SEISMIC RESPONSE OF RC FRAME
BUILDINGS WITH SOFT FIRST STOREYS
JASWANT N. ARLEKAR SUDHIR K. JAIN C.V.R. MURTY.
Research Associate Professor Assistant Professor
Department of Civil Engineering, Indian Institute of Technology Kanpur,
Kanpur 208 016
‘Tel. nos. : (0512) 597867, 598367, Fax : (0512) 597995, 590260, 590007
ABSTRACT
Open first storey is a typical feature in the morden multistorey constructions in urban India, Such
features are highly undesirable in buildings built in seismically active areas. The Jabalpur
earthquake of 22 May 1997 has illustrated the potential hazards associated with buildings having
open first storey. In this paper, a parametric study is performed on an example building with
open first storey to bring out the importance of explicitly recognizing the presence of the soft
first storey in the analysis. It is observed that most of the Indian buildings with open first storey
fall under the category of soft first storey buildings as per the draft Indian seismic code, The error
involved in modeling such buildings as complete bare frames, neglecting the stiffness of infills in
the upper storeys, is brought out through this study. Altemate measures, namely increasing the
column size in the open first storey and introduction of a concrete core, are proposed to reduce
the stiffness irregularity introduced by the open first storey. However, it is observed that the soil
flexibility can significantly reduce the effectiveness of the concrete core. Hence, under flexible
soil conditions, the analytical model of buildings with shear walls should include the foundation
flexibility.
INTRODUCTION
Many urban multistorey buildings in India today have open first storey as an unavoidable feature.
This is primarily being adopted to accommodate parking or reception lobbies in the first storeys.
‘The upper storeys have brick infilled wall panels. The draft Indian seismic code classifies a soft
storey as one whose lateral stiffness is less than 50% of the storey above or below {Murty and
Jain, 1996). Interestingly, this classification renders most Indian buildings, with no masonry
infill walls in the first storey, to be “buildings with soft first storey.
“Whereas the total seismic base shear as experienced by a building during an earthquake is
dependent on its natural period, the seismic force distribution is dependent on the distribution of
13.stiffness and mass along the height. In buildings with soft first storey, the inter-storey drift in the
soft first storey is large, The strength demands on the columns in the first storey for these
buildings is also large, as the shear in the first storey is maximum. For the upper storeys,
however, the forces in the columns are effectively reduced due to the presence of very stiff brick
inf! walls which share the forces. Further, the force distribution in the soft drst storey buildings
may not be parabolic as given in 1S:1893-1984, due to the uneven distribution of stifiness along,
the height.
Many earthquakes in the past, have demonstrated the potential hazard associated with soft first
storey buildings. Major damage to many reinforced concrete and steel buildings in the Kobe
apan) earthquake of January 17, 1995 (AU, 1995], and to critical hospita! facilities in the San
Fermando (California, USA) earthquake of 1971, were attributed to the soft fist storey. Alarming
amount of damage to the buildings with open basements for parking has beer reported during the
Northridge (California, USA) earthquake of January 17, 1994 [Hall, 1994; EQEI, 1994).
‘The recent Jabalpur earthquake of 22 May 1997, in the state of Madhya Pradesh in central Inc
also illustrated the poor performance of Indian buildings with soft first storey [Jain et al, 1997].
‘The magnitude of this earthquake was 6.0 on the Richter scale. The intensity of shaking in the
affected areas of Jabalpur, which are referred to in this paper was about VII on the MSK scale.
This earthquake, the first one in the recent years iri an urban neighborhood in India, provided an
opportunity to assess the performance of engineered buildings in the country during ground
shaking, Performance of the Himgiri, Ajanta and Nalanda apartments in the city of Jabalpur are
‘very good examples of the inherent risk involved in the construction of buildings with soft first
storey. The Himgiri apartments is a five-storey RC frame building with shops on the two
adjacent faces at the first storey with brick masonry partition walls (Fig.1). The remaining comer
14of the building is meant for parking. Upper storeys of the building have residential apartments
with a large number of brick masonry infills. The first storey columns in the parking area were
badly damaged. Damage consisted of spalling of concrete cover, snapping of lateral ties,
buckling of longitudinal reinforcement bars and crushing of core concrete (Fig.2). The columns
in the shopping area on the same storey (having brick infills) had much lesser level of damage in
them, Damage in the upper storeys consisted of only minor cracks in the filler walls, This is a
clear case of columns damaged as a result of the “soft first storey.”
