Engineering Physics Colloquium
Tuesday, February 19
4:00 PM to 5:00 PM
106 Engineering Research Building
Speaker: Professor Matt Allen, University of Wisconsin - Madison
"Validation and Methods of Dynamic Substructuring"
Abstract: "Dynamic substructuring provides an analyst the ability to perform tests on small, simple structures and combine them into a single built-up structure. This methodology often decreases the cost of analysis because expensive tests and simulations on large models can be avoided in favor of smaller, component-level tests. Dynamic substructuring has been successfully utilized in the analytical domain for several decades; however, experimental substructuring has lagged behind. Experimental substructuring often gives poor results due to sensitivity of the applied constraints to experimental errors, causing some analysts to be reluctant to use it. A new method for substructuring, dubbed the Transmission Simulator Method, has shown promise to reduce this sensitivity by softening the constraints. However, even with this new methodology, substructuring can give poor results. It would be helpful to be able to perform a validation experiment to give confidence in the results predicted by substructuring. This validation experiment is ideally inexpensive to perform and analyze as to not add significant cost to performing a substructuring analysis. This presentation gives an overview of substructuring methodologies and explores a new validation approach that can be used to assess the experimental errors which can cause substructuring to go awry. A Monte-Carlo simulation is performed in which random errors are applied to the system before substructuring is performed. The approach is then validated by simulating experimental-analytical substructuring of an engine-generator system."