Speaker
Description
Thermal analysis for space systems encounters a significant bottleneck in computing radiative heating contributions to the heat transfer problem, particularly when transparent and specular materials come into play. In this study, we introduce an innovative implementation that harnesses the remarkable parallel processing capabilities of ray tracing on Graphical Display Units (GPUs). Specifically designed for NVIDIA GPUs, our implementation efficiently handles parallel view factor and/or direct radiative heating calculations through Monte Carlo ray tracing.
In a previous study, we demonstrated that our GPU-based algorithm exhibited extraordinary speed gains compared to the CPU implementation in Simcenter 3D Thermal-Multiphysics, specifically for surface-to-surface enclosure radiation. In this presentation, we extend the application of our GPU-accelerated methods to encompass Solar and Orbital heating radiation calculations. By comparing the computation times and accuracy of our legacy CPU methods with our new GPU techniques, we reveal the advantages of our approach for Solar and Orbital heating analyses. For the Solar heating functionality, we will be looking at a detailed lunar surface model based on LRO topology data that was constructed for the Canadien Lunar Rover mission. For the orbital heating mission, we will focus on a satellite model including articulating components.
