Vol.11, No.

EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION

December, 2012

Earthq Eng & Eng Vib (2012) 11: 541-551

DOI: 10.1007/s11803-012-0140-2

Effect of nonlinear soil-structure interaction on seismic response of low-rise SMRF buildings

Prishati Raychowdhury and Poonam Singh
Indian Institute of Technology, Kanpur, India

Abstract: The nonlinear behavior of a soil-foundation system may alter the seismic response of a structure by providing
additional exibility to the system and dissipating hysteretic energy at the soil-foundation interface. However, the current design practice is still reluctant to consider the nonlinearity of the soil-foundation system, primarily due to lack of reliable modeling techniques. This study is motivated towards evaluating the effect of nonlinear soil-structure interaction (SSI) on the seismic responses of low-rise steel moment resisting frame (SMRF) structures. In order to achieve this, a Winklerbased approach is adopted, where the soil beneath the foundation is assumed to be a system of closely-spaced, independent, nonlinear spring elements. Static pushover analysis and nonlinear dynamic analyses are performed on a 3-story SMRF building and the performance of the structure is evaluated through a variety of force and displacement demand parameters. It is observed that incorporation of nonlinear SSI leads to an increase in story displacement demand and a signicant reduction in base moment, base shear and inter-story drift demands, indicating the importance of its consideration towards achieving an economic, yet safe seismic design.

Keywords: soil-structure interaction; Winkler modeling; nonlinear analysis; seismic response; low-rise steel moment
resisting frame

1 Introduction
Nonlinear behavior of a soil-foundation interface due to mobilization of ultimate capacity and consequent energy dissipation during an intense seismic event may alter the response of a structure in several ways. Foundation movement can increase the period of the structure by introducing additional exibility, material nonlinearity and hysteretic energy dissipation at the soil-foundation interface may reduce the force demand to the structure, foundation deformations may alter the characteristics of input ground motion, and so on. However, todate, the current design practice is reluctant to account for the nonlinear soil-structure interaction (SSI), for two principal reasons: (a) in anticipation that consideration of SSI leads to more conservative design, and (b) absence of reliable nonlinear modeling techniques. In the last few decades, a number of analytical and experimental studies have been conducted to understand the effect of SSI on the seismic behavior of structures (Chopra and Yim (1985); Yim and Chopra (1985);
Correspondence to: Prishati Raychowdhury, Indian Institute of Technology, Kanpur, India Tel: +91 512 259 6692; Fax: +91 512 259 7395 E-mail: prishati@iitk.ac.in Assistant Professor; Former Graduate Student Received January 2, 2012; Accepted September 20, 2012

Xiong et al. (1993, 1998); Nakaki and Hart (1987); Gazetas and Mylonakis (2001); Stewart et al. (2003); Dutta et al. (2004); Allotey and Naggar (2007); Harden and Hutchinson (2009); Raychowdhury and Hutchinson (2009, 2010, 2011); Gajan et al. (2010); Figini et al. (2012) and so on). These studies have indicated that the nonlinear soil-foundation behavior under intense earthquake loading has a considerable effect on the response of structure-foundation system (Fig. 1). Design and rehabilitation provisions (e.g., ATC-40, 1996; NEHRP, 2003; FEMA 356, 2000; ASCE-7, 2005) have traditionally focused on simplied pseudo-static force-based or pushover type procedures, where the soil-foundation interface is characterized in terms of modied stiffness and damping characteristics. However, the above-mentioned approaches cannot capture the complex behavior of nonlinear soil-foundation-structure systems, such as hysteretic and radiation damping, gap formation in the soil-foundation interface and estimation of transient and permanent settlement, and sliding and rotational deformations of the foundation. In this study, the seismic response of a ductile steel moment resisting frame (SMRF) building has been evaluated considering nonlinear soil-foundation interface behavior using a Beam-on-NonlinearWinkler-Foundation (BNWF) approach, where the soil-foundation interface is assumed to be a system of closely-spaced, independent, inelastic spring elements.

542

EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION

Vol.11
Large displacements cause frame damage , large

High forces cause shear wall damage

, small

Foundation yielding and rocking protects shear wall

Small displacement protect frame from damage

(a) Stiff and strong

(b) Flexible and weak

Fig. 1 Effect of foundation exibility on a typical structure (after ATC 40, 1996)

The effect of nonlinear SSI on the structural response is studied in terms of base moment, base shear, story displacement, and interstory drift angle. Nonlinear time history analysis is carried out along with static pushover analysis and eigenvalue analysis. The analysis is conducted using nite element models in the framework of Open System for Earthquake Engineering Simulation (OpenSees, 2008). The details of the modeling technique, soil and structural properties considered, and analysis procedures adopted are discussed below.

Poissons ratio: 0.4. The effective shear modulus is obtained by reducing the maximum shear modulus corresponding to small strain values by 50% to represent the high strain modulus during signicant earthquake loadings. The foundation was designed in such a way that it has a bearing capacity three times the vertical load coming to it (i.e., a static vertical factor of safety of 3). Further details of the structure and soil properties and analysis procedure can be found in Singh (2011).

