Comparative study of two simulation methods for thermal analysis of the Ceramic matrix composite plate

Ceramic matrix composites (CMCs) are regarded as ideal candidates for hot-end components in future aeroengines due to their high-temperature resistance and low density. However, the anisotropy of their thermal conductivity and the multi-phase nature pose significant challenges for numerical simulations. Currently, thermal simulation methods for CMCs can be classified into macro-scale and meso-scale approaches. In this study, both methods are employed to simulate a CMC plate, and the resulting temperature and temperature gradient distributions on internal characteristic sections are compared. The results indicate that the macro-scale method inherently lacks the resolution to capture internal local temperature and gradients, whereas the meso-scale method provides such resolution. Moreover, the internal thermal details are considerably more complex than macro-scale averages suggest. Local temperature fluctuations exhibit a peak relative difference of 17.21%, and the maximum error in gradient prediction reaches 7 × 10⁵ K/m, corresponding to a relative difference of 470%. Such substantial prediction inaccuracies directly lead to erroneous thermal stress calculations, posing a significant design risk. Consequently, the meso-scale simulation method is more suitable than the macro-scale method for simulating CMC hot-end components, and can therefore provide valuable guidance for the thermal design of such components.