2846
Aziz Ahmed and Kiang Hwee Tan
A Novel Seismic Demand Envelope Development Process Aziz Ahmed and Kiang Hwee Tan1
Abstract Soft soil above the rock amplifies the seismic shear waves, which in turn impose large lateral loads on buildings with natural period close to the site period. Computation of amplification factor requires shear wave velocity of every soil layer. The formulae available in the literature to calculate the shear wave velocity are only valid for a particular region. Based on extensive literature review and field measurements, this paper proposes a simple and novel approach to calibrate the formula to determine the shear wave velocity for Singapore soil sites. Finally, this paper presents the development of seismic demand envelope for Singapore soil sites followed by its application to determine the vulnerability of a typical high rise building in Singapore. Keywords Kallang soil • Performance point • Seismic demand envelope • Soft soil amplification
1 Introduction In the global seismic hazard map, Singapore falls in a low seismic hazard zone. This is because Singapore is located on a stable part of the Eurasian Plate, with the nearest earthquake fault about 400 km away in Sumatra. More than 70% of the land area of the world has similar low seismic hazard classification. Taking advantage of this, buildings in Singapore are gravity-load designed (GLD) structures, according to BS8110 (1985), which does not have any seismic provision. However, due to the far-field effects of earthquakes in Sumatra (Balendra et al. 2002, 2003), buildings in Singapore, of which most are reinforced concrete (RC) shear wall and frame structures, occasionally suffers from tremors that occur at the Sumatra fault due to the subduction of the IndianAustralian plate into the Eurasian Plate. Aziz Ahmed ( ) • Kiang Hwee Tan Department of Civil and Environmental Engineering, National University of Singapore, Singapore 117576 e-mail:
[email protected] © Bloomsbury Publishing India Pvt Ltd., 2015 V. Matsagar (ed.), Modeling, Simulation and Analysis ISBN: 978-93-84898-72-4
2846
A Novel Seismic Demand Envelope Development Process
2847
Therefore, the need to evaluate the seismic vulnerability of buildings in Singapore in case a larger or nearer earthquake occurs, has been felt and studied to some extent (Balendra et al., 2007). In order to evaluate the seismic adequacy of such buildings, seismic demands due to base motions generated from the worst earthquake scenario need to be established. The procedure to determine the seismic demands for buildings in Singapore includes:(1) seismic hazard analysis to identify the worst earthquake scenario originating in Indonesia as the design earthquake;(2) establishment of attenuation relationships to obtain the bedrock motions at Singapore due to the design earthquake;(3) consideration of soil profile amplification to get the surface motions at the building base. Previous studies by Balendra et al. (2002, 2003) identified and proposed the bedrock motions that may originate from the worst case scenario earthquake of Moment magnitude (MW) = 9.5, occurring at the Sumatran subduction zone about 600 kM away from Singapore. That leaves us with the development of Seismic demand curve of Singapore based on soil profile amplification of the bedrock motion. Several studies have contributed to obtain seismic demands in Singapore. Pan (1995) pointed out that the response of a building to long-distance earthquakes is dependent on the type of structural systems and the local geological conditions. He found that tall buildings founded on Quaternary deposit, i.e. the Kallang Formation, are particularly vulnerable to the long-distance Sumatra earthquakes. Balendra et al. (2002) analyzed the seismic faults that cause long distance earthquakes for Singapore and identified the worst earthquake scenarios in each of these faults: a Mw=7.8 Sumatran-fault earthquake 400 km away from Singapore, and a Mw=8.9 Sumatran-subduction earthquake 600 km away. A validated seismic hazard predictive model estimated the bedrock motions in Singapore, that is, the Component Attenuation Model (CAM). Based on their findings, the elastic base shear demand corresponding to the bedrock motions when accounted for amplification by soft soil is 6%–10% of the weight of the building. To develop a seismic demand envelope, this study generates response spectra for a number of representative soil sites. In total ten soil sites, spread all over Singapore Island; provide the borehole data in order to develop the seismic demand envelope. First, the paper outlines the various soil formations in Singapore and identifies the most vulnerable soil type. Next, it presents the procedure to derive the response spectra using SHAKE91 (Idriss and Sun 1992) computer program, which leads to the development of a formula to calculate shear wave velocity for Singapore soil sites. Final part of the paper, briefs the procedure to derive the proposed seismic demand envelope for Singapore soil sites follows by an example of its use to determine the vulnerability of a typical highrise building in Singapore to far field earthquake. Table 1 Soil factor Clay
