%0 Journal Article
%A Bhoopal, Rajpal S.
%A Sharma, P. K.
%A Singh, Ramvir
%A Beniwal, R. S.
%D 2013
%I Begell House
%K effective thermal conductivity, artificial neural network, metallic foams, volume fraction, feedforward backpropagation
%N 7
%P 585-596
%R 10.1615/JPorMedia.v16.i7.10
%T APPLICABILITY OF ARTIFICIAL NEURAL NETWORKS TO PREDICT EFFECTIVE THERMAL CONDUCTIVITY OF HIGHLY POROUS METAL FOAMS
%U http://dl.begellhouse.com/journals/49dcde6d4c0809db,513068641bda6b07,51d8587d03f25578.html
%V 16
%X This paper presents the applicability of artificial neural networks to predict effective thermal conductivity of highly porous metal foams. Artificial neural network models are based on feedforward backpropagation network with training functions such as gradient descent (GD), gradient descent with adaptive learning rate (GDA), gradient descent with momentum (GDM), gradient descent with momentum and adaptive learning rate (GDX), and scaled conjugate gradient (SCG). Volume fraction of fluid phase and thermal conductivity of solid and fluid phases are input parameters for the artificial neural network to predict the effective thermal conductivity. The training algorithm for neurons and hidden layers for different feedforward backpropagation networks runs at the uniform threshold function TANSIG-PURELIN for 500 epochs. Better agreement of predicted effective thermal conductivity values is obtained by using artificial neural networks with the experimental results. A comparison with other models is also made and it is found that the values of effective thermal conductivity predicted by using the present model are in good agreement with the reported experimental values.
%8 2013-06-28