Numerical and experimental study of inverse natural convection-conduction heat transfer in an inclined rectangular cavity

In this work, the inverse three-dimensional (3D) computational fluid dynamics (CFD) method combined with overdetermined experimental data and least squares method is applied to predict the unknown heat transfer rates and the suitable turbulent flow model in an inclined rectangular cavity. The conjugate heat transfer of two 3D natural convection systems at different temperatures separated by a cavity is solved by coupling. The effect of inclination angle theta on the appropriate flow model is investigated. The RMSE obtained by the appropriate flow model is smaller than that obtained by other flow models. The obtained Nusselt number on the hot wall also needs to be closer to the proposed correlation. An interesting finding is that the zero-equation model and RNG k-epsilon model are respectively more suitable for problems with 0 <=theta <= 30 degrees and 60 degrees <=theta <= 90 degrees, respectively. The number of natural circulation decreases with increasing theta. The obtained estimates of Nuh agree well with the proposed correlation and existing experimental results. A review of the literature revealed that these findings have not yet appeared in the open literature.