Selection of structure and soil properties

3 Numerical modeling of SSI

In this study, the soil-foundation interface is modeled using a Beam-on-Nonlinear-Winkler-Foundation (BNWF) concept, where the soil-foundation interface is assumed as an assembly of closely spaced, independent nonlinear spring elements. Each spring consists of different elements such as: drag and closure spring, dashpot, near-eld plastic spring, and a far-eld elastic spring (Fig. 3). These elements together are responsible for capturing several salient features of a foundation-soil interface, such as transient and permanent deformations, gap formation between foundation and surrounding soil, radiation and hysteretic damping of the adjacent soil, strength and stiffness degradation, and so on. Vertical springs (q-z elements) distributed along the length of the footing are intended to capture the rocking, uplift and settlement, while horizontal springs (t-x and p-x elements) are intended to capture the sliding and passive resistance of the footing, respectively. The constitutive relations used for the q-z, p-x and t-x mechanistic springs are represented by nonlinear backbone curves that were originally developed by Boulanger (2000), based on an earlier work of Boulanger et al. (1999); and later calibrated and validated by Raychowdhury (2008) for more appropriate utilization towards shallow foundation behavior modeling. Each spring material model consists of an initial elastic portion followed by a smooth inelastic curve (as shown in Fig. 4). In the elastic portion of a q-z material, the instantaneous load q is assumed to be linearly proportional with the instantaneous

A 3-story, 4-bay steel moment resisting frame (SMRF) building adopted from Gupta and Krawinkler (2000) is considered for the present study. The building was designed based on weak-beam strong-column mechanism, with a oor area of 36:6 36:6 m and four bays at an interval of 9.15 m in each direction (Fig. 2). The section properties and geometric details of the structure have been taken from Gupta and Krawinkler (2000). The columns of the building are assumed to be supported on a mat foundation resting on dense silty sand of the Los Angeles area (site class-D, NEHRP); with the following soil properties: cohesion: 70 kPa, unit weight: 20 kN/m3, shear modulus: 5.83 MPa and

6 bays @ 9.15 m

4 bays @ 9.15 m

3 @ 3.96 m

Fig. 2 SMRF structure considered in the study (adopted from Gupta and Krawinkler (2000))

No.4

Prishati Raychowdhury et al.: Effect of nonlinear soil-structure interaction on seismic response of low-rise SMRF buildings

543

Near-structure plastic response

Drag Plastic Damper

Closure Zero length Elastic G.S. Nodes

Far-structure elastic response

Elastic beam-column elements

p-x elements

G.S.

q-z elements

t-x elements

Fig. 3 Numerical modeling of nonlinear SSI: BNWF approach

1.0
Normalized load, q/quit Normalized load, p/puit

1.0
Normalized load, t/tuit

1.0 0.5 0

0.5

0.5 0

-0.5 -1.0 -16

-0.5 -1.0 -16

-0.5 -8

0 8 Vertical displacement (mm) (a)

16

-8 0 8 16 Horizontal displacement (mm) (b)

-8 0 8 16 Horizontal displacement (mm) (c)

Fig. 4 Material models for soil-foundation interface elements: (a) q-z element; (b) p-x element and (c) t-x element

displacement z: q = kinz (1)

where kin is the initial elastic (tangent) stiffness. The range of the elastic region is dened by the following relation: q0 = Cr qult (2) where q0 is the load at the yield point, Cr is a parameter controlling the range of the elastic portion, and qult is the ultimate load. In the nonlinear (post-yield) portion, the backbone curve is described by cz50 q = qult (qult q0 ) p p cz50 + z z0
n

(3)

where z50 is the displacement at which 50% of the ultimate load is mobilized, zpo is the displacement at the yield point, zp is the displacement at any point in the post-yield region, and c and n are the constitutive parameters controlling the shape of the post-yield portion of the backbone curve. The expressions governing both

p-x and t-x elements are similar to Eqs. (1) to (3), with variations in the constants n, c, and Cr, controlling the general shape of the curve. Furthermore, it may be noted that q-z material has a reduced capacity in the tension side (Fig. 4(a)), which allows the footing to uplift without losing contact with the soil beneath it during a rocking movement. On the other hand, the p-x material is characterized by a pinched hysteretic behavior (Fig. 4(b)), whereas the t-x material is characterized by a large initial stiffness and a broad hysteresis (Fig. 4(c)). It is evident from Eqs. (1) to (3) that the shape of spring backbone curves, which is basically a controlling factor for the SSI behavior, is mainly dependant on two physical parameters related to soil characteristics; namely, capacity (qult), and initial elastic stiffness (kin). The capacity and elastic stiffness of each spring is obtained by distributing the global footing capacity and stiffness utilizinga proper tributary area of each spring. The footing capacity is derived using the general bearing capacity equation from Terzaghi (1943) with shape, depth and inclination factors after Meyerhof (1963) as shown in Eqs. (4) to (7).