‘The Ajanta and Nalanda apartments buildings are two almost identical four-storey RC frame
buildings located side-by-side, In Nalanda apartments there are two apartments in each storey
including the first storey. Ajanta apartments has two apartments each in the upper storeys, but
Figure 2 :: Damage to columns in Himgiri apartment,
‘onfy one apartment in the first storey; the remaining space here is meant for parking. Whereas,
only nominal damages were observed in Nalanda apartments, the first storey columns of Ajanta
apartments were very badly damaged. The damage consisted of buckling of longitudinal bars,
snapping of ties, spalling of cover and crushing of core concrete.
Youth Hostel building in Jabalpur is a two-storey C-shaped RC frame building supported on stilt
columns on plies. The two storeys above the stilt storey have brick infilled wall panels. This
makes the upper storeys very stiff as compared to the storey at the stilt level. There was no
damage to the columns in the storeys above. However, extensive damage took place in the stilt
columns (Fig.3). The “soft first storey” at the stilt level is clearly the primary reason for such a
severe damage.
In this paper, a parametric study is performed on a building with soft first-storey. Various
alternative models of the same building are carefully examined for their empirical and analytical
15Figure 3 :: Damage to columns in the stilt storey of Youth Hostel building.
fundamental natural periods, as well as the stiffness of the open first storey in relation to the
stiffness of the upper storeys. The force resultants in the first storey columns are compared for
different models. It is seen that the drift and strength demands on the first-storey columns of such
buildings are very large in comparison to those in the upper storeys. Finally, some solutions to
reduce the hazards due to soft first storey are discussed, The influence of soil flexibility on the
recommended solution is also discussed.
BUILDING STUDIED
‘The plan layout of the reinforced concrete moment resisting frame building with open first storey
and unreinforced brick infill walls in the upper storeys, chosen for this study is shown in Fig.4.
‘The building is kept symmetric in both orthogonal directions in plan to avoid torsional response
under pure lateral forces, Further, the columns are taken to be square to keep the discussion
focused only on the soft first storey effect, without being distracted by the issues like orientation
of columns. The building is considered to be located in seismic zone II and intended for
residential use. The building is founded on medium strength soil through isolated footings (of
size 2mx2m) under the columns. When a central concrete service core is used, a 2m wide footing
is taken to go all around under the wall in the core. To show the effect of soil flexibility, the
modulus of subgrade reaction of the soil is taken as 30,000 kN/m’. Elastic moduli of concrete and
masonry are taken as 28,500 MPa and 3,500 MPa, respectively, and their Poison’s ratio is
0.2. Performance factor (K) has been taken as /.0 (assuming ductile detailing). The unit weights
of concrete and masonry are taken as 25.0 kN/m’ and 20.0 kW/m’. The floor finish on the floors is
1.0 kN/m?. The weathering course on roof is taken as 2.25 kN/m”. The live load on floor is taken
‘as 2.0 kN/m? and that on roof as 0.75 kN/m’. In the seismic weight calculations, only 25% of the
floor live load is considered.
169m (3 @ 3m cic}—»!
' 22.5m (5 @ 4.5m cle) >
Figure 4 :: Plan of the example building.
Nine different models of the building are studied (Fig.5). These are :
Model I
Modei II
Model [fT ::
Model IV ::
Model Vi
Model VE:
Model VII ::
Model VITE::
Model IX ::
1 Building has no walls in the first storey and one full brick infill masonry walls
(220 mm thick) in the upper storeys.
Building has no walls in the first storey and half brick infill masonry walls (110
‘mm thick) in the upper storeys.
Building modeled bare frame ignoring the stiffness contribution of walls.
However, masses of the walls as in mode! J are included in the model.
Building has one full brick infill masonry wall (220 mm thick) in all storeys,
including the first storey.
Building has one full brick infill masonry walls (220 mm thick) only in the upper
storeys. Further, a.central service core is introduced in the first storey, by providing
220 mm thick brick masonry walls within the frame panels formed by the columns
and beams in the central bay.
Building has one full brick infill masonry walls (220 mm thick) only in the upper
storeys, Again, a central service core is introduced in the building, by providing
reinforced concrete walls within the frame panels formed by the columns and
beams in the central service core in all storeys. Concrete wall thickness is taken as
250 mm in the first storey and 100 mm in the upper storeys.
Building has no walls in the first storey and one full brick infill masonry walls
(220 mm thick) in the upper storeys. However, the columns in the first storey are
much larger (850mmx850mm) than those in the upper storeys (400mm. 400mm) to
reduce the stiffness irregularity besween the open frst storey and the storey above.
