. For both bareand infill frames, the generated storey shear forces under theEl Centro earthquake records for both x and y-directions arein general larger than those induced under the Loma Prietarecords for the same loading directions (compare Figs. 7 and8). Masonry infill action magnifies the storey shear valueswith about 2.5 and 1.5 times as compared to bare frame caseunder the considered El Centro and Loma Prieta earthquakemotions, respectively. The obvious explanation for this canbe due to the El Centro records are characterized as near-fault

records, while the Loma Prieta records are of far-fault char-acteristics. From the plotted curves for storey shear forcesunder the El Centro records, it has been noticed that the exis-tence of a soft storey at the lowest level controls the outerrange of the calculated storey shear forces in both x andy-directions. More specifically, with the application of near-fault records of the El Centro earthquake load in x-direction,the infill model with soft storey at base has been found toinduce peak positive and negative values of 14,444kN and−15329 kN, respectively. Similar to x-direction, the infillmodel with soft storey at base produced outer range val-ues of 14,584kN and 15061 kN during the application ofEl Centro records in y-direction. Contrary to the near-fault

Fig. 9 Induced storeydisplacements under the ElCentro earthquake records for ax-direction loading, by-direction loading

123

Arab J Sci Eng

records, the application of far-fault motion to excite the mod-els clearly indicates that the existence of soft storey at higherlevels produced peak values of shear forces at base controlthe outer range of the plotted curves. Under the applied far-fault records of Loma Prieta earthquake in x-direction, it hasbeen found that the infill model with soft storey at the ninthfloor induces peak positive and negative shear values at baseof 9386 and −10448kN respectively. These captured valuesrepresent the range of the plotted curves of storey shear inx-direction. With the application of the earthquake load iny-direction, it has been found that the infill model with softstorey at the twelfth floor produces the peak positive and neg-ative values of 8490 and −8894kN, respectively, to form theouter range of the plotted curves in y-direction. It is worth

noting that the bare frame model produced the inner rangeof storey shear values regardless the direction of loading aswell as the type of the applied ground motion records eithernear-fault or far-fault records.

Peak displacement patterns of the 12-storey bare framebuilding model and fully infill building model as well asthe building model with soft storeys at different levels underthe El Centro and the Loma Prieta earthquake records arepresented in Figs. 9 and 10, respectively. The two earth-quake records are applied in two orthogonal directions. Asshown in this figures, the existence of soft storey causesa sudden change in the obtained peak displacements. Thisabrupt change leads to an increase in storey displacementsjust after passing the soft storey level which is highly pro-

Fig. 10 Induced storeydisplacements under the LomaPrieta earthquake records for ax-direction loading, by-direction loading

123

Arab J Sci Eng

nounced under the El Centro records. The bare frame modelproduces higher peak storey displacements as compared tothe masonry infill building frame models without and withsoft storeys under both the El Centro and Loma Prieta earth-quakes. This can be due to infill frame building systems withand without soft storeys have higher stiffness than the bareframe buildingmodel under the applied dynamic lateral load.This added stiffness to the infill system is due to the pres-ence of masonry infill walls. In contrast, the infill wall modelwithout soft storey produces the lowest storey displacementsunder both the El Centro and Loma Prieta records for theconsidered two directions of loading. If the induced lateral

deflections due to the existence of soft storey become toolarge, the resulting P −� effect may lead to an instability tothe building structure and potentially results in collapse. Forthe masonry infill model without and with soft storeys, slightdifferences between the induced peak storey displacementscan be seen from the plotted curves under the Loma Prietaground motion records applied in both x and y-directions. Ingeneral, at the upper storeys of the structural models con-sidered and under the two earthquake records in terms ofEl Centro and Loma Prieta, greater lateral displacements areallocated as compared to the induced deflections of the lowerstoreys regardless the loading direction.

Fig. 11 Induced storeymoments under the El Centroearthquake records for ax-direction loading, by-direction loading

123

Arab J Sci Eng

Results of the storey moment patterns for the bare framemodel, model with infill walls and models with soft storeysat 3rd, 6th, 9th and 12th floor levels under El Centro groundmotion records are presented in Fig. 11 considering twodirections of loading. The corresponding results under theLoma Prieta earthquake records are presented in Fig. 12.Comparing Figs. 11 with 12, it can be noticed that theinduced peak moments under the Loma Prieta earthquakevary significantly as the level of the soft storey changesespecially at lower storeys. However, under the El CentroEarthquake records, the variation in the obtained peak storeymoments seems to be insignificant and shows slight changewith the variation of the soft storey level. Under both excita-tion records and irrespective the loading direction, the bare

frame building model induces the lower storey moments.Contrary to the bare frame model, the full masonry infillbuilding model induces storey moments of higher values ascompared to the other models. This can be seen clearly whenusing Loma Prieta earthquake to excite the building models.Under the El Centro earthquake records, the value of inducedpeak moment considering x-direction of loading shows goodagreement with the one obtained considering y-direction ofloading. However, under the Loma Prieta earthquake loadapplied in x-direction, the peak moment value obtained ishigher than the one obtained one in y-direction. Conse-quently, a magnification factor for the moments has to beconsidered in the design stages. The reason behind the higherstoreymoments in case of fully infilledmodels in comparison

Fig. 12 Induced storeymoments under the Loma Prietaearthquake records for ax-direction loading, by-direction loading

123

Arab J Sci Eng

to the other models can be due to the increase in stiffness ofthe fully infill model which is assumed to be the summationof the bare frame model stiffness and the infill wall stiffness.The increase in stiffness leads to an increase in the inducedstraining action in terms of storey moments.

