A smart prediction tool for estimating heat transfer and overall pressure drop from shell-and-tube heat exchanger
Almost all thermal/chemical industries are equipped with heat exchangers in order to enhance the thermal efficiency. The performance of a shell and tube heat exchanger depends significantly on the design parameters like the tube cross-sectional area, tube orientation, baffle cut, etc. However, there are no specific relationships among these parameters to obtain an optimal design, such that the heat transfer rate is maximized and the pressure drop is minimized. Therefore, experimental and numerical simulations are carried out for a heat exchanger at various process parameters. Heat exchanger considered in this investigation is a single shell-multiple pass type device. For the performed experimental datasets, a generalized regression neural network is applied to generate a relation among the input and output process parameters.