Artificial intelligence approach in mixed convection heat transfer under transverse mechanical vibrations in a rectangular cavity

In the current research, the mixed convection heat transfer in a rectangular chamber with walls with sinusoidal oscillations and mechanical vibrations has been investigated. Mechanical vibrations on the chamber, sinusoidal oscillations of the hot wall and flow due to buoyancy are considered. The finite volume method is utilized for simulation. The efficacy of changes in governing parameters, such as frequency, oscillation amplitude, $\mathrm{Ra}$, chamber length to width ratio, change of fluid type on $\mathrm{Nu}$ has been investigated. The results indicate that there is an optimal ratio of chamber dimensions that has the maximum $\mathrm{Nu}$ in the fixed $\mathrm{Ra}$, and this ratio depends on the type of fluid. In the presence of sinusoidal oscillations of the hot wall and transverse mechanical vibrations of the cylinder, it increases the heat transfer by about 96 % and 75 %, respectively, compared to the state without vibration. The increase in the frequency and amplitude of oscillations in the case where the hot wall oscillates sinusoidally is negligible on the Nusselt number. Increasing the frequency and amplitude of oscillations of transverse vibrations of the chamber has a significant efficacy on $\mathrm{Nu}$, and the amplitude of oscillations has a greater efficacy than the frequency of oscillations on heat transfer. Based on the available data and using artificial intelligence, GMDH, $\mathrm{Nu}$ has been estimated. The results indicate that this modeling has been able to estimate the Nusselt number with good accuracy with $R^2=0.948$.