PSI - Issue 78

Atilla Ansal et al. / Procedia Structural Integrity 78 (2026) 2133–2140

2136

of possible variation is ± 20%, and the assigned layer thicknesses are the mean possible variation of ± 10%. 100 soil profiles were generated for each soil boring. 3. Probabilistic Approach for Estimating Acceleration time Histories on the Ground Surface In conducting time history dynamic analysis for engineering structures, one popular option is to select acceleration time histories compatible with the design spectrum for the engineering structure. However, this approach neglects the need for representative time histories with known exceedance probabilities. Even though the selected representative acceleration time histories may be scaled with respect to probabilistic acceleration spectrum or peak ground acceleration obtained based on probabilistic site response analysis; the probability of the selected acceleration time histories are unknown. The only possible option is to estimate probabilistic acceleration time histories with predetermined exceedance probabilities to enhance fully probabilistic site and structural response analysis. One option is to use the calculated acceleration time histories on the ground surface from site response analysis. However, there is no straight forward procedure to determine the probability of these calculated acceleration time histories on the ground surface. The approach presented below is partially empirical with the aim of selecting the best fitting procedure to the structural design spectrum. In the site response analyses conducted for this manuscript for one soil profile from Istanbul with 100 soil profiles generated by Monte Carlo simulation with respect to layer thickness and layer shear wave velocity and using 20 hazard compatible acceleration time histories. Following these 100 site response analysis with 20 input acceleration records, yielding 2000 acceleration time histories on the ground surface. At this stage an optimization scheme was utilized to determine the best fitting acceleration time histories with respect to their acceleration response spectra in comparison to the 10% exceedance design spectrum defined for the engineering structures as shown in Figure 4.

0.7

10% EXCEDANCE SPECTRUM 1-1163 (0.791963) 2-623 (0.797138) 3-843 (0.812126) 4-1968 (0.812398) 5-1023 (0.813554)

0.7

0.6

0.6

AVERAGE of 2000 spectra 10% EXCEEDANCE for all 2000 spectra

0.5

0.5

0.4

0.4

6- 83 (0.815988) 7-403 (0.816527) 8-1683 (0.820104) 9-203 (0.822878) 10-903 (0.823731)

0.3

0.3

0.2

0.2

0.1 SPECTRAL ACCELERATIONS (g)

0.1

SPECTRAL ACCELERATION (g)

0

0

0

0.5

1

1.5

2

2.5

3

3.5

4

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

(b)

PERIOD (s)

(a)

PEROD (s)

Figure 4. (a) 10% design spectrum and average spectrum for all the calculated 2000 spectra based on 1D site response analysis (b) comparison of the best fitting 10 acceleration spectra with respect 10% exceedance spectrum of the acceleration time histories calculated by site response analyses The 6 calculated acceleration time histories determined based on the best fits as given in Figure 4b are shown in Figure 5. As can be observed the selected acceleration time histories are slightly different from each other and look reasonable for conducting dynamic structural analysis. Even though the approach adopted is not totally probabilistic, it may still be considered that the acceleration time histories thus selected have probabilistic background preferably in comparison to randomly selecting previously recorded acceleration time histories.