THE EFFECT OF SOIL STRUCTURE INTERACTION ON SEISMIC

BEHAVIOUR OF MID AND HIGH-RISE BUILDINGS

Bayram Tanik CAYCI1, Zeynep Gokcen ICOZ2, Mehmet INEL3

ABSTRACT

This study aims to evaluate the effect of soil-structure interaction on seismic behavior of reinforced concrete (RC)
buildings considering nonlinear behavior of structural members. Two-dimensional 16-, 12- and 8-story reinforced
concrete (RC) frame buildings are used in the scope of the study. Two-dimensional SSI and fixed base nonlinear
models are compared for roof and displacement profiles to better understand behavioral differences of two
approaches. Four ground motion records and 2 soil types with different stiffness were taken into account during
nonlinear time history analyses. The results show that seismic demands tend to decrease for SSI models as soil
stiffness decreases. The effectiveness of SSI is related to height/width ratio of superstructures. While soil damping
is more predominant on seismic behavior of superstructures with increasing height-width (H/W) ratio, it is more
complex for structures with lower H/W ratio. The consideration of soil-structure interaction is also effective on
damage pattern of structures. The outcomes show that the interstory drifts at lower stories tend to increase for fixed
base models while the interstory drift ratios of SSI models increase at the upper stories due to soil damping at the
base of structure. The SSI effects are less remarkable for stiffer soil as expected. The dynamic amplification due
to frequency content of surface motion is more effective especially for structures with lower H/W ratios. The
consideration of SSI affects the damage of beam and column elements. The number of damaged members increases
when soil structure interaction is taken into account.

Keywords: Nonlinear dynamic analysis; Reinforced concrete buildings; Soil structure interaction

1. INTRODUCTION

The existence of soil has significant effect on the seismic behavior of structures. The dynamic response of
structure above the soil changes due to soil deformations and rotations at the base of foundation. The
existence of structure also affects the frequency content of motion. The soil-structure interaction is even more
significant for softer soil profiles with stiffer-taller superstructures. The recent destructive earthquakes
occurred in Turkey emphasize the importance of this phenomenon (Ozmen et al. 2013; Ozmen et al. 2014).
The fixed base assumption is very common due to easily generated structural models and less computational
efforts for structural design and performance evaluation studies. The soil-structure interaction (SSI) is
neglected for the fixed base case. However, ignoring these effects may cause significant errors on seismic
demand estimates. The various approaches are used for the soil-structure interaction in literature. The
simplest form of considering soil effects on superstructure is based on period lengthening and increase of
damping ratio due to existing of flexible soil (Veletsos and Meek, 1974) which limits the consideration of
inertial effects as shown in Figure 1. The rotational and lateral stiffness (𝑘𝜃 , 𝑘𝑢 ) of foundation is simplified
as spring members in model. The flexible base period 𝑇̃ and damping ratio of structure with flexible soil 𝜁̃
can be obtained using the following equations:
𝑇 𝑘 𝑘ℎ2
= 1+ + (1)
𝑇 𝑘𝑢 𝑘𝜃

1
Asst. Professor Dr., Pamukkale University Civil Engineering Department, Denizli, Turkey, bcayci@pau.edu.tr
2
Grad. Student, Pamukkale University Civil Engineering Department, Denizli, Turkey, gokzey94@gmail.com
3
Professor. Dr., Pamukkale University Civil Engineering Department, Denizli, Turkey, minel@pau.edu.tr
 𝜁
𝜁 = 𝜁0 + (2)
(𝑇/𝑇)3
where T and is fixed base period of SDOF structure 𝜁̃0 and 𝜁 are contribution of foundation damping and
material damping of fixed base model, k is stiffness and h is height of superstructure. Various methods are
generalized for better definition of inertial interaction problem (Stewart et al. 1999). The modeling of SSI
requires knowledge of various parameters such as ground motion characteristics, soil properties and
transmitting boundary conditions (Kutanis, 2001; Saez et al. 2013; Wolf, 2001). In general, it is assumed that
the increase in period and damping ratio is beneficial to reflect SSI on seismic demands. However, this is not
always the case as mentioned in the previous studies. The consideration of soil may cause ductility demand
increment in certain cases (Mylonakis and Gazetas, 2000). The studies in literature conclude that the SSI is
an extremely complex problem and the outcome of analyses are highly sensitive the input motion
characteristics. Seismic demands may significantly increase or decrease for individual ground motion records
although the SSI effects are not significant in average sense. Besides, soil-structure interaction is also affected
by the existence of adjacent buildings. Base shear and moment demands may increase with SSI in certain
cases (Trombetta et al. 2015).

