Simulation of Mechanical System Kinematic Operation Subsequent to High Intensity Loading

The goal of the Phase II effort is to develop finite element (FE) tools to predict the post-damage kinematic response of shipboard mechanical systems. Validation of the methods will be accomplished through a series of increasingly complex tests and validations. Starting with physics-level coupon testing through component/subassembly and eventually system-level tests, best practices will be determined to enable accurate and efficient FE modeling for applications involving contact and friction, nonlinear material response, and large articulated motions.In conjunction with the test-based validation, an efficient and scalable stochastic framework will be developed to enable economical study on the effects of a wide range of physical parameters. This analytical approach will utilize artificial neural networks (ANNs) to serve as extremely fast-running (<< 1 sec) surrogates for large (+10M DOF) FE models to enable hundreds or thousands of simulations to be performed in a matter of minutes or hours, rather than weeks or months for a comparable number of FE simulations. ATA Engineering will leverage and expand on previous successes with ANNs for FE systems and apply these machine learning algorithms to analyzing simple components and subassemblies, with later efforts focused on representing larger systems.