Horse Herd Optimization and LSTM Configuration for Minimizing Pressure Drop and Predicting Thermal Performance in Shell and U-Tube Heat Exchanger

Abstract

The industrial component that transfers heat from one fluid to another most frequently uses Shell and Tube Heat Exchangers (STHE). Enhancing the heat transfer efficiency of heat exchangers has garnered more attention as a result of scarce energy resources and high energy expenditures. In STHE, the pressure drop is considered an important issue that causes cracks and economic losses. An essential factor in improving the performance of a heat exchanger with low pressure drop was the angle and distance of the baffles. Several methods were developed to reduce pressure drop and speed up heat transfer. But those methods were not provide a satisfactory pressure drop reduction, so the optimal baffle configuration was still a task in the heat exchanger. In the proposed model, Horse-herd Optimization Algorithm (HOA) based baffle design and neural network based thermal performance prediction arrangement was developed to reduce the pressure drop and predict the rate of transferring heat. Shells and tubes were developed at the corresponding material, inside the shell, a baffle was designed to barrier the flow of cold water. The optimal solution of baffle configuration was solved through HOA, which finds the appropriate baffle’s distance and angle by reducing the pressure drop. After the water flow modelling, the seven key parameters values were observed, and create a dataset. Using this data, a thermal performance prediction system was developed to analyze each period input value to predict the net energy, heat transfer rate, and Nussle number. The proposed model provides 52 Pa pressure drop, 0.59 effectiveness, 0.59 NTU, 417 U, and 92% accuracy. The output of the suggested approach is contrasted with that of other current methods for validation. The proposed model offers a high heat transferring capacity and reduces pressure effects risk.