Figure 1. Simplified model of SSI with inertial interaction

The aim of this study is to evaluate the effect of soil-structure interaction on seismic behavior of existing
reinforced concrete (RC) buildings considering nonlinear behavior of structural members. Two-dimensional
16-, 12- and 8-story reinforced concrete (RC) frame buildings were used in the scope of the study. The
selected buildings were designed according to modern Turkish Earthquake Code (2007) considering
both gravity and seismic loads. A design ground acceleration of 0.40g assuming the highest seismicity
zone in Turkey and soil class Z3 that is similar to the class C soil of FEMA-356 is assumed (FEMA-
356, 2000) for the design of building according to the 2007 Turkish Earthquake Code. Compressive
concrete strength value of 35 MPa is considered while the yield strength of both longitudinal and
transverse reinforcement is assumed to be 420 MPa. Four ground motion records and 2 soil types with
different stiffness were taken into account during nonlinear time history analyses. The soil layer is assumed
to be elastic in the scope of the study. All buildings have the same plan dimensions as shown in Figure 2.
The interior frame in X direction considered for the 2D models are also shown on the figure. Schematic
illustration of 16-, 12-, and 8-story building models are given Figure 3. The story heights are 3 m for all
buildings.

2
 Figure 2. Plan view of building models

Figure 3. Schematic illustration of two-dimensional building models.

2. METHODOLOGY

This study uses finite element method (FEM) and 2D linear and nonlinear time history analysis with general-
purpose structural analysis program SAP2000. Viscous boundary approach was used as transmitting
boundary conditions to eliminate propagating waves (Lysmer and Kuhlemeyer, 1969). Substructure method
was used during SSI analysis. In the first step, free field motions were calculated for each case from two-

3
dimensional soil model. SAP2000, using the SOLID element, can be used to calculate either one, two or three
dimensional free-field motions if the soil material is considered as linear (Wilson, 2001). Then, free field
motions were applied to structure for both SSI and fixed base models. The mass of soil was considered in
SSI model even the motions were applied to the structure alone. Two different soil cases were considered for
this study. Soil dimensions were taken as 200 m in longitudinal (X), and 30 m in depth. Variable soil layers
were taken into account throughout to height. The bottom of soil layer was assumed to be on the rock defined
as fixed at the bottom in the models. Mesh length of soil is taken as 0.5 m. Nonlinear behavior of soil was
not taken into account in modeling. Detailed information about soil properties are given in Table 1. The
foundation system of model is assumed as rigid and located in shallow medium. The analysis steps of SSI
and the fixed base models are shown in Figure 4.

a) calculation of free field motion

b) analysis of SSI model c) analysis of fixed base modes

Figure 4. Analysis steps of SSI and fixed base models

Table 1. Soil properties used in this study

Soil S1 S2
Vs Density Poisson Density Poisson
Depth Damping (%) Vs (m/s) Damping (%)
(m/s) (kN/m3) Ratio (kN/m3) Ratio
0-5 100 15.69 0.4 10% 200 17.65 0.35 5%
5-10 150 15.69 0.4 5% 200 17.65 0.35 5%
10-15 150 15.69 0.40 5% 400 19.61 0.30 5%
15-20 200 17.65 0.35 5% 400 19.61 0.30 5%
20-30 200 17.65 0.35 5% 800 21.57 0.25 3%

