DEVELOPMENT OF A CLOSED-FORM EQUATION AND FRAGILITY CURVES FOR PERFORMANCE-BASED SEISMIC DESIGN OF GLASS CURTAIN WALL AND STOREFRONT SYSTEMS

Open Access
Author:
O'Brien Jr., William Charles
Graduate Program:
Architectural Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
March 18, 2009
Committee Members:
  • Ali M Memari, Thesis Advisor
Keywords:
  • Curtain Walls; Storefronts; Glass; Design; Seismic
Abstract:
This report provides research that will eventually lead to seismic design guidelines for glazing systems that will be utilized by professionals as a way to mitigate glass damage. Researchers at The Pennsylvania State University and University of Missouri have conducted many experimental studies on various curtain wall and storefront configurations, and based on the results from this extensive database a new closed-form equation and fragility curves were developed. Furthermore, conditions between the laboratory and field were investigated for practical application of the results. To expand the set of existing experimental data, new testing on glass curtain walls with various glass-to-frame clearances was performed to study the effect that glass –to-frame clearances specifically have on the seismic performance of glass panels. Furthermore, sensor testing was conducted on the racking facility to measure if any significant flexibility existed. Fragility curves were developed for twenty-four different glass configurations as a way to predict the seismic performance of glass in a probabilistic manner for gasket, cracking, and glass fallout damage states according to economic and life safety consequences. Then, a closed-form equation was formulated to predict the seismic cracking drift of a glass system. The equation uses the ASCE equation as its base, and then considers effects from glass type, glazing configuration type, substandard clearances, frame system type, and aspect ratio through the application of defined factors. An analysis showed that the proposed equation increases the accuracy of failure prediction by 33% compared to the ASCE equation.