In this Phase II we will transition the successful Phase I demonstration of a unified theoretical methodology for surrogate model development at the step and process levels, and of a heterogeneous software architecturethe FAD Tool Frameworkthat together are able to automate the development and production of fast running models (FRMs) into a fully functional standalone FAD Tool.FRMs are surrogate models, often based on high-fidelity physics-based codes, which are an increasingly important Modeling and Simulation (M&S) element.Areas of interest to AFRL where FRMs address critical needs include immediate turn-around weaponering predictions for analyzing mission scenarios in MOUT and buried contexts, parametric what-if studies to assess consequences for incompletely specified problems, exploratory investigations for unspecified materials or other unknown (i.e., noise) parameters, and analysis surveys intended to identify critical parameters, quantify problem sensitivities and uncertainties, and perform design optimizations.