Figures 13 and 14 show the results of maximum storeydrift ratios of 12-storey structure under the El Centro andLoma Prieta ground motion records. These obtained resultsdemonstrate the differences among the drift profiles of thebuilding structuremodelled as bare frame, fully infilledbuild-ing model and infilled building models with soft storeys.As it can be seen from the figures, the bare frame building

model has drift ratios of higher values than those associ-ated with the considered fully infill frame building modelunder the near-fault El Centro records and the far-faultLoma Prieta records. In addition and as it can be seen fromthe figures, the captured values of the drift ratios at thespecified soft storeys of the infill frame building modelsexceed those values obtained considering the building struc-ture modelled as bare frame under the applied two groundmotions. However, this increase in drift ratio can be morepronounced under near-fault El Centro records as comparedto the far-fault LomaPrieta records regardless the direction ofloading.

Fig. 13 Induced storey driftsunder the El Centro earthquakerecords for a x-directionloading, b y-direction loading

123

Arab J Sci Eng

Compared to the fully infilled frame building model, theinfill frame models with soft storeys have sudden increaseat the specified soft storey level. This observed trend hasbeen noticed for all the considered soft storey models underthe two ground motion records considered. This can be dueto the building frame models with soft storeys suffer fromstrength and stiffness discontinuities of those above andbelow storeys.

The unexpected movements of building structure underlateral seismic actions due to soft storeys can be highlysignificant. These may lead to pounding between adjacent

buildings or structurally separate portions of the same build-ing that do not have adequate separation distance in between.These collisions between structural elements can result inincrease in floor acceleration as well as significant localizeddamage between these structural elements.

From seismic design point of view, these excessivelateral movements and drifts due to existence of a softstorey can significantly affect the structural elements evenif these elements are part of the lateral forces resistingsystem.

Fig. 14 Induced storey driftsunder the Loma Prietaearthquake records for ax-direction loading, by-direction loading

123

Arab J Sci Eng

7 Conclusions

The current research study has been carried out on reinforcedconcrete framed buildings fully as well as partially infilledunder seismic loads. Dynamic time-history analysis has beenperformed employing two groundmotions from near and far-fault regions. The influence of infillwall action on the seismicperformance of the developed structural models with softstorey has been investigated. The effect of variation of softstorey level has been studied as well. The following resultssummarize the main findings of the considered different sce-narios of the structural models.

1. The masonry infill action has a significant influence onthe global performance of the building structure wherethe induced structural responses for bare frame case dosignificantly vary with the different configurations asso-ciated with fully or partially masonry infill walls undereither near-fault or far-fault earthquake loads.

2. Considering masonry infill action reduces the inducedstorey displacements as compared to the bare frame case.However, the induced storey moments and storey shearforces increase with the incorporation of masonry infillaction.

3. Masonry infill walls enhance the seismic performance ofthe building structure during earthquake excitations interms of displacement control, storey drifts and lateralstiffness.

4. The level of soft storeyhas a significant role on the inducedstorey shear forces under the far-fault Loma Prieta earth-quake especially at lower storeys. However, under ElCentro earthquake, the level of soft storey seems to beslightly significant.

5. Contrary to the induced storey shear forces, the inducedstorey displacements and moments are significantlyaffected by the variations of soft storey level undernear-fault motion. However, under far-fault motion, theseresponses seem to be unaffected with the variation of thesoft storey level.

6. Compared to the fully infilled frame building model, theinfill framemodels with soft storeys have sudden increasein the obtained responses at the specified soft storey levelsregardless direction of loading and the type of the appliedearthquake records as well.

7. Although themasonry infill action decreases the values ofinduced storey drift as compared to the bare frame case,the existence of a soft storey at a specified level highlymagnifies storey drift at that levelwith values exceed thoseassociated with the bare frame case.

8. TheNational Building Codes should provide amagnifica-tion factor for the storey response in terms of storey shearforces and overturning moments. On the other hand, in

case of ignoring masonry infill actions, a reduction factorfor storey displacements should be provided.