Building as in Model VI with soil flexibility introduced only under the concrete
walls in the central core area; the columns are assumed to be fixed at their base.
‘The soil is assumed to be medium stiff
Building as'in Model VI with soil flexibility introduced under the concrete walls in
the central core area as well as under all the columns.
17Be
Sen
laf
‘Model J and I Model It ‘Model IV.
‘Model VI, VIII and IX ‘Model VII
Figure 5 :: Elevation of different building models (model I has one full brick thick walls, model
I has half brick thick walls. Model VI has concrete core on fixed foundation, model
VII considers soil flexibility under the concrete core only, and-model IX considers
Soil flexibility under all the foundations).
ANALYSIS OF THE BUILDING
Linear elastic analysis is performed for the nine models of the building using ETABS analysis
package (Habibullah, 1995). The frame members are modeled with rigid end zones, the Walls are
modeled 2s panel elements without any openings, and the floors are modeled as diaphragms rigid
in-plane. The soil flexibility is introduced as linear Winkler springs under the footing, When the
central service core is used in models VIII and IX, the walls in the core are discretised finely into
250 mm wide vertical strips to enable the modeling of a continuous soil support through linear
Winkler springs. Two different analyses are performed on each model, namely the equivalent
static analysis and the multi-modal dynamic analysis. These are briefly deseribed below.
Equivalent Static Analysis
‘The fundamental natural period of the building is calculated by the expression T= 0.09H/VD,
18given in 1S:1893-1984, wherein H is the height and D is the base dimension of the building in
meters, in the considered direction of vibration. Thus, the fundamental period for all the models
in this method, is the same. The lateral load calculation and its distribution along the height is
done as per the empirical equation in 1$:1893-1984.
Multi-Modal Dynamic Analysis
‘The natural period values are calculated by ETABS, by solving the eigen value problem of the
model. Thus, the total earthquake load generated and its distribution along the height correspond
to the mass and stiffness distribution as modeled by ETABS. Analysis is performed for 5%
damped design spectrum given in IS:1893-1984.
RESULTS AND DISCUSSIONS
Storey Stiffness
‘The storey stiffness of the first and second storeys in different models are shown in Table 1.
‘The storey stiffness is defined as the magnitude of the force couple required at the floor levels
adjoining the storey to produce a unit lateral translation within the storey, letting all the other
floors to move freely. The stiffness irregularity in building models with soft first storey is evident
Table 1 :: Storey stiffness of first and second storeys for different building models.
‘Storey Stiffness (kN/mm)
Building Model Transverse | Longitudinal
Direction Direction
First | Second | First | Second
Storey | Storey | Storey | Storey
te T_ Open First Storey: Bo; 3448] 227| 5263
220mm thick walls in upper storeys
TW Open First Storey:: 25, 2083] 220) 3030
110mm thick walls in upper storeys
I_Bare Frame ves | 365| 166 231
TV_Brick Infilled Completely Rial ssvi| 3871 | 5263
V_ Open First Storey 4743333] 694 [5000
= Brick Service Core
VI_ Open First Storey 7346 | 4349 a167 | 7143
= Concrete Service Core
Vil Open First Storey with Stiffer a1 | 3846/2778 | 5556
Columns
Vill Open First Storey
~ Concrete Service Core 300| 3125] 308) 4546
= Flexible Soil under Core only
TX Open First Storey 205; 1613 220] 2837
= Concrete Service Core
= Flexible Soil
19from the fact that the stiffness of the first storey for model J, IZ, and J” is about 59%,10%, and
15%, respectively, of the second storey stiffness. Models / and // represent the realistic situations
for buildings with open first storeys. It is seen that the reduction of the wall thickness in the
upper storeys (model 2) and addition of the brick service core (mode! 11’) reduces the stiffness
irregularity only marginally. The stiffness of the first storey in model ZI7 (bare frame) is about
55% of that of the second storey. The use of RC service core (model #f) or stiffer columns
(model VZ/) in the first. storey significantly reduces the stiffness irregularity. The first storey
stiffness in these models is more than 50% of the second storey stiffness. It is interesting to note
that a careful choice of concrete core (mode! VIZ) can bring the first storey stiffness close to that
of model with full brick infills in the first storey (model /V). However, results of models VII and
‘1X show that the foundation flexibility can significantly reduce the effectiveness of the concrete
core. Clearly, for buildings supported on soft or medium soils, one needs to choose the
foundation carefully from this view point.
