Prediction and experimental validation of temperature-time coupled compression recovery performance in high-temperature seals

Thermal protection is a critical issue to be addressed for high-speed aircraft. High-temperature seals based on composite materials can effectively seal gaps between different components of high-speed aircraft and are widely used in high-speed flight applications. These seals exhibit favorable compression-recovery characteristics under high-temperature conditions, allowing them to compress and rebound in response to external load variations, thereby filling structural gaps and providing effective thermal insulation. However, long-term exposure to high-temperature environments leads to gradual degradation of seal performance, which may result in thermal sealing failure of high-speed aircraft and cause serious accidents. Therefore, accurate prediction of the compression-recovery behavior of high-temperature seals is essential to ensure reliable thermal sealing in high-speed aircraft. Current research on high-temperature seals primarily focuses on experimental testing and statistical analysis of compression-recovery properties, without establishing the relationship between compression-recovery characteristics and compression time or operating temperature. Furthermore, existing studies fail to describe the hysteresis behavior caused by plastic deformation during cyclic compression-recovery processes. As a result, the prediction accuracy remains insufficient to meet practical application requirements. In this study, a combined approach of finite element simulation and experimental investigation is employed to conduct a detailed analysis of the compression-recovery process of high-temperature seals. The nonlinear and hysteresis characteristics during the recovery process are identified and validated, and the variation patterns of compression-recovery properties with temperature and compression time are summarized. Based on these findings, a prediction method for the compression-recovery behavior of high-temperature seals is proposed using fuzzy evidence theory. By constructing fuzzy sets, high-precision interpolation and evidence theory-based fusion of compression-recovery data are achieved, enabling accurate prediction of the seal behavior. Comparison with experimental results shows a prediction error of only 5.2 %. Moreover, the proposed method improves prediction accuracy by 46.9 % and 73.7 % compared to traditional neural network and reference database algorithms, respectively, significantly outperforming existing methods. The research findings provide critical support for the reliable application of high-temperature seals in the aerospace field.