Hardware Accelerated Code for Hybrid Computational Electromagnetics
Small Business Information
315 S. Allen St., Suite 416, State College, PA, -
Director, Propagation Sof
Director, Propagation Sof
AbstractABSTRACT: A hybrid computational Electromagnetic solution will be developed that combines multiple physics techniques in order to handle far-field propagation, near-field interactions, and detailed interactions with complex objects, such as human anatomical models, while taking advantage of the computational power of graphics processing units (GPU) and other optimizations. The approach will leverage existing, mature physics models with demonstrated capabilities, existing GPU acceleration, and mature graphical user interfaces. Physics interfaces prototyped and demonstrated in Phase I will be fully implemented and integrated into the code baselines. Acceleration and optimization techniques tuned to the problem will be implemented, benchmarked, and verified. A preliminary hybrid graphical user interface will be developed to allow setup, execution and visualization of hybrid computational EM simulation results. The final solution will be a full end-to-end modeling tool that provides high-fidelity and optimal run times, with seamless interfaces between the physics techniques, controlled through an integrated, user-friendly graphical user interface. BENEFIT: The outcome of this SBIR will be a modeling suite that seamlessly integrates high-fidelity electromagnetic simulation in the near-field of antennas and in the vicinity of human anatomical models, with high-fidelity propagation calculations over rough terrain or within urban settings. Antenna designers, engineers, and health physicists could use this tool to assess health and safety risks in a variety of environments by determining the potential for radiation exposure to personnel. Its hybrid capabilities will allow it to be used to perform detailed assessments of fields or specific absorption rate (SAR) for near-field analysis or far-field analysis, well beyond the range where high-fidelity calculations would be feasible with current modeling solutions. GPU and other acceleration techniques will ensure reasonably optimal run times for calculations that would otherwise take significant time to complete. The combined set of capabilities also has potential for use in other fields, providing the capability to perform high fidelity electromagnetic analysis near any type of complex object.
* information listed above is at the time of submission.