Data-driven adaptive micro element reconstruction method for efficient ray tracing in time-continuous aero-optical flow fields

Numerical simulation is widely employed to analyze optical transmission performance in flow fields. However, time-continuous flow fields produce large data volumes, and dynamic aero-optical analysis requires substantial computational resources, resulting in low efficiency with the conventional method. Consequently, this study proposes the data-driven adaptive micro element reconstruction (DAMER) method. DAMER significantly improves computational efficiency in aero-optical simulations of time-continuous flow fields through eliminating redundant data, reducing computational load and interpolation requirements, and enabling continuous, precise adaptive step-size adjustment. In addition, unlike conventional approaches that empirically determine step-size ranges, DAMER introduces a method for defining step-size ranges across diverse computational fluid dynamics (CFD) grid scales. Simultaneously, DAMER simplifies the identification of intersections between rays and irregular optical windows in 3D flow fields. In the cases evaluated, DAMER achieved speedups of 131.10 and 11.56 compared to the conventional method for 2D and 3D continuous flow fields, respectively. These advancements improve aero-optical analysis by significantly enhancing computational efficiency and accuracy.