544

EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION

Vol.11

qult=cNcFcsFcdFci+DfNqFqsFqdFqi +0.5BNFsFdFi (4) where qult is the ultimate vertical bearing capacity per unit area of footing, c the cohesion, the unit weight of soil, Df the depth of embedment, and B the width of footing. Bearing capacity factors, Nc, Nq and N are calculated after Meyerhof (1963):

or nonlinear region of the backbone curves, may be expressed as: (cz50 ) n kp = n(qult q0 ) (12) n +1 (cz50 z0 + z ) Note that Eq. (12) was obtained by rearranging Eq. (3). Also, note that the shape and instantaneous tangent stiffness of the nonlinear portion of the backbone curve is a function of the parameters c and n, which are derived by calibrating against a set of shallow footing tests on different types of soil. It is evident from the preceding equations that the spring responses are primarily dependent on basic strength and stiffness parameters of soil such as cohesion (c), friction angle (), unit weight (), shear modulus (G) and Poissonsratio ()of soil. The BNWF model has shown good predictive capability in capturing the key features of a shallow foundation behavior. Figure 5 shows a sample result showing the BNWF model performance in predicting experimentally observed behavior of a strip footing of size 2.8 m 0.65 m (prototype scale) resting on dry dense sand with a relative density of 80%, and a vertical factor of safety of 5.2 carried out at the centrifuge facility at the University of California. Note that the BNWF model can reliably predict the following behavior of a seismically loaded footing: (a) peak moment, (b) peak shear, (c) initial and post-yield stiffness of moment-rotation and shear-sliding responses, (d) transient and permanent sliding, settlement and rotation, and (e) overall shape of the moment-rotation and shear-sliding loops. More comparison results can be found in Raychowdhury (2008), Raychowdhury and Hutchinson (2009), and Gajan et al. (2008, 2010). In order to investigate the effect of foundation nonlinearity efciently, in addition to the nonlinear base condition (i.e., BNWF model), two additional base conditions are considered: (1) xed base, where the movement of the foundation is neglected indicating absolutely no SSI, and (2) linear base condition, where the force-deformation behavior of the interface springs are assumed to be linear elastic with initial stiffness the same as that considered in the nonlinear base condition. The initial elastic stiffness and vertical capacity of the soil springs are calculated based on Gazetas (1991) and Terzaghi (1943), respectively. Springs are distributed at a spacing of 1% of the footing length at the end region and 2% of the footing length at the mid region. The end region is dened as a high stiffness region extending 10% of the footing length from each end of the footing; while the mid region is the less stiff middle portion. This variation in the stiffness distribution is provided based on the recommendations of ATC-40 (1996) and Harden and Hutchinson (2009) in order to achieve desirable rocking stiffness of the foundation.

N q = tan 2 (45 + )e tan 2

N c = ( N q 1) cot N = ( N q 1) tan(1.4 )

(5) (6) (7)

For the p-x material, the ultimate lateral load capacity is determined as the total passive resisting force acting on the front side of the embedded footing. For homogeneous backll against the footing, the passive resisting force can be calculated using a linearly varying pressure distribution resulting in the following expression: pult = 0.5 Df2 K p (8) where pult = passive earth pressure per unit length of footing, and Kp = passive earth pressure coefcient. For the t-x material, the lateral load capacity is the total sliding (frictional) resistance, which can be dened as the shear strength between the soil and the footing as: (9) where tult = frictional resistance per unit area of foundation, Wg = vertical force acting at the base of the foundation, = angle of friction between the foundation and soil (typically varying from 1/3 to 2/3 ) and Ab = the area of the base of footing in contact with the soil (=L B). The initial elastic stiffness (vertical and lateral) of the footing are derived from Gazetas (1991) as follows: kv = GL B 0.73 + 1.54( )0.75 1 v L GL B 2 + 2.5( )0.85 2v L (10) tult = Wg tan + Ab c

kh =

(11)

where kv and kh are the vertical and lateral initial elastic stiffness of the footing, respectively; G is the shear modulus of soil; is the Poissons ratio of soil; and B and L are the footing width and length, respectively.The instantaneous tangent stiffness kp, which describes the load-displacement relation within the post-yield

No.4

Prishati Raychowdhury et al.: Effect of nonlinear soil-structure interaction on seismic response of low-rise SMRF buildings 400
Moment (kN.m) Shear force (kN)

545

80 40 0 -40 -80 20 0
Settlement (mm)

200 0 -200 -400 20 0 Experiment BNWF simulation

Settlement (mm)

-20 -40 -60 -80 -100 -0.08 -0.04

-20 -40 -60 -80

-100 -20 0 20 40 60 0 0.04 0.08 Sliding (mm) Rotation (rad) Fig. 5 Predictive capability of BNWF model: Comparison of load-deformation behavior of a footing for BNWF simulation and SSG02_03 centrifuge test carried out by Gajan (2006)

4 Eigenvalue analysis
In order to understand the effect of SSI on the eigen properties of the structure, eigenvalue analysis is carried out for the xed base, linear base and nonlinear base structure. The results are compared and tabulated in Table 1. It has been found that the fundamental period of the xed base structure (i.e., ignoring the SSI effects) is obtained as 1.03 s, which is in accordance with the period obtained by Gupta and Krawinkler (2000). However, when the base exibility is introduced, the fundamental period is increased to 1.37 s, indicating a signicant period elongation (about 33%) due to the SSI effects. Further, the period elongation obtained using methods outlined in NEHRP (2003) gives very close results (Table 1), indicating that inelastic behavior of foundations does not signicantly affect the fundamental period and the rst mode shapes of a structure.

