A dataset for performance-based assessment of number of equivalent uniform stress cycles

Published: 12-01-2021| Version 2 | DOI: 10.17632/njg3gg4s8m.2
Kemal Onder Cetin,
Eray Altinci,
Habib Tolga Bilge


A critical review of currently existing studies regarding number of equivalent uniform stress cycle concept revealed that; i) significantly extended earthquake catalogs are now available, which may enable improved assessments of cyclic stress conversion of irregular transient earthquake records to uniform stress cycles, ii) based on the findings of Cetin and Bilge (2012), weighting factors (m values) of stress conversions are now known to be stress, strain and density state dependent, and iii) these weighting factors extend to ranges, which exceed the limits of earlier studies. Inspired by these, a semi-empirical probability-based model for the estimation of equivalent number of uniform stress cycles as functions of earthquake, site and performance (either strain or pore pressure-based) parameters is proposed by Cetin et al. (2020). The dataset presented herein is used for the purpose, which lists column-wise the region, the name and the magnitude of the earthquake event and the shear wave velocity of the upper 30 m of the strong ground motion recording site. Additionally, the closest, epicentral, and Joyner and Boore distances are documented followed by the weighting factor m used for estimating the equivalent number of uniform stress cycles. Last but not least, in the last four columns the directivity parameters also widely referred to as Summerville parameters are given.


Steps to reproduce

The database used for the purpose is composed of 4,134 accelerograms compiled from 150 earthquakes, which are documented in the Pacific Earthquake Engineering Research Center’s earthquake catalog (www.peer.berkeley.edu). Due to lack of a common time step in accelerograms, all data were decimated to a common time step of 0.02 seconds by first applying a low-pass filter having a corner frequency of 25 Hz (4th order Chebyshev) and then re-sampling it with time step of 0.02 seconds, consistent with the procedure followed by Liu et al (2001). By using a spreadsheet micro, transient acceleration time histories are converted to equivalent uniform stress cycles by following the procedure defined in Seed et al. (1975) and Liu et al. (2001). The number of uniform stress cycles are counted for 17 different m values (weighting factors) of 0.2, 0.25, 0.3, 0.37, 0.41, 0.45,0.5, 0.55, 0.6, 0.65,0.7,0.75, 0.8, 0.85, 0.9, 0.95 and 1.0. m= 0.37, 0.41 and 0.5 are particularly selected to enable direct comparisons with the predictions of Liu et al. (2001). The resulting database is composed of 70,278 N-data points.