Two-phase mixture numerical and soft computing-based simulation of forced convection of biologically prepared water-silver nanofluid inside a double-pipe heat exchanger with converging sinusoidal wall: Hydrothermal performance and entropy generation analysis

Abstract

The cooling performance of biological water-silver nanofluid (NF) in three double-pipe heat exchangers with converging sinusoidal inner wall (SCDHX), converging inner wall (CDHX), and plain inner wall (PDHX) was examined numerically using the first and second laws of thermodynamics. The two-phase mixture model is employed to conduct the simulations. Based on the results, the sinusoidal wall increases the flow mixing, and thereby the convective heat transfer coefficient in SCDHX enhances by 50% and 18% over the PDHX type for Res of 500 and 2000, respectively. Moreover, the highest NF frictional entropy generation rate (S˙f,m,c) was obtained for SCDHX; 67% and 80% higher than those for CDHX and PDHX, respectively. The efficiency criteria ratio (η) of SCDHX was obtained as roughly 1.1 for Res of 500 and 1000, which is 27.27% higher than that for CDHX type. Also, the efficiency criteria ratio of SCDHX over the CDHX was in the range of 1.21-1.41. Besides, a robust soft computing, namely the Gaussian process regression (GPR) approach, was proposed to accurately estimate the total entropy of cold NF, the total entropy of hot NF, and the performance ratio based on the nanoparticle concentration and Reynolds number.