5 Pushover analysis
Following eigenvalue analysis, static pushover analysis is conducted on the building with three different base conditions: xed, linear and nonlinear. Figure 6 shows the pushover curves for different base conditions. It is observed that for linear modeling of the soilfoundation interface, the global response of the system only slightly varies from that of a xed base case, with almost no change in the shape of the curve (depicting a typical elastic-plastic behavior common for a steel building). However, when the soil springs are modeled as nonlinear, the curve becomes much softer, resulting in lower yield force and higher yield displacement demands, which may be associated with yielding of the soil beneath the foundation. Table 2 provides the numerical values of the displacements and forces corresponding to the rst yield point for these three

Table 1 Eigenvalue analysis results Fundamental period from Gupta and Krawinkler (2000) 1.03 s Fundamental period from present study (Fixed base case) 1.03 s Fundamental period from present study (Linear base case) 1.37 s Fundamental period from present study (Nonlinear base case) 1.37 s Period elongation /T ) (T from present study 1.34 Period elongation (T /T ) using NEHRP provisions 1.33

Table 2 Pushover analysis results Base condition Fixed base Linear base Nonlinear base Yield displacement (m) 0.19 0.25 0.28 Yield force (MN) 4.53 4.28 2.86

546

EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION

Vol.11

4 Base shear (MN)

3 Fixed base Linear base Nonlinear base

0.2

0.4 0.6 Roof displacement (m)

0.8

Fig. 6 Global pushover curves for different base conditions

cases. It can be observed that the yield displacement increases about 32% from xed base to linear base case, whereas it increases about 47% for the nonlinear base. The yield force, on the other hand, decreases only slightly (about 5.5%) from the xed base to linear base, but signicantly (about 37%) when base nonlinearity is introduced. This signicant difference in the force and displacement demands for the nonlinear base condition may be due to the capacity mobilization and yielding of the soil-foundation interface spring elements. Since pushover analysis is a well-accepted and widely-used method for assessing the seismic demand of structures, these observations provide a crucial indication that nonlinear SSI may signicantly alter the seismic demand of a structure.

6 Nonlinear dynamic analysis

In this section, nonlinear dynamic analysis is carried out using 60 ground motions (SAC motions from Somerville et al., 1997). These ground motions are divided into three sets (each set having 20 motions), representing the probabilities of exceedance of 50%, 10% and 2% in 50 years; with return periods of 72 years, 475 years, and 2,475 years, respectively. Table 3 provides the details of the ground motions used in this study. Figure 7 shows the acceleration response spectra of the motions, where spectra for individual motions are shown with light grey lines and the median and 84th percentile curves are shown with darker and thicker lines. A 2% Rayleigh damping is used for the dynamic analysis. In this paper, the three sets of ground motions are denoted as 50/50, 10/50 and 2/50, respectively, for brevity. See Somerville et al. (1997) for further details regarding these motions. Since the geometry and material properties of the moment frame is taken from Gupta and Krawinkler

(2000), the structural response is rst compared with that of the above-mentioned study in order to check the validity and reliability of the current xed-base model. Figure 8 shows a comparison of frame responses from the present study (xed base case) with that from Gupta and Krawinkler (2000). The observation indicates that the results are comparable for all oor levels and all ground motions, ensuring the reliability of the structural model. However, to predict the seismic responses of the frame in a more accurate way, seismic SSI effects must be taken into account. The following paragraphs provide results showing the effect of considering both linear and nonlinear SSI effects. Figures 9 through 13 provide the statistical results of various force and displacement demands obtained from the nonlinear dynamic analysis considering xed, linear and nonlinear base conditions using the aforementioned 60 ground motions. The maximum absolute value of each response parameter (such as moment, shear, story displacement, and story drift angle) is considered as the respective demand value. Figure 9 demonstrates the median values of story displacements for three sets of ground motions and three base conditions. Note that story displacement increases up to 40% from xed base to linear base, and up to about 2.5 times from xed base to nonlinear base. This increase in story displacement in exible bases (both linear and nonlinear) is caused by the overall reduction in the global stiffness resulting from the induced foundation movements. Further, this trend is consistent for all ground motions and all oor levels. These results indicate that lack of proper consideration of SSI will lead to an underestimation of the story displacement demands, and may crucially affect several design decisions. Although the absolute displacements at story levels are greater in the case of exible-base conditions (both linear and nonlinear), the relative displacements show a decreasing trend when base nonlinearity is introduced, as indicated in Fig. 10. The statistical values of the story drift angle (story drift divided by story height), presented in Fig. 10 indicate that the relative story drift, which is generally known as the inter-story drift ratio, only slightly alters when the base condition is changed from xed to elastic SSI, but reduces signicantly (as much as one-third) when nonlinear SSI is incorporated. Since the story drift demand can be related to the global demands, such as root drift and spectral displacement, as well as to the element level force and displacement demands, it can be considered as an important parameter to assess the performance of a structure (Gupta and Krawinkler, 2000). Current design provisions (such as NEHRP, 2003) provide indicative values for acceptable story drifts at various performance levels, making these parameters crucial from a performance-based design

