You are on page 1of 11
10NCEE
10NCEE

Tenth U.S. National Conference on Earthquake Engineering Frontiers of Earthquake Engineering July 21-25, 2014 Anchorage, Alaska

SEISMIC RESPONSE OF HILL BUILDINGS SUBJECTED TO BI-DIRECTIONAL EXCITATION

Y. Singh 1 , V.R. Yeluguri 2 , and D.H. Lang 3

ABSTRACT

In the present study, the term ‘Hill Buildings’ refers to buildings located on hill slopes with foundations placed at different elevations. The structural configurations of these buildings are quite different as compared to regular buildings located on flat terrain. Hill buildings are characterized by stiffness and strength irregularities which are mainly due to the presence of short stiff columns on the building’s uphill side. In order to identify typical structural configurations of hill buildings in the Indian Himalayas, field studies were conducted in the northern Indian cities of Mussoorie and Nainital. The step-back configuration, which is the most prevalent building configuration in Indian hill regions, has been chosen for detailed investigation. A generic building with step-back configuration is designed for different design levels as per relevant Indian codes. The seismic performance of the various building models has been studied using bi-directional incremental dynamic analysis (IDA). It is observed that the primary failure mode is governed by shear failure of short columns, while it can be generally said that hill buildings possess much lower seismic capacity than their counterparts located on flat ground.

1 Professor, Dept. of Earthquake Engineering, Indian Institute of Technology Roorkee, Roorkee – 247 667, INDIA 2 Graduate Student, Dept. of Earthquake Engineering, Indian Institute of Technology Roorkee, Roorkee – 247 667, INDIA 3 Senior Research Engineer, NORSAR, Kjeller 2027, NORWAY

Singh Y, Yeluguri VR, Lang DH. Seismic response of hill buildings subjected to bi-directional excitation. Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

SEISMIC RESPONSE OF HILL BUILDINGS SUBJECTED TO BI-DIRECTIONAL EXCITATION

Y. Singh 1 , V.R. Yeluguri 2 , and D.H. Lang 3

ABSTRACT

In the present study, the term ‘Hill Buildings’ refers to buildings located on hill slopes with foundations placed at different elevations. The structural configurations of these buildings are quite different as compared to regular buildings located on flat terrain. Hill buildings are characterized by stiffness and strength irregularities which are mainly due to the presence of short stiff columns on the building’s uphill side. In order to identify typical structural configurations of hill buildings in the Indian Himalayas, field studies were conducted in the northern Indian cities of Mussoorie and Nainital. The step-back configuration, which is the most prevalent building configuration in Indian hill regions, has been chosen for detailed investigation. A generic building with step-back configuration is designed for different design levels as per relevant Indian codes. The seismic performance of the various building models has been studied using bi-directional incremental dynamic analysis (IDA). It is observed that the primary failure mode is governed by shear failure of short columns, while it can be generally said that hill buildings possess much lower seismic capacity than their counterparts located on flat ground.

Introduction

The scarcity of flat lands in hilly areas can be seen as a major challenge for the construction of conventional buildings. The situation in the Indian Himalayas is particularly acute, as the Himalayan hills, being among the youngest in the world, are characterized by very steep slopes while, at the same time, this region is one of the densely populated parts of India. In order to create flat dwelling spaces, people have developed a variety of ingenious structural configurations, which are not covered by any building standard. In the present study, these often highly irregular structural configurations are collectively termed as ‘Hill Buildings’. Apart from the amplification of seismic ground motion due to the geometry of the topographic features as well as earthquake-induced slope failure (landslides, rock falls), these buildings have an intrinsically increased vulnerability due to their irregular structural configuration alone.

For the present study, field surveys were conducted in the cities of Mussoorie and Nainital (both located in the state of Uttarakhand, Northern India) in order to identify the

1 Professor, Dept. of Earthquake Engineering, Indian Institute of Technology Roorkee, Roorkee – 247 667, INDIA 2 Graduate Student, Dept. of Earthquake Engineering, Indian Institute of Technology Roorkee, Roorkee – 247 667, INDIA 3 Senior Research Engineer, NORSAR, Kjeller 2027, NORWAY

Singh Y., Yeluguri VR., Lang DH. Seismic response of hill buildings subjected to bi-directional excitation. Proceedings of the 10 th National Conference in Earthquake Engineering, Earthquake Engineering Research Institute, Anchorage, AK, 2014.

