Investigation of the thermal analysis of a wavy fin with radiation impact: an application of extreme learning machine

Abstract

The combined impact of radiation and convection on the heat transfer of a wavy fin is scrutinized in the present analysis. The novelty of this research work is that it proposes a deterministic machine learning model known as an extreme learning machine to address the heat transfer problem of a wavy fin. The effect of radiation on convective heat transfer and the Rosseland approximation for the radiation heat exchange have been considered in the investigation. The nonlinear ordinary differential equation (ODE) is converted to its nondimensional form using the appropriate dimensionless variables. Runge-Kutta-Fehlberg's fourth-fifth order technique (RKF 45) is used to solve the nondimensional ODE numerically. The roles of convection-conduction, radiation-conduction, thermal conductivity, and radiation parameters have been discussed for satisfying a prescribed temperature distribution in rectangular and wavy fins with graphical visualization. A rise in convection-conduction and radiation-conduction variables decreased the thermal distribution of both the wavy fin and rectangular fin. Further, ANSYS simulation analyzes the variation of temperature and total heat flux in both rectangular and wavy fins. The study demonstrates the effectiveness of the model selected through the obtained results, which indicate the potential of the regression model for providing an accurate prediction.