Nonlinearity of structural models was defined with lumped plasticity approach by defining plastic hinges at
both ends of beams and columns. Inel and Ozmen (2006), studied possible differences on the results of
pushover analysis due to default and user-defined nonlinear component properties. They observed that
although the model with default hinge properties seemed to provide reasonable displacement capacity for the
well-confined case, the displacement capacity was quite high compared to that of the poorly-confined case.
Thus, this study implements user-defined hinge properties. As shown in Figure 5, five points labelled A, B,
C, D and E define force-deformation behaviour of a typical plastic hinge. The values assigned to each of
these points vary depending on type of element, material properties, longitudinal and transverse steel content,
and axial load level on the element. Plastic hinge length is assumed to be equal to half of the section depth
4
as recommended in 2007 Turkish Earthquake Code and other documents such as ATC-40 (ATC-40, 1996)
and FEMA-356. Also, effective stiffness values are obtained per the code; 0.4EI for beams and values
between 0.4 and 0.8EI depending on axial load level for columns. Shear hinges were also defined at the
middle of columns to reflect brittle behaviour of members. Shear hinges were not effective on results in the
scope of this study since none of column members reached the shear capacity.
The Mander confined concrete model (Mander, 1988) was used for moment-curvature analyses of RC
sections. Newmark average acceleration method was used for nonlinear time history analyses in
SAP2000.

Figure 5. Force-deformation relationship for a typical plastic hinge

Four different ground motions recorded on soil type A and B were used during dynamic time history analyses.
Table 2 lists the records considered in this study (Pacific Earthquake Center Strong-Motion Virtual Data
Center). The peak ground acceleration given in the table is the original value recorded at the rock layer. This
record is carried out as the free field motion using soil properties assumed for this study. Therefore, the peak
ground acceleration values of free filed motions are higher than the values provided in the table.

Table 2. Ground motion records used in this study

ID Date Station Comp. Vs30 PGA

3. RESULTS

Nonlinear time-history analysis results of SSI and fixed base frames were compared for two soil types
under for different ground motion records. Soil type S1 has lower stiffness than soil type S2 as seen in
Table 1. Therefore, it is expected that the SSI effects are more predominant for soil type S1. Table 3
lists maximum displacement demands for the fixed base and SSI models for the free field motions of
the ground motion records. The average values of maximum roof drift demand ratios are compared in
Figure 6 for the SSI and fixed base models. The horizontal axis represents the buildings in terms of
height-width ratio. It is obvious that the seismic demands of the SSI models tend to decrease as the
height-width ratio increases for low-stiffness soil profile as it seen in Figure 6a. The difference between
SSI and fixed based models is not significant for the 8-story model. With the increase of soil stiffness,
dynamic amplification effects are more predominant for superstructure models. While roof drift demand
ratio of SSI and fixed base models is found 0.89 for 8-story model for soil type S1 in average, it increases
to 1.49 for soil type S2. The reason of higher roof drift demands of 8-story SSI model compared to the
fixed base model is related to dynamic amplification due to change in dominant period rather than soil
behavior. Displacement profile along the building height is an indicator for demand pattern of seismic
loads. It shows sudden changes of story displacement for irregular structures. Displacement pattern of
the structures may differ for different cases although they have similar roof drifts. Thus, the average
displacement profiles of the used models are compared for the fixed base and SSI approaches at
maximum roof displacement in order to understand the differences between approaches more closely in
Figures 7 and 8.

5
 Table 3. Maximum displacement demands for the fixed base and SSI models of the building models
considered in the study

Soil type S1
SSI Fixed
Record 16-story 12-story 8-story 16-story 12-story 8-story
mm mm mm mm mm mm
chichi-tcu 37.60 93.87 133.02 87.39 102.24 130.23
kobe-nis 74.69 98.68 140.44 127.72 114.78 159.28
kocaeli-izt 32.02 58.38 54.02 66.31 79.94 78.18
lomap-hsp090 36.22 48.23 67.07 56.10 82.56 72.63
Soil type S2
SSI Fixed
Record 16-story 12-story 8-story 16-story 12-story 8-story
mm mm mm mm mm mm
chichi-tcu 17.83 22.30 83.32 25.64 28.15 35.86
kobe-nis 25.86 31.94 32.43 32.37 46.87 28.44
kocaeli-izt 14.03 14.57 13.24 22.98 21.35 16.41
lomap-hsp090 8.61 10.41 16.80 14.42 16.57 16.86

1 1.75

1.5
0.75 R² = 0.9887 1.25
SSI/Fixed

SSI/Fixed

1
0.5 R² = 0.7548
0.75

0.5
0.25
0.25

0 0
0.5 1 1.5 2 0.5 1 1.5 2
H/W H/W
a) Soil type S1 b) Soil type S2
Figure 6. Comparison of roof drift demand ratios with H/W ratios of buildings.

