Global Engineering and Materials, Inc. — Department of Defense STTR Phase I: N23A-T001

Description

Global Engineering and Materials, Inc. (GEM), along with its team member Professor Ming-Chen Hsu’s research group at Iowa State University (ISU), proposes the development of an innovative and effective octree-based adaptive meshing tool for efficient multi-physics simulations for aircraft design and analysis. The proposed tool is capable of autonomously generating a common mesh from Computer-Aided Design (CAD) geometry with adaptive global and local refinement capabilities for coupled aero-thermal-structural analysis and optimization, which can also enable Virtual Reality (VR)-based real-time interactive designs. The novelty of the proposed modeling approach includes an efficient octree-based adaptive meshing technique for fluid and solid domains, high-quality mesh generation around complex CAD models with no human intervention, efficient local/global meshing to capture relevant physics in critical regions, and immersogeometric analysis (IMGA) based efficient numerical method to perform compressible flow analysis on non-body-fitted mesh. Moreover, the hybrid IMGA/BF (immersogeometric/boundary-fitted) approach is proposed to handle both body-fitted and non-body-fitted meshes in one framework. The development aims to address the challenges posed by increasing geometrical accuracy requirements and physical complexities in mesh generation for multi-physics simulations involving high-speed aerodynamics, structural dynamics, and thermodynamics. The lack of autonomous and geometry-aware mesh generation techniques and the difficulty in automatic mesh adaption with local and global refinements without prior knowledge of the problem have been identified as significant bottlenecks in computational fluid dynamics (CFD) workflow. The proposed tool will overcome these challenges by enabling autonomously-generated geometry-aware mesh generation and adaption for various air vehicle geometries, allowing automatic mesh adaption with no or minimal dependence on prior knowledge in multi-physics simulations for capturing critical quantities at critical locations.