Chiou and Youngs PEER-NGA Empirical Ground Motion Model for ...

Chiou and Youngs PEER-NGA Empirical Ground Motion Model for ... Chiou and Youngs PEER-NGA Empirical Ground Motion Model for ...

peer.berkeley.edu
from peer.berkeley.edu More from this publisher
08.04.2013 Views

with increasing rupture depth. The data in the PEER-NGA data set are insufficient to test if Equation (13) performs better than piece-wise linear functions of RRUP, M and ZTOR. SITE EFFECTS Near-Surface Geology: The incorporation of the effects of near-surface geology or site classification has gone through an evolution in the past 10 years. At the beginning of this period, ground motion models typically contained a scaling parameter based on site classification (e.g. Boore et al., 1993), or presented different models for “rock” and “soil” sites (e.g. Campbell, 1993; Sadigh et al. 1997). Classification of recording sites into rock or soil sites varied among investigators. Boore et al. (1997) introduced the explicit use of the average shear wave velocity in the upper 30 meters, VS30, in the ground motion model. Abrahamson and Silva (1997) building on an earlier model by Youngs (1993) introduced the explicit modeling of non-linear site effects in the ground motion model. The model we have developed for incorporating near-surface geology combines these concepts. ln( site ref Site S 30 ref y ) = ln( y ) + f ( V , y ) (14) The parameter yref is the ground motion on the reference site condition derived from the source and path scaling models described in the previous section. The reference site shear wave velocity was chosen to be 1130 m/sec because of the initial use of site response data to develop the functional form (Silva, 2004) and because it is expected that there will not be significant nonlinear site response at this velocity. As indicated on Figures 4 and 5, there are very few data with values of VS30 greater than 1100 m/sec. The reference motion is defined to be the spectral acceleration at the spectral period of interest for two reasons. Bazzurro and Cornell (2004) indicate that the spectral acceleration at spectral period T is “the single most helpful parameter” for the prediction of site amplification at that period. In addition, the estimation of the coefficients of the ground motion model is performed using random (mixed) effects regression in which the reference motion includes the random event term representing the deviation of the average ground motions from a given earthquake from the global population mean. Use of the reference spectral acceleration at period T to estimate surface ground motions at the same period eliminates the need to include the correlation in the random effects between those at period T and those at another period, such as pga at “zero” period. The function form for the site response model fSite with yref = SA1130 is given by: where f Site ( V S 30 , T, SA 1130 ) = a( V a( V S 30 S 30 , T ) + b( V b( Vs30, T ) = φ ( T ) exp 2 ⎡ VS 30 ⎤ , T ) = φ1 ln⎢1130⎥ ⎣ ⎦ c( T ) = φ ( T ) ⎡ SA1130 ( T ) + c( T ) ⎤ , T ) ln⎢ c( T ) ⎥ ⎣ ⎦ { φ ( T ) × ( V − 360) } C&Y2006 Page 29 4 3 S 30 S 30 (15)

The interpretation of the parameters a, b, and c is illustrated in Figure 23. Parameter a represents the linear site response that occurs at small level of reference site motion. It is modeled as a linear function of ln[VS30] consistent with previous representations (e.g. Boore et al., 1997). Parameter c represents the ground motion level in the middle of the transition from linear to nonlinear behavior. Parameter b represents the nonlinear behavior in terms of a linear decrease in the natural log of site amplification, fS, with increasing amplitude of the reference motion. In general, a stronger nonlinearity in soil response corresponds to a more negative value for b (stronger dependence on SA1130). It is expected that the degree of nonlinearity is a function of the stiffness of the site soils and is represented by making b a function of VS30. fSite a c Figure 23: Soil amplification function ln(SA1130) The ability of Equation (15) to represent nonlinear site response is illustrated in Figures 24 and 25. Figure 24 shows site amplification factors for pga derived from Silva’s (2004) simulations. Two soil property models were used by Silva (2004), the EPRI (1993) set of soil modulus and damping relationships and the less nonlinear set Peninsular Range set developed by Silva et al. (1996). The solid lines shown on the figure are the result of fitting Equation (15) to the combined amplification factors for the two soil model sets and indicate that the function form can well represent the behavior implied by nonlinear (equivalentlinear) site response. Figure 25 shows the values of parameters a and b derived from the results of Silva (2004). The dashed lines show that the function forms for a and b in Equation (15) provide a good match to the site response results. C&Y2006 Page 30 b

The interpretation of the parameters a, b, <strong>and</strong> c is illustrated in Figure 23. Parameter a<br />

represents the linear site response that occurs at small level of reference site motion. It is<br />

modeled as a linear function of ln[VS30] consistent with previous representations (e.g. Boore<br />

et al., 1997). Parameter c represents the ground motion level in the middle of the transition<br />

from linear to nonlinear behavior. Parameter b represents the nonlinear behavior in terms of<br />

a linear decrease in the natural log of site amplification, fS, with increasing amplitude of the<br />

reference motion. In general, a stronger nonlinearity in soil response corresponds to a more<br />

negative value <strong>for</strong> b (stronger dependence on SA1130). It is expected that the degree of<br />

nonlinearity is a function of the stiffness of the site soils <strong>and</strong> is represented by making b a<br />

function of VS30.<br />

fSite<br />

a<br />

c<br />

Figure 23: Soil amplification function<br />

ln(SA1130)<br />

The ability of Equation (15) to represent nonlinear site response is illustrated in Figures 24<br />

<strong>and</strong> 25. Figure 24 shows site amplification factors <strong>for</strong> pga derived from Silva’s (2004)<br />

simulations. Two soil property models were used by Silva (2004), the EPRI (1993) set of<br />

soil modulus <strong>and</strong> damping relationships <strong>and</strong> the less nonlinear set Peninsular Range set<br />

developed by Silva et al. (1996). The solid lines shown on the figure are the result of fitting<br />

Equation (15) to the combined amplification factors <strong>for</strong> the two soil model sets <strong>and</strong> indicate<br />

that the function <strong>for</strong>m can well represent the behavior implied by nonlinear (equivalentlinear)<br />

site response. Figure 25 shows the values of parameters a <strong>and</strong> b derived from the<br />

results of Silva (2004). The dashed lines show that the function <strong>for</strong>ms <strong>for</strong> a <strong>and</strong> b in<br />

Equation (15) provide a good match to the site response results.<br />

C&Y2006 Page 30<br />

b

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!