6
 Fixed SSI 16-story S1 Fixed SSI 12-story S1
16 12

14
10
12
8
10
Story

Story
8 6

6
4
4
2
2

0 0
0 25 50 75 100 125 0 25 50 75 100 125
Displacement (mm) Displacement (mm)
Fixed SSI 8-story S1
8

5
Story

0
0 25 50 75 100 125
Displacement (mm)

Figure 7. Comparison of average displacement profiles for soil type S1

The profiles start to deviate for soil type S1 due to effect of soil damping of SSI models. The difference
between SSI and fixed base approaches tends to decrease as number of stories decreases. The higher mode
effects are clearly observed for soil type S2 in Figure 8. Another important point is the increase of seismic
demands for 8-story model for the SSI model. The effect of soil deformations is quite limited with soil type
S2. While the demand pattern of 16- and 12-story models is similar for SSI and fixed base models, the
observed difference for 8-story model can be attributed to dynamic amplification.

7
 Fixed SSI 16-story S2 Fixed SSI 12-story S2
16 12

14
10
12
8
10
Story

Story
8 6

6
4
4
2
2

0 0
-25 0 25 50 -25 0 25 50
Displacement (mm) Displacement (mm)
Fixed SSI 8-story S2
8

5
Story

0
-25 0 25 50
Displacement (mm)
Figure 8. Comparison of average displacement profiles for soil type S2

Interstory drift ratios (IDR) obtained from remarkable cases are illustrated in Figures 9 and 10 for soil type
S1 and S2 respectively. The difference between SSI and fixed base models is obvious when the interstory
profiles are examined for individual records. The outcomes show that the interstory drifts at lower stories
tend to increase for fixed base models as seen in Figure 9. The interstory drift ratios of SSI models tend to
increase for upper stories due to soil damping at the base of structure. The interstory drift ratios of 16- and
12-story building models have similar trend for soil type S2 in Figure 10. Maximum IDR values are higher
for 8-story SSI model as stated in comparison of demand pattern.

8
 Fixed SSI 16-story Fixed SSI 12-story
16 Kobe-nis S1 12 Chichi-tcu S1

14
10
12
8
10

Story
Story

8 6

6
4
4
2
2

0 0
-1 0 1 2 3 -1 0 1 2 3
IDR (%) IDR (%)

Fixed SSI 8-story

5
Story

0
-1 0 1 2 3
IDR (%)

Figure 9. Comparison of IDR profiles for remarkable cases at soil type S1

Plastic hinge damage states of 8-story SSI and fixed base models subjected to Chichi-tcu ground motion
record at soil type S2 are given in Figure 11 in order to better understand the behavioral differences of SSI
and fixed base model. Damage states are represented as yielding point (B), Immediate Occupancy (IO), Life
Safety (LS), and Collapse Prevention (CP). As it seen in the figure, damage state of beam and column
members are remarkably higher for the SSI model. While 100% of beam members of the first story are at
LS-CP damage state for SSI model, none of beam member is at LS-CP damage level for the fixed base model.
Similarly damage level of column members are higher for the SSI model.

9
 Fixed SSI 16-story Fixed SSI 12-story
16 Kobe-nis S3 12 Kobe-nis S3

14
10
12
8
10
Story

Story
8 6

6
4
4
2
2

0 0
-1 0 1 2 -1 0 1 2
IDR (%) IDR (%)

Fixed SSI 8-story

5
Story

0
-1 0 1 2
IDR (%)

Figure 10. Comparison of IDR profiles for remarkable cases at soil type S2

10
 Column Beam
B-IO% IO-LS% LS-CP% B-IO% IO-LS% LS-CP%
100 100
90 90
80 80
70 70
60 60
50 50

%
%

40 40
30 30
20 20
10 10
0 0
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
Story Story
a) fixed base
Column Beam
B-IO% IO-LS% LS-CP% B-IO% IO-LS% LS-CP%
100 100
90 90
80 80
70 70
60 60
50 50
%

40 40
30 30
20 20
10 10
0 0
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
Story Story
b) SSI
Figure 11. Plastic hinge damage states of 8-story model for soil type S2 under chichi-tcu record

