A machine learning approach and numerical investigation for intelligent forecasting of entropy generation rate inside a turbulator-inserted solar collector tube

The irreversibility rate analysis on the solar thermal collectors can be used to determine the regions in the geometry with high intensification of the thermal and viscous entropy generation rates ( Sth and Sfr), thereby improving the geometry. Although the numerical analysis of the hydrothermal characteristics of the solar thermal collector has been widely studied, the entropy generation rate inside a turbulator insert absorber tube is less understood. In addition, there is still the lack of an accurate predictive model for the estimation of the total entropy generation rate inside the turbulator insert absorber tube. This study deals with the determination of Sth and Sfr of water flow inside a tube considering three twisted tape inserts with pitch distances (H) of 50 mm, 100 mm, and 150 mm as well as the Re numbers of 500, 1000, and 1500. The results showed that the escalation of H from 50 mm to 150 mm at Re numbers of 500, 1000, and 1500, Sfr reduces by 15.40%, 24.50%, and 29.67%, respectively. Also, a 90% increase in Sfr was observed as Re intensified from 500 to 1500. In addition, the increment in Re from 500 to 1500 reduces by 43%, 41%, and 34% at H values of 50 mm, 100 mm, and 150 mm, respectively. Besides, the increase in H from 50 mm to 150 mm escalates Sth b by 55%, 63%, and 60%, respectively, for Res of 500, 1000, and 1500. Meanwhile, the entropy generation contour plots revealed that the rate of irreversibilities near the blade and tube wall is high indicating a turbulator geometry modification is needed for the lower entropy generation rates. Moreover, the results of the artificial neural network model showed the high accuracy of the introduced correlation for the estimation of the total entropy generation rate in terms of Re and H.