archetypical configurations that are representative for this region. Out of the six different configurations that were identified during the study, the most commonly used has been considered for detailed investigation. Since current design codes do not address the design of buildings with these types of structural configurations, the buildings have been designed for different ‘code levels’ following the provisions of existing codes that address buildings on flat topography. In addition, different ground shaking intensities are considered and the response of the designed buildings investigated. As the chosen structural configuration is irregular both in plan and elevation, it is subjected to bi-directional ground motion. Seven records from the PEER ground motion database [1] have been selected that best match the spectral shape of the design response spectrum as per Indian seismic design code IS : 1893 (Part 1) - 2002 [2]. The acceleration time histories were further scaled with respect to the spectral acceleration value of the major component of the ground motion, at the fundamental period of the building in the considered direction. Incremental Dynamic Analysis [3] is performed in both orthogonal directions of the building models in order to evaluate the sustained intensity of ground motion before collapse.

Identification of Representative Structural Configurations

A typology survey of buildings in the cities of Mussoorie and Nainital was conducted under the Indo-Norwegian EQRisk collaboration project [1]. In typology surveys for conventional buildings located on flat ground, vertical (wall type) and horizontal (floor and roof type) load- bearing systems are the main classification criteria [4-5]. However, in case of hill buildings, it was observed that identifying wall and roof/floor types only, is not adequate, as the building’s geometric configuration also affects its seismic response. Therefore, the geometric configuration was also considered as yet another classification criterion. During the survey, it was observed that in case of steep slopes, it is not possible to have foundations at the same level. In order to create a flat foundation ground, deep cutting of soils or rocks is required. This procedure is not only economically unviable, it also creates vertical cuts into the soil/weathered rock slopes which may dramatically increase susceptibility of slope failures, even without earthquake action. Due to these reasons, there is a general tendency among the local builders in following the terrain of the hill for supporting the buildings. This has given ri se to a number of interesting support systems, some of them given in Fig. 1. Fig. 2 schematically illustrates the typical geometric configurations commonly observed during the survey, which can be classified into six main categories based on the arrangement of supports to the building. The most natural structural configuration is to accommodate the shape of the slope in foundation arrangement through a step-back configuration (SC Types C and D, Fig. 2). This is basically achieved by providing separate foundations at gradually increasing levels. In some cases, the superstructure, in addition, follows the shape of the slope, resulting in a combination of step-back and set-back configurations (SC Type D, Fig. 2).

In case of very steep slopes on rocky terrain, a conventional solution for hill buildings is to provide foundations at two (dual) levels. Few of the columns or walls are supported at or close to the uphill road level, whereas the remaining columns or walls are supported downhill at the toe of the slope (SC Type B, Fig. 2). These buildings may have several stories below and a few above the road level. To create this configuration type, it is required to cut the soil in order to prepare the plane patches of land supporting individual or a group of columns or walls. The

vertical cuts in hard rock may sustain themselves against sliding, but in case of soil and weathered rock strata, parts of the buildings have to retain the backfill, resulting in lateral pressure on the vertical building components. Retaining of the soil by the building can be avoided by creating a flat platform above the slope made of landfill, instead of cutting the soil (SC Type E, Fig. 2). However, this comes along with the disadvantage that the filled up platform requires a retaining wall on the downhill side to support the filling. Different locations of the retaining wall with respect to the buildings’ outer edges have been observed during the survey (Fig. 2e). The most convenient, but perhaps the most vulnerable configuration to create flat base for a building is to use columns or posts (stilts, Fig. 2f) of varying height, resulting in extremely irregular configurations.

vertical cuts in hard rock may sustain themselves against sliding, but in case of soil and
vertical cuts in hard rock may sustain themselves against sliding, but in case of soil and
(a) (b) (c) (d)
(a)
(b)
(c)
(d)

Figure 1. Commonly observed support systems and geometric configurations of buildings in hilly areas in Northern India: (a) downhill side supported on stilts, (b) a flat platform created by a retaining wall on the downhill side, (c) step-back configuration, and (d) step-back and set-back configuration.

