Geoscience Reference
In-Depth Information
possible earthquake damage, depending on the seismicity of the investigated region,
deterministicapproachcanbepreferable(Erdiketal.,2004).Independentofthemethod-
ology adopted for the earthquake hazard evaluation, whether it is probabilistic or deter-
ministic, realistic recorded (Bommer and Acevedo, 2004) or simulated acceleration time
histories areneeded toconduct siteresponse analyses for the investigated area.
2. Input motion
One of the sources for the major uncertainty in ground response analysis arises from
the variability of the input earthquake motion (Boore, 2004). Adopting a probabilistic
approach, two hazard compatible alternatives, real and simulated acceleration time his-
tories, were investigated. For both options earthquake hazard compatibility with respect
to the expected fault type, fault distance, and magnitude was taken into account in the
selectionandinthegenerationoftheaccelerationtimehistories.Thesiteresponseanaly-
seswereconductedusing1Dequivalentlinearmodel(IdrissandSun,1992)forthesame
area and for different site profiles to evaluate the reliability and variability of earthquake
characteristics on theground surface.
2.1. REAL ACCELERATION RECORDS
In using previously recorded real acceleration time histories, scaling scheme of the
selectedaccelerationrecordsbecomesanimportantdecisionpoint.Toevaluatetheeffects
ofdifferentscalingoptions,theselectedsetofearthquakeaccelerationtimehistorieswere
scaled with respect to peak ground acceleration (PGA), peak ground velocity (PGV),
and Arias intensity (AI). Site response analyses for the same soil profiles are repeated
andthevariabilityofearthquakecharacteristicsonthegroundsurfaceinducedbyscaling
is evaluated. Scaling of input time histories was carried out in time-domain that involves
only the amplitude of the timeseries (Ansal et al.,2006a; Durukal et al., 2005, 2006).
The selected site is located in Golcuk; a town in the epicentre area of the 1999 Kocaeli
earthquake. Detailed site investigations were carried out in the town as a part of the
post-earthquake studies (Ansal et al., 2000). The selected soil profile extending down
to 45m depth is analysed by Shake91 (Idriss and Sun, 1992). The regional earthquake
hazard was estimated with earthquakes in the magnitude range of 7.0-7.5 that can be
characterised by PGA
s (Ambraseys, 1995; Joyner and Boore,
1988). AI is estimated as 2m/s for the magnitude and distance range considered based
on the empirical attenuation relationship proposed (Travasarou et al., 2003; Siyahi et al.,
2001).
=
0
.
35g, PGV
=
30cm
/
The ground motion sets were downloaded from PEER website (PEER, 2006). The
criteria used in the selection were magnitude range of M
w
=
5, strike slip earth-
quake mechanism, and site conditions with NEHRP (BSSC, 2001) site classification of
B/C boundary. The site distance range of 0-40km was considered with respect to four
increments as; 0-10km (11 records), 10-20km (14 records), 20-30km (10 records),
7
.
0-7
.