Fine sand
Med sand
Coarse Sand
Sand and Gravel
Gravel
1
1.09
1.07
1.11
1.15
1.45
2848
Aziz Ahmed and Kiang Hwee Tan
Fig. 1 Soil formation in Singapore (Pitts 1992)
2 Development of Seismic Demand Envelope for Singapore Soil Sites 2.1 Singapore Soil Types The soil on Singapore Island typically falls into one of the six major formations known as the Kallang formation, Old Alluvium, Jurong Formation, Bukit Timah Granite, Gombak Norite and Sajahat Formation (Pitts 1992). Figure 1 shows these formations based on Pitts (1992). The Old Alluvium is mainly in the eastern part of the Singapore Island. This formation has remained virtually uninterrupted at the surface since its deposition or is situated below the sediment layers which have been deposited later on. The Jurong formation comprises sedimentary rocks of the Jurassic age and covers most of the western part of Singapore. The Bukit Timah Granite covers one third of Central Singapore, in particular the central water catchment area. The Sajahat formation is a comparatively minor formation in central Singapore. Gombak Norite is a Paleozoic rock formation similar to Sajahat formation generally found in central Singapore. The Kallang formation, which isabundant in much of the coastal plain and immediate offshore zone, is a relatively recent deposit and comprises soil of marine, alluvial and estuaries origins. It is estimated that this formation covers one fourth of the Singapore Island (Pitts 1992). Marine Clay is the main constituent of Kallang formation. The marine clay formation varies in thickness; it is usually between 10 m to 15 m near the estuaries, but in some instances it
A Novel Seismic Demand Envelope Development Process
2849
can be thicker than 40 m. The marine clay is deposited in two layers, typically referred to as the upper and lower marine clay, separated by a stiffer intermediate layer, widely considered to be the dried out crust of the lower marine clay. As Kallang soil formation comprises most of Singapore’s inhabited area and it is mainly comprised of soft marine clay which can significantly amplify the bed rock motion, this study mainly focuses on Kallang formation.
2.2 Generation of Response Spectra Ahmed (2012) presents details of the bedrock accelerograms and soil profiles of the selected sites (Fig. 1). Current study uses SHAKE91 to calculate the acceleration response spectra at surface. This program bases on equivalent linear seismic response analyses of horizontally layered soil deposits. It is based on the original SHAKE program published by Schnabel, Lysmer and Seed (1972). In SHAKE91, the soil profile is idealized as a system of homogeneous, viscoelastic sublayers of infinite horizontal extent. The response of this system is calculated by considering vertically propagating shear waves. An equivalent linear procedure is used to account for the nonlinearity of the soil using an iterative procedure to obtain values for modulus and damping, which are compatible with the equivalent uniform strain induced in each sub-layer. The assumptions of the analysis are: 1. Each sublayer is completely defined by its shear modulus, critical damping ratio, total unit weight and thickness. These values are independent of frequency. 2. Responses in the soil profile are caused by the upward propagation of shear waves from the underlying rock formation, which are specified as acceleration ordinates at equally spaced time intervals. However, there is no universal formula to calculate the shear wave velocity of a soil layer, thus it is a general practice to select or develop an appropriate formula to calculate shear wave velocity for a particular study. This study uses the formula by Ohta and Goto (1978) to develop the shear wave velocity formula for Kallang formation. The formula by Ohta and Goto (1978) is as follows: Vs = C1 N SPT
0.17
0.19
Z F1 F2
(1)
where, C is an empirical constant, Ohta and Goto (1978) suggested 69 NSPT symbolizes N value of the standard penetration test for soil (blow/30 cm) is taken Z refers to depth of soil in m where blow count F is the age factor, 1 for Holocene age (alluvial deposits) i.e. younger soil, 1.3 for Pleistocene age (dilluvial deposits), older soil F is the soil factor, the value are given in Table 1.
