Execution of a smart prediction tool to evaluate thermal performance in a heat exchanger by using single elliptical leaf strips with altered angle

Abstract

Heat exchangers are prominent industrial applications where engineering science of heat transfer and Mass transfer occurs. It is a contrivance where transfer of energy occurs to get output in the form of energy transfer. This paper aims at finding a solution to improve the thermal performance in a heat exchanger by using passive method techniques. This experimental and numerical analysis deals with finding the temperature outlets of cold and hot fluid for different mass flow rates and also pressure drop in the tube and the annular side by adding an elliptical leaf strip in the pipe at various angles. The single elliptical leaf used in experiment has major to minor axes ratios as 2:1 and distance of 50 mm between two leaves are arranged at different angular orientations from 00 to 1800 with 100 intervals. Since it’s not possible to find the heat transfer rates and pressure drops at every orientation of elliptical leaf so a generalized regression neural network (GRNN) prediction tool is used to get outputs with given inputs to avoid experimentation. GRNN is a statistical method of determining the relationship between dependent and independent variables. The values obtained from experimentation and GRNN nearly had precise values to each other. This analysis is a small step in regard with encomiastic approach for enhancement in performance of heat exchangers.