SC Type A (a) SC Type C (c) SC Type E (e) SC Type B (b)

SC Type A

SC Type A (a) SC Type C (c) SC Type E (e) SC Type B (b)
(a) SC Type C (c) SC Type E (e)
(a)
SC Type C
(c)
SC Type E
(e)
SC Type A (a) SC Type C (c) SC Type E (e) SC Type B (b)

SC Type B

SC Type A (a) SC Type C (c) SC Type E (e) SC Type B (b)
(b) SC Type D
(b)
SC Type D
(d) SC Type F
(d)
SC Type F

(f)

Figure 2. Schematic representation of different structural configurations of buildings, commonly found in hilly areas of Northern India: (a) foundation of the entire building on the same level, (b) building is founded on two different levels, (c) step-back configuration, (d) step-back and set-back configuration, (e) building is (partly) founded on an artificially created platform that is supported by a retaining wall on the downhill side, and (f) parts of the buildin g are supported by stilts on the downhill side.

The most common geometric configuration, out of those described above, is the step- back configuration whose foundation structure more or less follows the natural shape of the slope. This configuration has been considered for further investigation in the present study. As in case of regular buildings on flat terrain, wall and floor/roof types play an important role in the seismic response of hill buildings, in addition to the geometric configuration. In the hilly areas of Northern India, reinforced-concrete frame construction is one of the most common type of contemporary construction, as it can be easily adapted to the undulating terrain. The size and

plan of the building, arrangement of columns, as well as spans of the beams vary widely, depending on the available land and usage of the building. Based on the observations made during the survey, a generic building configuration with plan and elevation as shown in Fig. 3 has been considered in the present study. The columns of the building have been assumed to be founded on rock and accordingly considered to be fixed at their base. It can be observed from Fig. 3, that the building plan is symmetric in both general axes. However, due to the varying heights of the columns along the slope, an inherent asymmetry in the direction across the slope is observed.

plan of the building, arrangement of columns, as well as spans of the beams vary widely,
(a) (b)
(a)
(b)

Figure 3. Building configuration considered in the present study: (a) plan, and (b) elevation.

Two design levels of the building have been considered. As many buildings in India are still constructed without any consideration for earthquake forces, the first model is designed for gravity loads only, and designated as ‘GLD’ (Gravity Load Design). The second model is designed according to provisions of the current Indian seismic design code IS : 1893 - 2002 [2], which are actually applicable to buildings on flat terrain only. In addition to the usual gravity loads, earthquake forces corresponding to Indian seismic zone IV (Zone Factor 0.24 g) are

considered. The model has been detailed for ductility requirements of ‘Special Moment Resisting Frames (SMRF)’ according to IS : 13920 - 1993 [6]. It is to be noted that the SMRF of Indian code is equivalent to Intermediate Moment Resisting Frame (IMRF) of ASCE 7/ACI 318 [7, 8] in terms of the reinforcement detailing and response reduction factor [9]. This is the most ductile category as per Indian code, and there is no building class in Indian code that would correspond to ‘SMRF’ of ASCE 7. Table 1 shows the typical member dimensions for two design levels of the building. The dimensions have been proportioned so that the columns have longitudinal reinforcement in the range 2–4% of the cross-sectional area, whereas the beams have longitudinal reinforcement around 1% on each face.

Table 1. Typical sizes of the members in buildings with different design levels.

Members

GLD Building

SMRF Building

All beams

  • 0.23 m × 0.30 m

  • 0.23 m × 0.30 m

Short columns

  • 0.35 m × 0.35 m

  • 0.50 m × 0.50 m

All other columns

  • 0.35 m × 0.35 m

  • 0.40 m × 0.40 m

Modeling and Analysis

Nonlinear space frame models of the two buildings are developed using the software SAP 2000 Nonlinear [10]. Beams and columns are modeled using frame elements with lumped plasticity. The rigid diaphragm action of floor and roof slabs is simulated using diaphragm constraints on all the nodes at one floor level. Effective stiffness of the cracked RC members is considered according to ASCE 41 Update [11] and so is the joint stiffness. Flexural plastic hinges are assigned at both ends of the beams, whereas P-M2-M3 interactive plastic hinges are assigned at the ends of the columns. Shear hinges are also assigned at the mid-height of the columns, mainly to simulate the shear failure of short columns on the uphill side. The properties of plastic hinges are defined according to ASCE 41 [12] guidelines. The transverse reinforcement categories ‘Non-conforming (NC)’ and ‘Conforming (C)’ are used for ‘GLD’ and ‘SMRF’ buildings, respectively. Shear failure of columns is considered as being force-controlled [12].

