Abstract:
A Compact Heat Exchanger (CHE) is a device used to transfer energy from one fluid to another, vital role in the efficient energy transfer. The optimum design of these Compact heat exchangers is a challenging skill for making minimum pumping power in terms of minimum pressure drop and efficient heat transfer. The generation of thermo-hydraulic performance factors, Colburn factor ‘j’ and fanning friction factor ‘f’ correlations for CHE take a large number of simulations in numerical analysis and the same is very expensive in modelling the model for Experimental analysis. In the open literature, experimentally developed fins performance as a function of Reynolds number and geometric parameters, which is expensive. The Numerical model has been developed by simulating Reynolds number Re and geometric parameters, such as fin height h, fin spacing s and fin thickness t for generation of fanning friction factor f and Colburn factor j. The 144 fin geometric parameters are used in the numerical model to develop a correlation. The numerical model is analyzed using ANSYS Fluent®. This tremendous process of correlation development is faster by using Artificial Neural Networks (ANN). This Paper focused on developing design data requirements for rectangular plain fin compact heat exchangers using CFD and Neural Networks(NN). The Analogy of correlation developed by Neural Network Prediction and CFD Fluent are verified and validated using open literature. The correlations from Neural Networks using limited data are faster and has better agreement with open literature.