No.4

Prishati Raychowdhury et al.: Effect of nonlinear soil-structure interaction on seismic response of low-rise SMRF buildings Table 3 Details of selected ground motions (adopted from Somerville et al. (1997))

547

Hazard Level

SAC Name LA41 LA42 LA43 LA44 LA45 LA46 LA47 LA48 LA49 LA50 LA51 LA52 LA53 LA54 LA55 LA56 LA57 LA58 LA59 LA60 LA01 LA02 LA03 LA04 LA05 LA06 LA07 LA08 LA09 LA10 LA11 LA12 LA13 LA14 LA15 LA16 LA17 LA18 LA19 LA20 LA21 LA22 LA23 LA24 LA25 LA26 LA27 LA28 LA29 LA30 LA31 LA32 LA33 LA34 LA35 LA36 LA37 LA38 LA39 LA40

Record Coyote Lake, 1979 Coyote Lake, 1979 Imperial Valley, 1979 Imperial Valley, 1979 Kern, 1952 Kern, 1952 Landers, 1992 Landers, 1992 Morgan Hill, 1984 Morgan Hill, 1984 Parkeld, 1966, Cholame 5W Parkeld, 1966, Cholame 5W Parkeld, 1966, Cholame 8W Parkeld, 1966, Cholame 8W North Palm Springs, 1986 North Palm Springs, 1986 San Fernando, 1971 San Fernando, 1971 Whittier, 1987 Whittier, 1987 Imperial Valley, 1940, El Centro Imperial Valley, 1940, El Centro Imperial Valley, 1979, Array #05 Imperial Valley, 1979, Array #05 Imperial Valley, 1979, Array #06 Imperial Valley, 1979, Array #06 Landers, 1992, Barstow Landers, 1992, Barstow Landers, 1992, Yermo Landers, 1992, Yermo Loma Prieta, 1989, Gilroy Loma Prieta, 1989, Gilroy Northridge, 1994, Newhall Northridge, 1994, Newhall Northridge, 1994, Rinaldi RS Northridge, 1994, Rinaldi RS Northridge, 1994, Sylmar Northridge, 1994, Sylmar North Palm Springs, 1986 North Palm Springs, 1986 1995 Kobe 1995 Kobe 1989 Loma Prieta 1989 Loma Prieta 1994 Northridge 1994 Northridge 1994 Northridge 1994 Northridge 1974 Tabas 1974 Tabas Elysian Park (simulated) Elysian Park (simulated) Elysian Park (simulated) Elysian Park (simulated) Elysian Park (simulated) Elysian Park (simulated) Palos Verdes (simulated) Palos Verdes (simulated) Palos Verdes (simulated) Palos Verdes (simulated)
5 Spectral acceleration (g)

Earthquake Magnitude 5.7 5.7 6.5 6.5 7.7 7.7 7.3 7.3 6.2 6.2 6.1 6.1 6.1 6.1 6 6 6.5 6.5 6 6 6.9 6.9 6.5 6.5 6.5 6.5 7.3 7.3 7.3 7.3 7 7 6.7 6.7 6.7 6.7 6.7 6.7 6 6 6.9 6.9 7 7 6.7 6.7 6.7 6.7 7.4 7.4 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1

Distance (km) 8.8 8.8 1.2 1.2 107 107 64 64 15 15 3.7 3.7 8 8 9.6 9.6 1 1 17 17 10 10 4.1 4.1 1.2 1.2 36 36 25 25 12 12 6.7 6.7 7.5 7.5 6.4 6.4 6.7 6.7 3.4 3.4 3.5 3.5 7.5 7.5 6.4 6.4 1.2 1.2 17.5 17.5 10.7 10.7 11.2 11.2 1.5 1.5 1.5 1.5
10 Spectral acceleration (g) 8 6 4 2

Duration (s) 39.38 39.38 39.08 39.08 78.60 78.60 79.98 79.98 59.98 59.98 43.92 43.92 26.14 26.14 59.98 59.98 79.46 79.46 39.98 39.98 39.38 39.38 39.38 39.38 39.08 39.08 79.98 79.98 79.98 79.98 39.98 39.98 59.98 59.98 14.945 14.945 59.98 59.98 59.98 59.98 59.98 59.98 24.99 24.99 14.945 14.945 59.98 59.98 49.98 49.98 29.99 29.99 29.99 29.99 29.99 29.99 59.98 59.98 59.98 59.98

