CFD modeling and validation of heat transfer inside an autoclave based on a mesh independency study

Abstract

Autoclave processing is the main technology used in the manufacturing of structural aerospace composite parts. To optimize the autoclave process, the thermal behavior of the part and mold can be investigated through simulations. Computational fluid dynamics (CFD) provide a significant contribution to studies on heat transfer and airflow patterns, which are key points in an optimization applied to achieve a homogeneous temperature distribution inside composite parts. The solution is produced by solving the 3 D unsteady Navier–Stokes equations. This paper describes a systematic numerical study using the CFD approach to significantly improve the modeling efficiency of the heat transfer coefficient (HTC) inside an autoclave and maintain a high level of accuracy. Considering the modeling cost, calculation time, and accuracy of the results, a reasonable hybrid mesh is used based on a mesh independency study. The level of grid independence is examined using the general Richardson extrapolation method. In addition, a more robust autoclave model is presented, which is unaffected by the inlet turbulence. Further, the inlet fluid velocity and turbulence models have been identified as sensitive influencing factors. In this study, the Spalart–Allmaras turbulence model shows the best performance compared with the standard $k-\epsilon$ and $k-\omega$ SST models. Finally, the results are validated with the experimental data. The mean error of the simulated temperatures in the calorimeter for the front, middle and rear positions are ${2.1}^{°}$ C, ${0.8}^{°}$ C, and ${0.9}^{°}$ C, indicating a good agreement with the experiments. This paper provides guidelines on the use of a CFD simulation to predict the heat transfer during the autoclave curing process with high accuracy and reduced numerical effort.