4. CONCLUSIONS

This study evaluates the effect of soil-structure interaction on seismic behavior of existing reinforced concrete
(RC) buildings considering nonlinear behavior of structural members. Two-dimensional 16-, 12- and 8-story
reinforced concrete (RC) frame buildings were used in the scope of the study. Two-dimensional SSI and
fixed base nonlinear models are compared for roof and displacement profiles to better understand behavioral
difference of two approaches. It is concluded that soil-structure interaction becomes more effective on
seismic demands as the height/with ratio of superstructure increases. The SSI is a more complex problem for
buildings with lower height/width ratios. The consideration of soil-structure interaction is also effective on
damage pattern of structures. The outcomes show that the interstory drifts at lower stories tend to increase
for fixed base models while the interstory drift ratios of SSI models increase at the upper stories due to soil
damping at the base of structure. The SSI effects are less remarkable for stiffer soil profiles as expected. The
dynamic amplification due to frequency content of surface motion is more effective especially for structures
with lower H/W ratios on stiffer soil layer. The consideration of SSI affects the damages of beam and column
elements. The findings in the scope of this study is limited to mid and high-rise frame buildings without shear
walls. It should be noted that the existence of shear walls or structural irregularities may directly affect the
interaction between soil and structure.

11
5. ACKNOWLEDGMENTS

The authors acknowledge support provided by Pamukkale University Research Fund Unit (PAU-BAP)
for the conference participation grant and partial support of Pamukkale University Research Fund Unit
under Project No: 2015FBE011.

6. REFERENCES

FEMA-356(2000). Prestandard and commentary for seismic rehabilitation of buildings, Report No. FEMA-356,
Federal Emergency Management Agency, Washington, DC.
Inel M, Ozmen HB (2006). Effects of Plastic Hinge Properties in Nonlinear Analysis of Reinforced Concrete
Buildings, Engineering Structures, 28,1494–502. doi:10.1016/j.engstruct.2006.01.017.
Kutanis M (2001). Dynamic Soil-Structure Analysis,” Ph.D. Thesis, Sakarya University.
Lysmer J and Kuhlemeyer RL (1969). Finite Dynamic Model for Infinite Media, Journal of The Engineering
Mechanics, Div 95, 859–78.
Mander JB, Priestley MJN and Park R (1988). Theoretical stress-strain model for confined concrete. ASCE,
114(8), 1804-1826.
Mylonakis G and Gazetas G (2000). Dynamic Behavior of Building-Foundation Systems, Journal of Earthquake
Engineering, Vol. 4, No. 3, 277-301
Ozmen HB, Inel M and Cayci BT (2013), “Engineering Implications of the RC Building Damages After 2011 Van
Earthquakes,” Earthquakes and Structures, 5, 297–319. doi:10.12989/eas.2013.5.3.297.
Ozmen HB, Inel M, Akyol E, Cayci BT and Un H (2014). Evaluations on The Relation of RC Building Damages
with Structural Parameters After May 19, 2011 Simav (Turkey) Earthquake, Natural Hazards, 71, 63–84,
doi:10.1007/s11069-013-0900-y.
Pacific Earthquake Engineering Research Center ground motion database http://peer.berkeley.edu
Saez E, Lopez-Caballero F and Modaressi-Farahmand-Razavi A (2013). Inelastic Dynamic Soil-Structure
Interaction Effects on Moment-Resisting Frame Buildings, Engineering Structures, 51:166–77.
doi:10.1016/j.engstruct.2013.01.020.
SAP2000 V-19 CSI (2016). Integrated finite element analysis and design of structures basic analysis reference
manual, Berkeley, USA.
Trombetta NW, Mason HB, Hutchinson TC, Zupan JD, Bray JD and Kutter BL (2015). Nonlinear Soil–
Foundation–Structure and Structure–Soil–Structure Interaction: Engineering Demands, Journal of Structural
Engineering, doi: 10.1061/(ASCE)ST.1943-541X.0001127
Turkish Earthquake Code (TEC-2007) (2007). Specifications for buildings to be built in seismic areas. Ministry
of Public Works and Settlement. Ankara, Turkey.
Veletsos AS and Meek JW (1974). Dynamic Behavior of Building-Foundation Systems, Earthquake Engineering
and Structural Dynamics, Vol. 3, 121-138
Wilson EL (2001). Three Dimensional Static and Dynamic Analysis of Structures
Wolf JP. Dynamic Soil-Structure Interaction 2001, 1 st Ed., Prentice-Hall, New Jersey, USA

12