PGA (cm/s2) 578.34 326.81 140.67 109.45 141.49 156.02 331.22 301.74 312.41 535.88 765.65 619.36 680.01 775.05 507.58 371.66 248.14 226.54 753.70 469.07 452.03 662.88 386.04 478.65 295.69 230.08 412.98 417.49 509.70 353.35 652.49 950.93 664.93 644.49 523.30 568.58 558.43 801.44 999.43 967.61 1258.00 902.75 409.95 463.76 851.62 925.29 908.70 1304.10 793.45 972.58 1271.20 1163.50 767.26 667.59 973.16 1079.30 697.84 761.31 490.58 613.28

PGV (cm/s) 69.51 26.72 42.43 22.57 24.74 24.24 40.85 25.02 26.94 22.81 42.58 36.87 31.21 32.08 36.72 25.42 21.67 27.05 98.54 60.02 62.39 59.89 83.00 77.11 89.20 47.44 66.07 65.68 91.32 60.36 79.09 56.04 95.55 80.96 98.57 100.60 80.17 118.93 68.27 103.83 142.70 123.16 73.75 136.88 160.42 163.72 130.46 193.52 71.20 138.68 119.97 141.12 111.03 108.44 222.78 245.41 177.47 194.07 85.50 169.30

PGD (cm) 11.06 6.68 22.97 14.27 14.15 14.98 33.44 12.58 6.87 5.74 6.53 5.36 6.34 9.07 7.19 5.85 12.84 17.73 12.66 7.89 27.68 14.29 33.42 48.20 48.29 30.00 33.25 39.50 56.25 46.45 28.16 16.50 19.82 35.58 18.01 26.38 17.37 26.87 15.64 25.57 37.81 34.22 23.07 58.85 29.31 42.93 28.27 43.72 34.58 93.43 36.17 45.80 50.61 50.12 89.88 82.94 77.38 92.56 22.64 67.84

4 Spectral acceleration (g) 3 2 1 0 0.01

2% in 50 years

10% in 50 years

50% in 50 years

4 3 2 1 0 0.01 0.1 Period (s) (b) 1 10

Individual ground motion Median 84th percentile

0.1 Period (s) (a)

10

0 0.01

0.1 Period (s) (c)

10

Fig. 7 Acceleration response spectra of the ground motions considered in the study with hazard levels: (a) 50% in 50 years; (b) 10% in 50 years and (c) 2% in 50 years

548
4

EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION

Vol.11

3 Gupta and Krawinkler, 2000 (50/50 Median) Gupta and Krawinkler, 2000 (10/50 Median) Gupta and Krawinkler, 2000 (2/50 Median) Present study, xed base case (50/50 Median) Present study, xed base case (10/50 Median) Present study, xed base case (2/50 Median) Floor level

0.02

0.04

0.06 Story drift angle

0.08

0.10

0.12

Fig. 8 Comparison of responses from present study (xed-base case) with Gupta and Krawinkler (2000)

Floor level

3
50/50 Median (Fixed base) 50/50 Median (Linear base) 50/50 Median (Nonlinear base) 10/50 Median (Fixed base) 10/50 Median (Linear base) 10/50 Median (Nonlinear base) 2/50 Median (Fixed base) 2/50 Median (Linear base) 2/50 Median (Nonlinear base)

2 0 0.1 0.2 0.3 0.4 Story displacement (m)

0.5

0.6

0.7

Fig. 9 Median values of total story displacement

perspective. Therefore, according to the ndings of this study, the individual structural members are likely to be over-conservatively designed if the inelastic behavior of foundations is ignored. In addition to the displacement demands, the effects of the SSI on the force demands of the structure are also investigated. Figures 11(a) and (b) provide the normalized base moment demand and normalized base shear demand, respectively, plotted against the spectral acceleration corresponding to the fundamental period of the building. Note that both moment and shear demands signicantly reduce when nonlinearity of soil-foundation interface is considered, whereas linear modeling of SSI does not give signicantly different results when compared to a xed base case. This qualitative observation is systematically quantied by showing the statistical values of these demands. Figures 12(a) and (b) summarize the median and 84th percentile values, respectively, of the maximum absolute base moment demand normalized by the same demand parameter corresponding to the xed base case. It can be observed that when linear SSI is considered, the base moment is reduced up to 30% of the xed base moment, whereas consideration of nonlinear SSI leads to a reduction

of up to 60% of the xed base moment. A consistent observation is obtained for each set of ground motions for median as well as 84th percentile values. A similar trend is observed for the normalized base shear demand, as indicated in Fig. 13. It can be observed that the base shear reduces up to 10% of the xed base shear for the linear elastic base condition, whereas the reduction is as much as 60% when nonlinearity at the base is considered. These observations indicate that nonlinear modeling of SSI may lead to a signicant reduction in the structural force demand, mostly due to mobilization of the bearing capacity and energy dissipation of the underlying soil.

