Vulnerability of buildings to disproportionate (or progressive) collapse has
become an increasingly important performance issue following the collapses of the
Alfred P. Murrah Federal Building in Oklahoma City in 1995 and the World Trade
Center in 2001. Although considerable research has been conducted on this topic, there are still numerous unresolved research issues. This dissertation is aimed at developing structural models and analysis procedures for robustness assessment of steel building structures typical of construction practices in the United States, and assessing the performance of these typical structures.
Beam-column connections are usually the most vulnerable elements in steel buildings structures suffering local damage. Models of three typical frame connections for use in robustness assessment have been developed with different techniques, depending on the experimental data available to support such models. A probabilistic model of a pre-Northridge moment-resisting connection was developed through finite element simulations, in which the uncertainties in the initial flaw size, beam yield strength and fracture toughness of the weld were considered. A macro-model for a bolted T-stub connections was developed by considering the behavior of each connection element individually (i.e. T-stub, shear tab and panel zone) and assembling the elements to form a complete connection model, which was subsequently calibrated to experimental data. For modeling riveted connections in older steel buildings that might be candidates for rehabilitation, a new method was proposed to take advantage of available experimental data from tests of earthquake-resistant connections and to take into account the effects of the unequal compressive and tensile stiffnesses of top and bottom parts in a connection and catenary action.
These connection models were integrated into nonlinear finite element models of structural systems to allow the effect of catenary and other large-deformation action on the behavior of the frames and their connections following initial local structural damage to be assessed. The performance of pre-Northridge moment-resisting frames was assessed with both mean-centered deterministic and probabilistic assessment procedures; the significance of uncertainties in collapse assessment was examined by comparing the results from both procedures. A deterministic assessment of frames with full and partial-strength bolted T-stub connections was conducted considering three typical beam spans in both directions. The vulnerability of an older steel building with riveted connections was also analyzed deterministically. The contributions from unreinforced masonry infill panels and reinforced concrete slabs on the behavior of the building were investigated.
To meet the need for a relatively simple procedure for preliminary vulnerability assessment, an energy-based nonlinear static pushdown analysis procedure was developed. This procedure provides an alternative method of static analysis of disproportionate collapse vulnerability that can be used as an assessment tool for regular building frames subjected to local damage. Through modal analysis, dominant vibration modes of a damaged frame were first identified. The structure was divided into two parts, each of which had different vibration characteristics and was modeled by a single degree-of-freedom (SDOF) system separately. The predictions were found to be sufficiently close to the results of a nonlinear dynamic time history analysis (NTHA) that the method would be useful for collapse-resistant design of buildings with regular steel framing systems.
| Identifer | oai:union.ndltd.org:GATECH/oai:smartech.gatech.edu:1853/41079 |
| Date | 13 May 2011 |
| Creators | Xu, Guoqing |
| Publisher | Georgia Institute of Technology |
| Source Sets | Georgia Tech Electronic Thesis and Dissertation Archive |
| Detected Language | English |
| Type | Dissertation |
Page generated in 0.0022 seconds