Development of Artificial Neural Network Model for Predicting Dynamic Viscosity and Specific Heat of MWCNT Nanoparticle-Enhanced Ionic Liquids with Different [HMIM]-Cation Base Agents

Abstract

Abstract The specific heat and dynamic viscosity of various 1-hexyl-3-methylimidazolium [HMIM]-cation with multiwalled carbon nanotube (MWCNT) nanoparticles are measured and used to develop an artificial neural network (ANN) model. The specific heat values of [C12MIM][Tf2N], [HMIM][Tf2N], [HMIM][TfO], and [HMIM][Pf6] ionic-liquid-based MWCNT nanofluids decrease with increasing nanoparticle concentration and increase with temperature. Also, the dynamic viscosity of the MWCNT nanoparticle-enhanced ionic liquids decreases at low concentrations; however, it increases significantly when the concentration increases up to 1 wt%. A new ANN model for predicting the dynamic viscosity and specific heat is developed, and the predictive values agree with the experimental data with high accuracy. The mean square error and R-value of the proposed predictive ANN model are 0.001291 and 0.9985, respectively. The maximum margin of deviation of the proposed ANN model for dynamic viscosity and specific heat is 9.63% and 4.3%.