Due to variable column heights, both building models are considered to be irregular in elevation leading to torsional effects. Therefore, both are subjected to bi-directional excitation in order to account for the torsional coupling of the response in the two directions. Nonlinear dynamic (time-history) analysis is performed using the bi-directional components of seven recorded ground motion time histories. The time histories have been selected from the PEER Ground Motion Database [1] according to best match with the response spectrum of Indian seismic design code. The ground motions have been further scaled with respect to the spectral acceleration of the major component, corresponding to the fundamental period of the building in the two directions. Fig. 4 shows the response sp ectra of the major and minor components of the chosen time histories, scaled to match the design spectral acceleration at fundamental periods (T = 1.09 sec) of the SMRF building in longitudinal (along slope) direction. The component of the

ground motion having higher spectral acceleration corresponding to the fundamental period of the building is considered as the major component. Similar scaling is done with respect to the fundamental period in the other general direction and for the GLD building. The ground motions are applied with major components oriented alternately along the two directions. The records are gradually scaled to different levels, performing Incremental Dynamic Analysis (IDA) [3] of the developed nonlinear models of the buildings.

ground motion having higher spectral acceleration corresponding to the fundamental period of the building is considered

(a)

ground motion having higher spectral acceleration corresponding to the fundamental period of the building is considered

(b)

Figure 4. Response spectra of the seven recorded time histories scaled to S a (T 1 , 5%) of the SMRF building in along-slope direction: (a) major component (applied along the slope), and (b) minor component (applied across the slope). The gray curves show the response spectra of the individual recorded time histories, whereas the black curve represents the design response spectrum for rock and hard soil following the provisions of Indian seismic design code IS : 1893 - 2002 [2].

Results and Discussion

The IDA curves of the building models designed for two different levels (GLD and SMRF) are shown in Fig. 5. The curves remain almost linear right up to the incipient failure, which is considered as one of the three conditions whichever is reached first: (i) shear failure of short columns at ground level, (ii) flexural hinging of a sizable number of beams and columns resulting in unstable mechanism, or (iii) plastic rotation in a sizeable number (>50%) of members reaching the collapse limit state as per ASCE 41 [12]. The average maximum spectral acceleration and corresponding average PGA of the seven time histories, sustained by the buildings at incipient collapse, have been compared with the corresponding values used in design, and provided in Table 2. It is to be noted that the zone factor given in the code does not represent PGA but an average value of Effective Peak Ground Acceleration (EPGA) corresponding to the design ground shaking level. EPGA has slightly lower values than PGA and is defined in terms of average peak spectral acceleration in the short period range [2, 13].

(a) (c) (b) (d) Figure 5. Incremental Dynamic Analysis (IDA) curves for: (a) GLD building with

(a)

(a) (c) (b) (d) Figure 5. Incremental Dynamic Analysis (IDA) curves for: (a) GLD building with

(c)

(a) (c) (b) (d) Figure 5. Incremental Dynamic Analysis (IDA) curves for: (a) GLD building with

(b)

(a) (c) (b) (d) Figure 5. Incremental Dynamic Analysis (IDA) curves for: (a) GLD building with

(d)

Figure 5. Incremental Dynamic Analysis (IDA) curves for: (a) GLD building with major ground motion component along the slope, (b) GLD building with major ground motion component across the slope, (c) SMRF building with major ground motion component along the slope, and (d) SMRF building with major ground motion component across the slope. S a (T 1 , 5%) denotes the 5% damped elastic spectral acceleration of the major ground motion component, corresponding to the fundamental period in the building in considered direction, and θ denotes the maximum interstory drift.

It can be observed from Fig. 5 and Table 2 that the buildings with step-back configuration have much lower seismic capacity than expected. RC frame buildings on flat ground, designed as per Indian codes have been found [14] to possess sufficient capacity to sustain the Maximum Considered Earthquake (MCE) of the zone for which those have been designed, whereas in case of the considered hill buildings, the buildings are not able to even sustain the Design Basis Earthquake (DBE) though they have been designed for. Their performance is particularly poor in

the across-slope direction, being caused by torsional irregularities, resulting in failure at even lower intensities of shaking. The main reason of failure of these buildings, in both the cases of excitation (along and across the slope), is brittle shear failure of short columns at ground level (Fig. 6). The failure at intensities of shaking lower than that used for design, indicates that the current seismic design provisions for buildings on flat ground are not adequate for the design of hill buildings and special considerations need to be given to these building typologies. Further, the major portion of the existing building stock on hills, which is not designed for seismic forces, has very little seismic resistance and is, theoretically, expected to collapse at ground shaking levels as low as 0.02g.

the across-slope direction, being caused by torsional irregularities, resulting in failure at even lower intensities of

(a)

the across-slope direction, being caused by torsional irregularities, resulting in failure at even lower intensities of

(b)

Figure 6. Typical failure pattern of SMRF building model: (a) major component of ground motion applied along the slope, and (b) major component of ground motion applied across the slope.

