Simulation of Endurance Time Excitations via Wavelet Transform

Mohammadreza Mashayekhi, Homayoon E. Estekanchi, Hassan Vafai

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

This paper puts forth a wavelet-based methodology for generating endurance time (ET) excitations. Conventional simulating practice expresses signals by acceleration values which are then computed via unconstrained nonlinear optimization. Dynamic characteristics of signals, including frequency content, are not represented directly in this type of variable definition. In this study, a new algorithm is developed to generate ET excitations in discrete wavelet transform (DWT) space. In this algorithm, signals are represented by transform coefficients. In addition, objective functions are modified in order to obtain transform coefficients and return the objective function values. The proposed method makes the filtering of the optimization variables possible so that insignificant variables can be eliminated. New excitations are generated in filtered DWT space. Different generating scenarios are used, and the results are then compared. Results show improvement in the generated excitations. It is also observed that a filtered DWT space brings about higher match with target acceleration spectra. Further, significance of generating more matched ET excitations in dynamic response assessment is examined through analyzing a multidegree of freedom structure.

Original languageEnglish (US)
Pages (from-to)429-443
Number of pages15
JournalIranian Journal of Science and Technology - Transactions of Civil Engineering
Volume43
Issue number3
DOIs
StatePublished - Sep 1 2019
Externally publishedYes

Keywords

  • Endurance time method
  • Nonlinear optimization
  • Time history analysis
  • Wavelet transform

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Geotechnical Engineering and Engineering Geology

Fingerprint

Dive into the research topics of 'Simulation of Endurance Time Excitations via Wavelet Transform'. Together they form a unique fingerprint.

Cite this