Transient CFD- and Model Order Reduction-Based Solver for Rapid Thermal Signature Prediction

Existing methods for human and vehicle thermal signature prediction are computationally demanding or inaccurate, and therefore ill-suited for multi-condition parametric analysis, trade-space exploration, and design optimization of military ground vehicles. The proposed effort aims to develop an innovative Computational Fluid Dynamics (CFD) and Model Order Reduction (MOR) based computational tool for accurate and rapid prediction of transient turbulent flow, heat transfer and thermal signature. Key features of proposed solution include: (1) fully coupled CFD, thermal, and signature analysis framework with one integrated solver; 2) mathematically rigorous MOR technique to enable accurate and reliable analysis; 3) an online CFD-MOR analysis approach that could lead to orders-of-magnitude speedup over the traditional CFD methods; and 4) a modular software framework to facilitate seamless integration of different software components or models. In Phase I, key components, including CFD-thermal coupling with turbulence, a MOR engine with novel algorithms, and prototype integration of CFD-ROM will be developed and feasibility of the overall computational tool will be demonstrated with a military vehicle. The Phase II effort will focus on improving tool capability and computational performance (e.g., accuracy, speed, and reliability), enhancing software modularization and quality, developing graphical user interfaces (GUI), software integration, and extensive validation and demonstration.