Date of Award

8-2010

Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)

Department

Civil Engineering

Advisor

William Davids

Second Committee Member

Eric Landis

Third Committee Member

Roberto Lopez-Anido

Abstract

Recent research efforts have led to the development of light-weight, modular, FRP-coated wall panels (T-Panels) for force protection. The research presented in this paper was aimed at predicting the behavior of coated T-Panels for adaptation to various threat levels and to provide the ground work for analyzing other lightweight, ductile panel structures. Three research objectives were established to accomplish this goal: 1) quantify the hysteretic response of the coated T-Panels, 2) develop a model to predict the behavior of the coated T-Panel when subjected to a single blast load, and 3) generate PI diagrams to quickly assess expected damage in the coated T-Panel given a specific blast event. Quantifying the hysteretic response of the coated T-Panel was done by developing a hysteretic model using laboratory test data and field blast test data. 4-point bending tests were conducted on the coated T-Panels which forced the panel in positive and negative bending. A piecewise hysteretic model was adapted to fit the hysteretic response of the coated T-Panels using data from the 4-point bend tests. Thirteen parameters were used to describe the model and were determined by fitting the laboratory test data as well as field blast test data. Fitting of the parameters was done manually as well as using an unconstrained least squares optimization. For the dynamic blast analysis of the coated T-Panel, Newmark?s time stepping method was used to solve the equation(s) of motion. Both linear SDOF and MDOF systems were used as well as a nonlinear SDOF system which incorporated the hysteretic model derived from laboratory and field blast tests. The behavior of the coated T-Panels was best described using a nonlinear analysis with hysteretic parameters fitted from blast testing data. This model was able to capture the softening behavior of the panel as well as an increase in period which was likely due to damage that occurred in the panel during the blast events. PI diagrams were generated using the linear and nonlinear SDOF models of the coated T-Panel. Curves were generated using only the positive phase of the blast wave as well as the positive and negative phase of the blast wave. It was found that there is a significant difference in PI curves using a linear analysis when the initial and rebound displacement of the panel is considered along with using the negative phase of the blast wave. The nonlinear analysis led to inaccurate results when solving for points on the PI curve. This was due to the combination of the nonlinear properties, the negative phase of the blast wave, and the objective function used to solve for the PI points. Further research is needed to determine the exact origin of the inaccuracies in generating a PI diagram using the nonlinear SDOF analysis for the coated T-Panel. More testing would also help to determine if the negative phase of the blast wave is needed to generate accurate PI diagrams.

Download

Share