A PGSELU-DENN and CMDMO based heat transfer enhancement in Mylar plate heat exchanger

Abstract

In many industrial applications, Plate Heat Exchangers (PHEs) have been widely applied. For energy conversion and the improvement of the system economy via savings in capital investment, improving the thermal-hydraulic performance of PHEs is of crucial significance. Therefore, a number of researches were conducted on the PHEs to enhance Heat Transfer (HT). But those researches didn’t provide a satisfactory design for all the types of plates. Among those, the Mylar plate is considered, which is highly demanded than other plates. Considering the Mylar plates, the article proposes a thermal and hydraulic optimization of a Mylar Plate Heat Exchanger (MPHE). The objective of the proposed work is to improve the HT between two Mylar plates to preserve, increase, or decrease temperature in a wide variety of circumstances. Initially, by utilizing the parameters attained from the Computational Fluid Dynamic (CFD), the meta-model is created. Next, by employing the Newton Raphson with Gaussian Scaled Exponential Linear Unit-based Deep Elman Neural Network (NRGSELU-DENN) algorithm, the initial trial has been performed to mismatch prediction with the generated model. With the help of the Cauchy Mutated Dwarf Mongoose Optimization (CMDMO) methodology, the optimal design has been developed when there is a mismatch betwixt the pressure levels. Pressure Drop (PD), total cost, uniform heating, HT coefficient, and HT area are the multi-objective functions considered in this research work. Finally, to assess the effect of the parameters of MPHE on thermal and hydraulic performance, a simulation analysis was conducted. During the assessment, the mismatch is predicted between the pressure levels by the proposed model with 94.77% accuracy and 90.45% precision. The outcomes displayed that, when analogized with the common Heat Exchangers (HEs), the MPHE’s thermal efficiency is increased.