Software Calibration And Sensitivity Analysis Of Ordinary Standard Bridge Response

Abstract

Reliable seismic response models and software are necessary to support the assessment and design of California highway bridges. While nonlinear time history analysis using general purpose finite element tools typically give more accurate results compared to response spectrum and nonlinear static methods, the computed response is extremely sensitive to modeling parameters and analysis details. In addition, the computed time history response can vary significantly due to the finite element formulations implemented in different software packages. To address such differences, three-dimensional models of four ordinary standard bridges (OSBs) provided by Caltrans are analyzed using CSiBridge and OpenSees. With extensive calibration, these two software packages are able to give nominally similar results for modal and nonlinear static analyses, as well as for nonlinear dynamic analyses with multiple ground motions. Discrepancies between CSiBridge and OpenSees were found to lie primarily in the nonlinear abutment models and the cyclic concrete constitutive response of the columns. Additional sensitivity analyses carried out using the direct differentiation method (DDM) on only the OpenSees models confirm that the strength and stiffness of the abutments have a significant influence on the response of each OSB. From the software calibrations and sensitivity analyses, modeling guidelines and recommendations are made based on the choice of element formulation and constitutive model, and also the variability in response due to the choice of input parameter values.

Publication Date

1-1-2018

Publication Title

11th National Conference on Earthquake Engineering 2018, NCEE 2018: Integrating Science, Engineering, and Policy

Volume

5

Number of Pages

2994-3004

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

Socpus ID

85085496718 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/85085496718

This document is currently not available here.

Share

COinS