The Establishment and Optimization of a Multi-Thermodynamic-State Gas Radiation Model Based on Spectral Mapping Using Intervals of Comonotonicity

In infrared radiation calculations, the k-distribution method effectively improves the computational efficiency of solving the radiative transfer equation for uniform paths and achieves accuracy comparable to the line-by-line method. However, when applied to highly non-uniform scenarios involving multiple thermodynamic states, such as the infrared radiation from aero-engine nozzles, the computational error increases significantly. This paper proposes a spectral mapping method for multiple thermodynamic states, which iteratively partitions the spectral intervals of the target gas into multiple comonotonic sub-intervals using particle swarm and clustering algorithms. This approach eliminates the blurring effect of traditional k-distribution methods in strongly non-uniform scenarios and enhances the computational accuracy. The study examines the impact of sub-interval partitioning strategies on the accuracy of the gas radiation model, explores the mechanism behind constructing comonotonicity within sub-intervals, and reveals how variations in the comonotonic vector and spectral point clustering strengthen sub-interval comonotonicity. The proposed spectral mapping method and optimization techniques are applied to gas radiation models in typical infrared bands, and the performance of the model is evaluated using results from representative one-dimensional test cases. The results demonstrate that the optimized spectral mapping method reduces the overall relative error of the gas radiation model from 63% to 7.3%, achieving a maximum improvement in computational accuracy of 88.5%.