UC San Diego

The dissertation responds to the need for efficient techniques for testing and validating protective technologies to resist explosive loading on civil structures; the newly-developed Explosive Loading Laboratory allows investigators to simulate explosive blast loading at a lower cost and with better data than when using real explosives. Analytical tools are developed for the new laboratory and comparisons between real and simulated explosive tests confirm that the two are equivalent. These developments provide future investigators with confidence in the simulated blast methodology and tools to predict results and analyze the data, thus setting the stage for decades of fruitful research in a nascent field of simulated explosive testing. The simulated blast load is applied via specially designed apparatuses called blast generators, each of which is made of a high speed actuator that accelerates an impacting module towards the specimen. An elastomeric pad, called the programmer, on the front of the impacting module provides a blast-like pressure pulse to the specimen. The entire setup is housed in a post-tensioned, self-reacting, reinforced concrete reaction structure. An inaugural series of tests on column specimens provides data on structural behavior and on the functioning of the laboratory. New methods of data analysis are developed: the equivalent uniform load method reduces the load recorded on multiple channels to a single, rationally- derived value, and the equivalent charge method translates the experimental results to an equivalent real scenario. Because there is no fireball to obscure the specimen during the blast, the verbal descriptions and video stills herein are the first-ever visual record of structural behavior under blast loading. The laboratory results are corroborated with field test data via qualitative comparisons and demand-damage plots to demonstrate the validity of simulated blast testing. A methodology is developed for quickly and easily determining the necessary velocity of the impacting module to deliver a desired impulse to the specimen. Load-deformation programmer data are collected and a phenomenological model of programmer behavior is derived for use in detailed pre-test analyses, including finite element implementations. Both tools are derived theoretically, calibrated to experimental data, and verified against independent experimental data