Table
Empirical and analytical fundamental natural periods of different building models.
Fundamental Natural Period (sec)
Model | Transverse Direction | Longitudinal Direction
Empirical | Analysis | Empirical | Analysis |
027 0.42
0.27 038
z x 0.27 O71 |
(WI «. 0.27 0.15
E 0.27 0.26
. 0.27 0.13
Vir] 0.42 0.27 0.15
. 0.27 037
Lx 0.27 0.44
Fundamental Periods
‘Whe empirical (1S:1893-1984) and analytical (ETABS) natural periods of the building models are
shown in Table 2. It is seen that the analytical natural periods do not tally with the natural
periods obtained from the empirical expression of the code, The bare frame idealization in model
HT leads to a severe overestimation of the natural period compared to the (actual) open first
storey ‘building in mode! J. This leads to an underestimation of the design lateral force in model
Lateral Deformation
The lateral displacement profiles of the various models for the two different analysis are shown
in Fig.6. In these figures, the abrupt changes in the slope of the profile indicate the stiffness
inregularity. All displacement profiles corresponding to models having stiffness irregularity (ZI,
¥, VII and IX) have a sudden change of slope at first floor level. However, the other models i..,
IM, IV, VI and VIL, show sraooth displacement profiles. The displacements at first floor level are
shown in Table 3. The inter-storey drift demand is largest in the first storey for all the models
with soft ground storey. This implies that the ductility demand on the columns in the first storey,
20for these models, is the largest. For the models which do not have stiffness imegularity the first
floor displacement is small; approximately 10% of the corresponding values in model J. Thus,
the drift ductility demand in the first storey can be greatly reduced by ensuring that the storey
stiffness is atleast equal to 50% of the second storey.
VayIM Vo Viygy vm 1 .
= =
z z
3 z°
2 a
Transverse 2 Lorigitudinal
° ae
5 10 18 ° 5 10 15
‘Storey Drift (mm) Storey Drift (mm)
(a)
VaviNy, vnVENOC VIN VEY VELOC
2] 2
i 0 10
Es] 8
z
Be 6
=. ‘4
2 Transverse 2 Longitudinal
° ° =I
° 6 0 | tS ° 5 10 15
Storoy Drift (mm) ‘Storey Drift (mm)
@
Figure 6: Lateral Displacement Profile by (a) Equivalent Static Analysis and (b) Multi-Modal
Dynamic Analysis.
Bending Moment and Shear Force in Columns
The maximum bending and maximum shear forces in the columns in the first and the upper
storeys are shown in Table 3; the bending moment and shear force (strength) demands are
severely higher for first storey columns, in case of the soft first storey buildings.
The introduction of walls in the first storey (model /Y) reduces the force in the first storey
acolumns. As the force is distributed in proportion to the stiffness of the members, the force in the
columns of the upper storeys, for all the models (except mode! 1/7), are significantly reduced due
to the presence of brick walls. These forces (bending moment and shear force) are about 10-20%
of the corresponding values in the first storey columns. The use of brick service core is not very
effective in reducing the strength demand on the first storey columns. However, the force values
are around 5026 of the values in case of model J. When concrete service core is used, the demand
on the columns is significantly reduced (by a factor of about /0.0). Irterestingly, the drift
demands on the first storey columns in case of model JV (completely infilled) and model VI
(stiffer columns in first storey) are very close. This is true for strength demands also. Thus, a
careful choice of concrete core can effectively eliminate stiffness as well as strength irregularity
in open first storey buildings.
Table 3 :: Displacement at first floor, maximum forces in first storey columns and average of
the maximum forces in the columns of the storeys above for different models.