2850
Aziz Ahmed and Kiang Hwee Tan
Also, Seed et al. (1986) adopted this formula and Jamiolkowskiet al. (1988) later applied it in the following form. 0.17
Vs = C2 N 60 Z FF 1 2 0.19
(2)
where, C refers to an empirical constant, for which Jamiolkowski et al. (1988) used 53.5; and N represents the value of NSPT corrected for field procedures and apparatus In developing Eq.1, Ohta and Goto (1978) compared fifteen sets of empirical equations to estimate shear wave velocity by combining the four parameters in the equation. All sets were derived using about three hundred data, and the correlation coefficients between the measured and estimated shear wave velocities evaluate their accuracies. Eq.1produced the best correlation coefficient of 0.86 and included all the indexes. However, the authors stressed that this equation is only valid for the soil types they investigated and requires appropriate calibration for other soil sites. On the other hand, Jamiolkowski et al. (1988) used Eq.2 to calculate the shear wave velocity for Italian cohesion-less soil deposits and compared it with measured values which yielded reasonably accurate results for low N values. This study, calibrates Eq. 1 for Singapore soil based on measured data available in the literature (Leong et al. 2003; Pan et al. 2002). To do this, this paper proposes a correction factor to convert NSPT values to N is derived for Singapore context. Next, the comparison of response spectrum curves generated using the new formula and the previous ones demonstrates the effect of newly calibrated formula. Finally, using the response spectra generated by SHAKE91 analysis and adopting the ATC-40 (1996) recommended procedure; this study proposes a demand envelope for Singapore soil types.
2.3 Conversion of NSPT Value to
Value
NSPT data is corrected for a number of site specific factors to improve its repeatability. Burmister’s 1948 energy correction (1948) assumed that the hammer percussion system was 100% efficient (a 140-lb (63.5 kg) hammer dropping 30 inches (0.76 m) = 4,200 ft-lbs (5694 J) raw input energy). In a much later publication, the procedures for determining a standardized blowcount were presented (1986), which allow hammers of varying efficiency to be accounted for. This corrected blow-count is referred to as "N60", because the original NSPT (Mohr) hammer has about 60% efficiency, and this is the “standard” to which other blow-count values are compared. N60 is given as: N 60 = C E C B C S C R C N N SPT
where, CE is the correction factor for hammer efficiency; CB refers to the borehole diameter correction factor;
(3)
A Novel Seismic Demand Envelope Development Process
2851
CS represents the sample barrel correction factor; CR symbolizes the rod length correction factor; CN displays over burden pressure correction factor. Overburden pressure for Singapore soil is generally very low, so it can be considered equal to atomospheric pressure; hence: CN = PA /σ′ . = 1, where, represents atmospheric pressure and ′ refers to overburden pressure. Table 2 lists correction factors to account for these uncertainties (Sherif and Radding 2001). Based on general practice; the drilling machines used in Singapore can be regarded to have accuracy equivalent to that of a safety hammer. From the table the range of corresponding correction factor for energy is found to be 0.7~1.2. Hence, an average value of CE = 0.95 is ratio i.e. assumed. Generally, boreholes used in Singapore are 100 mm in diameter. It is within the range 65–115 mm; so correction factor to account for difference in borehole is equal to one. The length of the rods used in Singapore for SPT diameter, test is 15.5 m which is within the range of 10–30 m. Hence, correction factor to account for difference in sampling rod length, CR is also equal to one. Also correction factor for sampling method, CS is equal to one as generally standard samplers are used in Singapore. Table 2 Correction Factors for Calculating N
Factor
Equipment Variable
Energy ratio CE
Donut hammer Safety hammer Automatic hammer
Bore hole diameter CB
65–115 mm 150 mm 200 mm
Sampling method CS
Standard sampler Sampler without liners
Rod length CR
3–4 m 4–6 m 6–10 m 10–30 m >30 m
Over burden pressure CN
N/A
(Sherif and Radding, 2001)
Correction Factors 0.5–1.0 0.7–1.2 0.8–1.5 1.0 1.05 1.15 1.0 1.1–1.3 0.75 0.85 0.95 1.0
