Affiliation:
1. South China University of Technology, 510641 Guangzhou, People’s Republic of China
Abstract
A hypersonic airborne radome must possess multiband electromagnetic transparency, high specific strength, and thermomechanical resistance to ensure proper antenna operation. However, achieving these multifunctional properties in radome wall design is challenging due to conflicting material requirements. Moreover, existing design methods for arbitrary multiband and broadband hypersonic radomes are insufficient. This paper presents an efficient method that integrates artificial intelligence and an evolutionary algorithm to enable the multidisciplinary synchronous design of the radome wall. The designed radome walls using multilayered porous silicon-nitride ceramics achieve power transmission efficiency exceeding 90% in a broadband of 30 GHz or arbitrary multiple bands with bandwidths of 5 GHz, covering the centimeter-to-millimeter wave range. Moreover, the designed laminate also meets requirements of flexural strength surpassing 60 MPa and a maximum thermal stress–strength ratio below 1. Notably, the utilization of artificial intelligence accelerates the design process by at least 1093 times compared to conventional methods.
Funder
Natural Science Foundation of Guangdong Province, China
National Natural Science Foundation of China
the State Key Laboratory of Subtropical Building and Urban Science
the National Key R&D Program of China
the Guangdong Provincial Key Laboratory of Human Digital Twin
Publisher
American Institute of Aeronautics and Astronautics (AIAA)