Mathematical modeling of thermal contact resistance for different curvature contacting geometries using a robust approach

Nowadays having deep knowledge on thermal contact conductance (TCC) and thermal conduct resistance (TCR) existed between various type metals is interesting during heat transfer occurrence in the nuclear reactor, thermal control system of spacecraft, and heat exchangers. In this present contribution, artificial neural network (ANN) coupled with multi-layer perceptron (MLP) modeling was utilized for the prediction of transient temperature contour in various contacting surface such as flat-flat, flat-cylinder, cylinder-cylinder. In order to develop accurate transient model, position, time, and roughness parameter of metal was used as input parameter and temperature of solid bodies decided as target parameter of model. Modeling results indicates that ANN base modeling show great accuracy in comparison with other numerical methods. Also, values of average absolute relative deviation (AARD), coefficient of determination (R2) for the overall data is 0.056 and 0.996 respectively which clarifies the accuracy and robustness of the proposed model.