JFET_V12_N3_RP5 N.K. Chavda Journal on Future Engineering and Technology 2230 – 7184 12 3 35 43 Artificial Neural Network, Critical Heat Flux, Nanofluid, Pool Boiling, Prediction Critical heat flux is one of the significant factors during pool boiling to observe in order to reduce the risk of damaging or melting of metal. Increasing value of critical heat flux, not only increases the functionality of various thermal systems, but also ensure their safety. Out of the various methods available, one of the recent methods to increase the critical heat flux is application of various nanofluids. The enhancement in critical heat flux in pool boiling using nanofluid depends on different parameters. Thus it requires extensive experimentation to propose the appropriate nanofluid for the same. In the present paper, critical heat flux have been experimentally evaluated using various water based nanofluids, such as Al2O3, CuO, and TiO2 having 0.1% to 1.0% volume concentration when two types of test heaters with different diameters are used. On the basis of the experimental results, fifty different ANN models using various ANN architectures, such as FFN, CFB, EBP, FFDD, GR, and RB have been developed and trained considering four input parameters, such as type of nanoparticle, concentration of nanoparticle, test heater material, and test heater diameter to predict the critical heat flux. The trained ANN models have been used to simulate the critical heat flux value and errors in prediction have been calculated in terms of MSE, NMSE, MARD, MRE, and AAE. The ANN model C8 (Elman back propagation having eight neurons of hidden layer), which yields global minimum value of error in prediction is proposed as the suitable ANN model for the case. February - April 2017 Copyright © 2017 i-manager publications. All rights reserved. i-manager Publications http://www.imanagerpublications.com/Article.aspx?ArticleId=13436