Predictive modeling through physics‐informed neural networks for analyzing the thermal distribution in the partially wetted wavy fin

• Published in: Zeitschrift für angewandte Mathematik und Mechanik Vol. 104; no. 8
• Main Authors: Karthik, Kalachar, Sowmya, Ganeshappa, Sharma, Naman, Kumar, Chandan, Ravikumar Shashikala, Varun Kumar, Alur Shivaprakash, Siddesh, Muhammad, Taseer, Gill, Harjot Singh
• Format: Journal Article
• Language: English
• Published: Weinheim Wiley Subscription Services, Inc 01.08.2024
• Subjects: Conduction heating; Dimensional analysis; Dimensionless analysis; Heat conductivity; Heat transfer; Neural networks; Parameters; Prediction models; Runge-Kutta method; Temperature dependence; Temperature ratio; Thermal conductivity; Wavy fins
• ISSN: 0044-2267; 1521-4001
• DOI: 10.1002/zamm.202400180

The heat transport analysis and thermal distribution in partially wetted wavy profiled fin are investigated in the current study. Convective, radiative effects and temperature‐dependent thermal conductivity are all considered in this heat transfer analysis. The dimensional governing temperature equations of the partially wetted wavy extended surface are nondimensionalized utilizing the appropriate dimensionless terms. Further, the resulting nondimensional thermal equations of the wavy fin are solved by employing Physics‐Informed Neural Network (PINN). The values of the temperature equations obtained by the numerical procedure Runge Kutta Fehlberg's fourth‐fifth (RKF‐45) order scheme are compared with PINN outcomes. The results are portrayed with the aid of tables, and the significance of several dimensionless constraints on the partially wet wavy fin is exhibited using graphical illustrations. A rise in the thermal conductivity parameter values enhances the wavy fin's thermal profile. The temperature of the wavy fin diminishes as the convective‐conductive parameter, temperature ratio parameter, and radiation‐conduction parameter upsurges.