References

1. Fardis, M.N.; Panagiotakos, T.B.: Seismic design and response ofbare and infilled reinforced concrete buildings. Part I: bare struc-tures. J. Earthq. Eng. 1(1), 219–256 (1997)

2. IS 1893 (Part I): Criteria for Earthquake Design of Structures—Part I: General Provisions and Buildings (5th revision). Bureau ofIndian Standards, New Delhi (2002)

3. IBC, “International Building Code”, USA: International CodeCouncil Inc. Country Club Hills, Ill: ICC (2003)

4. Scarlet, A.: Design of soft stories—a simplified energy approach.Earthq. Spectra 13, 305–315 (1997)

5. UBC, “Uniform Building Code”, USA: International Code Coun-cil, Inc. Falls Church, Virginia (1997)

6. ASCE, “American Society of Civil Engineers” 1801 AlexanderBell Dr Reston, Virginia (2006)

7. Dolsek, M.P.; Fajfar, P.: Soft storey effect in uniformly infilledreinforced concrete frames. J. Earthq. Eng. 5(1), 1–12 (2001)

8. Demir, F.; Sivri, M.: Earthquake Response of Masonry InfilledFrames. In: ECAS2002 International Symposiumon Structural andEarthquake Engineering, METU, Ankara (2002)

9. Mulgund, G.V.; Patil, D.M.; Murnal, P.B.; Kulkarni, A.B.: Seismicassessment of masonry infill RC framed building with soft groundstorey. In: International Conference on Sustainable Built Environ-ment (ICSBE-2010) Kandy, 13–14 December (2010)

10. Kabir, M.Z.; Shadan P.: Seismic performance of 3D-panel wall onPiloti RC frame using shaking table equipment. In: Proceedings ofthe 8th International Conference on Structural Dynamics, EURO-DYN 2011 Leuven, Belgium, 4–6 July 201(2011)

11. Hirde S.; Ganga T.: Seismic performance of multistory buildingwith soft storey at different level with RC shear wall. In: Inter-national Journal of Current Engineering and Technology E-ISSN2277-4106, P-ISSN 2347–5161 (2014)

12. Karwar, D.B; Londhe, R.S.: Performance of RC framed structureby using pushover analysis. Int. J. Emerg. Technol. Adv. Eng. 4(6),488–491 (2014)

13. Setia, S.; Sharma, V.: Seismic response of R.C.C building with softstorey. Int. J. Appl. Eng. Res. ISSN 0973-4562, 7 (11) (2012)

14. Amit, S.G.: Seismic analysis of frame with soft ground storey. Int.J. Pure Appl. Res. Eng. Technol 1(8), 213–223 (2013)

15. Agrawal, N.: Analysis of masonry infilled RC frame with & with-out opening including soft storey by using “Equivalent DiagonalStrut Method”. International Journal of Scientific and ResearchPublications, ISSN 2250-3153, 3(9), (2013)

16. Pinarbasi, S; Konstantinidis, D.: Seimic isolation for soft storeybuildings. In: 10thWorld Conference on Seismic Isolation, EnergyDissipation and Active Vibrations Control of Structures, Istanbul,28–31 May (2007)

17. Morbiducci, R.: Nonlinear parameters of models for masonry. Int.J. Solids Struct. 40(15), 4071–4090 (2003)

18. Polyakov, S.V.: On the interaction betweenmasonry filler walls andenclosing frame when loaded in the plane of the wall. In: Transla-tion in Earthquake Engineering. Earthquake Engineering ResearchInstitute (EERI), San Francisco, pp. 36–42 (1960)

19. Asteris, P.: Lateral stiffness of brick masonry infilled plane frames.Struct. Eng. 129, 1071–1080 (2003)

20. FEMA-306: Evaluation of Earthquake Damaged Concrete andMasonry Wall buildings—Basic Procedures Manual. FederalEmergency Management Agency (1999)

123

Arab J Sci Eng

21. Raheem, Abdel: Evaluation of egyptian code provisions for seis-mic design of moment-resisting-frame multi-story buildings. Int.J. Adv. Struct. Eng. 5, 1–18 (2013)

22. Mahmoud, S.; Abdallah,W.: Response analysis of multi-storey RCbuildings under equivalent static and dynamic loads according toEgyptian code. Int. J. Civil Struct. Eng. Res. 2(1), 79–88 (2014)

23. Clough, R.W.; Penzien, J.: Dynamics of Structures, 3rd edn. Com-puters and Structures, Berkely (2003)

24. Chen, C.C.; Robinson, A.R.: Improved time history analysis ofstructural dynamics. I: treatment of rapid variation of excitation andmaterial nonlinearity. J. Eng. Mech. 119(12), 2496–2513 (1993)

25. Chen, X.; Mahmoud, S.: Implicit Runge-Kutta methods for lip-schitz continuous ordinary differential equations. SIAM J. Numer.Anal. 46, 1266–1280 (2008)

26. Mahmoud, S.; Chen, X.: A verified inexact implicit Runge–Kuttamethod for nonsmooth ODEs. Numer. Algorithms 47, 275–290(2008)

27. Chopra, A.K.: Dynamics of Structures: Theory andApplications inEarthquake Engineering. Prentice-Hall, Englewood Cliffs (1995)

123