Summary and conclusions

This study focuses on the effect of foundation nonlinearity on various force and displacement demands of a structure. A medium height SMRF building adopted from Gupta and Krawinkler (2000) has been used for this purpose. The nonlinear behavior of the soilfoundation interface is modeled using a Winkler-based model concept. Static pushover analyses and nonlinear dynamic analyses are carried out using SAC ground

No.4

Prishati Raychowdhury et al.: Effect of nonlinear soil-structure interaction on seismic response of low-rise SMRF buildings
4

549

Floor level

3 Median 84th percentile Median 84th percentile Median 84th percentile 0 0.02 0.04 0.06 0.08 Story drift angle (a)

(Fixed base) (Fixed base) (Linear base) (Linear base) (Nonlinear base) (Nonlinear base) 0.10 0.12

Floor level

0.02

0.04

0.06 0.08 Story drift angle (b)

0.10

0.12

Floor level

0.02

0.04

0.06 0.08 Story drift angle (c)

0.10

0.12

Fig. 10 Statistical values of inter-story drift demands for ground motions: (a)50% in 50 years; (b) 10% in 50 years and (c) 2% in 50 years hazard levels
3 Normalized moment demand (Mmax/WL) Normalized shear demand (Vmax/W) 3

Fixed base Linear base Nonlinear base 0 2 4 6 Sa (T1) (m/s2) (a) 8 10

Fixed base Linear base Nonlinear base 0 2 4 6 Sa (T1) (m/s2) (b) 8 10

Fig. 11 Seismic force demands of the building: (a) normalized base moment demand and (b) normalizedbase shear demand (W = weight of the building, L =length of the footing)

550

EARTHQUAKE ENGINEERING AND ENGINEERING VIBRATION

Vol.11

1.2

Fixed base Linear base Nonlinear base

1.2

|M|max |M|max_Fixed base

0.8

0.8

0.4

0.4

50/50

10/50 2/50 Ground motion hazard level (a)

50/50

10/50 2/50 Ground motion hazard level (b)

Fig. 12 Statistical values of base moment: (a) median and (b) 84th percentile
1.2 1.2 Fixed base Elastic base Nonlinear base 0.8

|V|max_Fixed base

0.8

|V|max

0.4

0.4

50/50

10/50 2/50 Ground motion hazard level (a)

50/50

10/50 2/50 Ground motion hazard level (b)

Fig. 13 Statistical values of base shear: (a) median and (b) 84th percentile

motions for three different hazard levels provided by Somerville et al. (1997). The following specic conclusions are made from the present study: (1) Pushover analysis results indicate that with incorporation of SSI, the global force demand of a structure reduces about 5.5%, while the roof displacement demand increases about 32%. However, this alteration is more signicant (as much as 37% reduction in force demand and 47% increase in displacement demand) when the inelastic behavior of the soil-foundation interface is taken into account. (2) The dynamic time history analysis indicates that the story displacement demands signicantly increase when base nonlinearity is taken into account. However, the inter-story drift angle values are observed to consistently decrease for nonlinear SSI cases, indicating lower design requirements for individual structural members upon consideration of nonlinear SSI. (3) The global force demands such as base moment and base shear of the columns are reduced to as much as 60% with incorporation of nonlinear SSI, indicating that ignoring nonlinear SSI may lead to an over-conservative estimation of the structural forces. Finally, it may be concluded from this study that

nonlinear behavior of the soil-foundation interface may play a crucial role in altering the seismic demands of a structure, indicating the necessity for incorporation of inelastic foundation behavior in modern design codes to accomplish more economic, yet safe structural design in a performance-based design framework.

References
Allotey N and Naggar MHE (2007), An Investigation into the Winkler Modeling of the Cyclic Response of Rigid Footings, Soil Dynamics and Earthquake Engineering, 28: 4457. ASCE-7 (2005), Seismic Evaluation and Retrot of Concrete Buildings, Structural Engineering Institute (SEI) and American Society of Civil Engineers (ASCE), Reston, Virginia. ATC-40 (1996), Seismic Evaluation and Retrot of Concrete Buildings, Applied Technology Council (ATC), Redwood City, California. Boulanger RW (2000), The PySimple1, TzSimple1, and QzSimple1 Material Models, Documentation for the OpenSees Platform. URL: http: //

No.4

Prishati Raychowdhury et al.: Effect of nonlinear soil-structure interaction on seismic response of low-rise SMRF buildings