‘Displacement (rim) ‘Maximum Moment (kNm) ‘Maximum Shear (KN) |
Model ‘at First Floor Transverse | Longitudinal | Transverse | Longitudinal —
Direction Direction Direction Direction
Trans. | Long. | First | Upper | First | Upper | First | Upper | Furst | Upper
Direction | Direction | storey | storeys | storey | storeys | storey | storeys | storey | storeys
‘Equivalent Static Analysis
1] 27 31_ [5647 627 627] 487 2727 45) 301] 3.6)
if 2. 25_| 450] 82] so4[ 7.2[ 216] 60] 24i[ 5&
ml 39 52 | 704| 439[ 853 499] 317] 323[ 365] 375
[03 02 66{ 62, 457 647 31) 44] 21] 35
v[_ 14 io [27s] 54) 208] 47[ 1] 37[ 100, 33
vif 03 02 67] 82] 397 63, 34] 58] 19] 43
vif 02 03 | 762] 53) ear] 3.8] 3297 39] 349] 29
Vitt[ 2.1 23 [468] 15.6] 458] 68[ 233] 107| 220| 48
Ix] 33 32 [rosy sii] 7iy 346, 345] 218] 354] 247
‘Multi-Modal Dynamic Analysis
i] 23 24 [481] 45] 47] 47] 232] 34] 235] 32
uy 19 19 [380] 60; 440] 71] 183| 45[ 188] 48
im] 25 26 | 452/234] 39.1] 224] 204] 175[ 190] 176
[03 02 s8[ 47[ 36{ 39, 28 34] 18[ 29
vi 13 o9 [259] 40] i77[ 33] 124] 28[ 84] 23
vif 03 01 61] 71] 3af sip 31] 50, 16] 35
vil 02 02 [65s] 407 678] 29, 284] 29/282] 23
vin 1.8 1g [ars] way 37277 Saf ai] 97] 179 36
Ix] 26 24 | 57s] 28.6] 55.27 300] 285] 200[ 278[ 214
Foundation Flexibility
From the above discussion, it is seen that the concrete service core is an effective way to reduce
the stiffness irregularity. However, if the concrete core is supported on flexible foundation
system, it can substantially impair its effectiveness. Thus, it is important to incorporate the soil
22flexibility in the modeling of the buildings with shear walls, failing which the drift and strength
demands in the first storey columns can be under-estimated, resulting in an incorrect design of
the building,
CONCLUSIONS
RC frame buildings with open first storeys are known to perform poorly during strong
earthquake shaking, In this paper, the seismic vulnerability of buildings with soft first storey is
showa through an example building. The drift and strength demands in the first storey columns
are very large for buildings with soft ground storeys. It is not very easy to provide such
capacities in the columns of the first storey. Thus, itis clear that such buildings will exhibit poor
performance during a strong shaking. This hazardous feature of Indian RC frame buildings needs
to be recognized immediately, and necessary measures taken to improve the performance of the
buildings.
The open first storey is an important functional requirement of almost all the urban multi-storey
buildings, and hence, cannot be eliminated. Alternative measures can help overcome this specific
situation of a “soft first storey.” One approach of solving this problem is by increasing the
stiffness of the first storey such that the first storey is at least 50% as stiff as the second storey.
‘The possible schemes to achieve the above are (i) provision of stiffer columns in the first storey,
and (ii) provision of a concrete service core in the building, These measures have been found to
significantly reduce the lateral drift demand, and hence the stress resultants, on the first storey
columns. The soil flexibility needs to be examined carefully before finalizing the analytical
model of a building, Flexible soil conditions for buildings with shear walls may require altemate
solutions other than those described in this paper, to reduce seismic drift demands on the
columns in the first storey.
ACKNOWLEDGMENTS
‘The financial support provided by the Department of Science and Technology (DST), New
Delhi, for conducting this post-Jabalpur earthquake study, is gratefully acknowledged.
REFERENCES
AU, 1995, “Preliminary Reconnaissance Report of the 1995 Hyogoken-Nanbu Earthquake,
Architectural Institute of Japan, Tokyo, Japan.
EQEI, 1994, “The January 17, 1994 Northridge, California Earthquake - An EQE Report,” EQE
{,jtternational, San Francisco, USA.
Habibuilah,A.. (1995), ETABS - Three Dimensional Analysis of Building Systems, Users Manual,
Computers and Structures, Inc., Berkeley, CA, USA.
Hall,F., (Editor), 1994, “Northridge Earthquake January 17, 1994 - Preliminary Reconnaissance
Report," Report No.94-01, Earthquake Engineering Research Institute, Oakland, USA.
2318:1893-1984, Indian Standard Criteria for Earthquake Resistant Design of Structures, Bureau
of Indian Standards, New Delhi, 1984,
Jain.SK., Murty,C.V.R,, Arlekar,J.N., SinhaR., Goyal,A., and Jain,CK., (1997), “Some
Observations on Engineering Aspects of the Jabalpur Earthquake of 22 May 1997," EERI
Speciai Earthquake Report, EERY Newsletter, Vol.31, No.8, August 1997, pp 1-8.
Murty,C.V.R., and Jain,S.K., (1995), “A Proposed Draft for 1S:1893 Provisions on Seismic
Design of Buildings- Part I: Code,” Journal of Structural Engineering, SERC, Madras,
Vol.22, No.1, April 1995, pp 21-29.
24