Table 2. Design and sustained levels of ground shaking

Design

Major

Design level of shaking

Sustained level of shaking

Level

component

       

of shaking

S a (T 1 , 5%)

EPGA

S a (T 1 , 5%)

PGA

GLD

Along-slope

0.03g

0.04g

Across-slope

0.02g

0.03g

SMRF

Along-slope

0.11g

0.12g

0.10g

0.11g

Across-slope

0.07g

0.12g

0.06g

0.10g

Conclusions

Typical hill building configurations have been identified based on a field survey in two north- Indian Himalayan towns. One of the most commonly used building configurations in hills has been investigated for expected seismic performance. It was found that the hill building configuration sustains much lower level of ground shaking than the level of shaking for which it

was designed for, using the code provisions for buildings located on flat ground. This highlights the need for development of special design provisions for hill buildings. Further, the majority of the existing building stock, which has not been designed for seismic forces at all, has very little seismic resistance and requires strengthening measures to avoid catastrophic collapse even under a minor earthquake.

Acknowledgments

The current study was conducted under the EQRisk (Earthquake Risk Reduction on the Indian Subcontinent) project which is funded by the Ro yal Norwegian Embassy to India (New Delhi) and administered by the Research Council of Norway (Oslo).

References

  • 1. PGMD. Peer Ground Motion Database (Beta Version updated November 8, 2011), University of California, Berkeley, Ca 94720-1792 (http://peer.berkeley.edu/Peer_Ground_Motion_Database), 2011.

  • 2. IS : 1893 (Part 1) - 2002, Criteria for earthquake resistant design of structures -general provisions and buildings. Bureau of Indian Standards, New Delhi, India, 2002.

  • 3. Vamvatsikos D, Cornell CA. Incremental dynamic analysis. Earthquake Engineering & Structural Dynamics 2013; 31 (3): 491–514.

  • 4. Haldar P, Singh Y, Lang DH, Paul DK. Comparison of seismic risk assessment based on macroseismic intensity and spectrum approaches using ‘SeisVARA’. Soil Dynamics and Earthquake Engineering 2013; 48: 267–281.

  • 5. National Building Institute and the Federal Emergency Management Agency (NIBS and FEMA), HAZUS-MH MR2 Technical Manual, Washington, D.C., 2006.

  • 6. IS : 13920 - 1993, Indian standard ductile detailing of reinforced concrete structures subjected to seismic forces – code of practice. Bureau of Indian Standards, New Delhi, 1993.

  • 7. ASCE 7-10. Minimum design loads for buildings and other structures. American Society of Civil Engineers, Reston, Virginia, 2010.

  • 8. ACI 318. Building code requirements for Reinforced Concrete. American Concrete Institute, Detroit, 2011.

  • 9. Khose, VN, Singh, Y and Lang, DH. A Comparative Study of Design Base Shear for RC Buildings in Selected Seismic Design Codes, Earthquake Spectra 2012; 28(3): 10471070.

    • 10. SAP2000. CSI analysis reference manual for SAP2000. Computers and Structures Inc., Berkeley, California,
      2010.

    • 11. ASCE/SEI-41 Supplement-1. Update to ASCE/SEI 41 concrete provisions. American Society of Civil Engineers, Reston Virginia, 2007.

    • 12. ASCE 41-06. Seismic rehabilitation of existing buildings. American Society of Civil Engineers, Reston, Virginia, 2007.

    • 13. Kramer SL. Geotechnical Earthquake Engineering. Prentice Hall, Inc., Upper Saddle River, New Jersey, 1996.

    • 14. Haldar P, Singh, Y. Seismic performance and vulnerability of Indian code designed RC frame buildings. ISET Journal of Earthquake Technology 2009; 46(1): 2945.