Sensitivity Analysis for Knowledge Discovery in Scramjet Intake Design Optimization Using Deep-Learning Flowfield Prediction

Chihiro Fujio,Hideaki Ogawa

DOI: https://doi.org/10.1016/j.ast.2024.109183

IF: 5.457

2024-05-08

Aerospace Science and Technology

Abstract:Scramjet engines are a promising propulsion technology for future space transportation systems, and its design exploration represents a significant task toward development of high-performance scramjets. While sensitivity analysis is expected to facilitate acquisition of physical insights for effective scramjet design, the substantial computational cost incurred due to numerical simulations required for statistical analysis poses a challenge to the application, limiting the capability of those analysis. The present study has conducted sensitivity analysis using deep-learning-based flowfield prediction that can significantly reduce the computational cost for numerical simulations providing flowfield data. Global sensitivity analyses have been employed to investigate influential design variables on the flowfield and performance. These have allowed for identifying the design variables that dominantly influence or determine the performance parameters. Local sensitivity analyses have been performed to elucidate the design rationales and the characteristic flow structures for high-performance intake designs. The sensitivity analysis methods in conjunction with flowfield prediction have enabled generation of rich insights that would otherwise be difficult to acquire without this approach, demonstrating the capabilities of the proposed methodology.

engineering, aerospace