Using Pegasus Scientific Workflows to Investigate the Effects of Material Property Variations on the Seismic Response of a Steel Special Moment Resisting Frame

Abstract

Material properties of structural steel alloy are variable. Variability may be observed in engineering parameters such as yield stress, modulus of elasticity and strain hardening modulus of the structural steel. Such variability may affect the seismic response of the steel structures. This research study focuses on use of scientific workflows using the Pegasus workflow system, which is available on the NEEShub (nees.org) as an OpenSees Laboratory tool (nees.org/resources/openseeslab) and can provide access to supercomputing resources through the Open Science Grid (OSG), to investigate the possible effects of steel engineering material properties on the simulated seismic response of a steel special moment resisting frame (SMRF). The Pegasus workflow system is used to develop a four-story four-bay steel SMRF model to run on OpenSees (opensees.berkeley.edu). The Pegasus workflow accepts separate coefficients of variation for steel yield stress, modulus of elasticity and ratio of modulus of strain hardening to modulus of elasticity of steel. The workflow then generates different material properties randomly within the defined ranges of the variability. These random material properties are then assigned to the girder and column elements of a series of steel SMRF OpenSees models. The response of these steel SMRFs to various earthquake ground motions are simulated using the OpenSees simulation platform. The output of the simulations are compared in terms of maximum roof displacement, the maximum story drift anywhere in the structure, the maximum roof acceleration response, and the maximum base shear demand. Individual and combined material engineering property variations are compared with results obtained from the reference structure that has no variation in its material properties and for which mean values the material properties are used. It is observed that variability in modulus of elasticity of steel, variability in combinations of modulus of elasticity and yield stress, and variability in yield stress respectively resulted in the highest variations in roof drift, story drift and base shear while variability in the modulus of strain hardening had minimal effect.