551

opensees.berkeley.edu. Boulanger RW, Curras CJ, Kutter BL, Wilson DW and Abghari A (1999), Seismic Soil-pile-structure Interaction Experiments and Analyses, ASCE Journal of Geotechnical and Geoenvironmental Engineering, 125(9): 750759. Chopra A and Yim SC (1985), Simplied Earthquake Analysis of Structures with Foundation Uplift, ASCE Journal of Structural Engineering, 111(4): 906930. Dutta SC, Bhattacharya K and R R (2004), Response of Low-rise Buildings under Seismic Ground Excitation Incorporating Soil-structure Interaction, Soil Dynamics and Earthquake Engineering, 24: 893914. FEMA 356 (2000), Prestandard and Commentary for the Seismic Rehabilitation of Buildings, American Society of Engineers, Virginia. Figini R, Paolucci R and Chatzigogos CT (2012), A Macro-element Model for Non-linear Soilshallow Foundationstructure Interaction under Seismic Loads: Theoretical Development and Experimental Validation on Large Scale Tests, Earthquake Engineering & Structural Dynamics, 41(3): 475493. Gajan S, Hutchinson TC, Kutter BL, Raychowdhury P, Ugalde JA and Stewart JP (2008), Numerical Models for Analysis and Performance-based Design of Shallow Foundations Subject to Seismic Loading, Report No. PEER-2007/04, Pacic Earthquake Engineering Research Center, University of California, Berkeley. Gajan S, Raychowdhury P, Hutchinson TC, Kutter B and Stewart JP (2010), Application and Validation of Practical Tools for Nonlinear Soil-foundation Interaction Analysis,Earthquake Spectra, 26(1): 111129. Gazetas G (1991), Formulas and Charts for Impedances of Surface and Embedded Foundations, Journal of Geotechnical Engineering, 117(9): 13631381. Gazetas G and Mylonakis G (2001), SSI Effects on Elastic & Inelastic Structures, Proc. of Fourth Int. Conf. On Recent Advances in Geotechnical Earthquake Engineering & Soil Dynamics and Symposium in Honor of Professor W. D. Liam Finn, San Diego, California, March 2631, 2001. Gupta A and Krawinkler H (2000), Behavior of Ductile SMRFs at Various Seismic Hazard Levels,ASCE Journal of Structural Engineering, 126(1): 98107. Harden CW and Hutchinson TC (2009), Beam-onnonlinear-winkler-foundation Modeling of Shallow, Rocking-dominated Footings,Earthquake Spectra, 25. Meyerhof GG (1963), Some Recent Research on the Bearing Capacity of Foundations, Canadian Geotechnical Journal, 1(1): 1626. Nakaki DK and Hart GC (1987), Upliting Response of Structures Subjected to Earthquake Motions, U.S.-Japan Coordinated Program for Masonry Building Research,

Report No. 2.1-3. Ewing, Kariotis, Englekirk and Hart. NEHRP (2003), Recommended Provisions for Seismic Regulations for New Buildings, Building Seismic Safety Council, Washington, D.C. OpenSees (2008), Open System for Earthquake Engineering Simulation, Pacic Earthquake Engineering Research Center, PEER, Richmond (CA, USA). http:// opensees.berkeley.edu/. Raychowdhury P (2008), Nonlinear Winklerbased Shallow Foundation Model for Permormance Assessment of Seismically Loaded Structures, PhD thesis, University of California, San Diego. Raychowdhury P and Hutchinson TC (2009), Performance Evaluation of a Nonlinear WinklerBased Shallow Foundation Model Using Centrifuge Test Results, Earthquake Engineering and Structural Dynamics, 38: 679698. Raychowdhury P and Hutchinson TC (2010), Sensitivity of Shallow Foundation Response to Model Input Parameters, ASCE Journal of Geotechnical and Geoenvironmental Engineering, 136(3): 538541. Raychowdhury P and Hutchinson TC (2011), Performance of Seismically Loaded Shearwalls on Nonlinear Shallow Foundations, International Journal for Numerical and Analytical Methods in Geomechanics, 35: 846858. Singh P (2011), Performance Evaluation of Steelmoment-resisting-frame Building Incorporating Nonlinear SSI, Masters thesis, Indian Institute of Technology Kanpur, India. Somerville P, Smith N, Punyamurthula S and Sun J (1997), Development of Ground Motion Time Histories for Phase 2 of the FEMA/SAC Steel Project. http://www.sacsteel.org/project/. Stewart JP, Kim S, Bielak J, Dobry R and Power MS (2003), Revisions to Soil-structure Interaction Procedures in NEHRP Design Provisions, Earthquake Spectra, 19(3): 677696. Terzaghi K (1943), Theoretical Soil Mechanics, J. Wiley, New York. Xiong Jianguo, Wang Danmin, Fu Tieming and Liu Jun (1998), Nonlinear SSI-simplied Approach, Model Test Verication and Parameter Studies for Seismic and Air-blast Environment, Developments in Geotechnical Engineering, 83: 245259. Xiong Jianguo, Wang Danmin and Fu Tieming (1993), A Modied Lumped Parametric Model for Nonlinear Soil-structure Interaction Analysis, Soil Dynamics and Earthquake Engineering, 12: 273282. Yim SC and Chopra A (1985), Simplied Earthquake Analysis of Multistory Structures with Foundation Uplift, ASCE Journal of Structural Engineering, 111(12): 27082731.