A web-based intelligent calculator for predicting viscosity of ethylene–glycol–based nanofluids using an artificial neural network model

Abstract

This study presents the development of an artificial neural network (ANN) model to predict the viscosity of ethylene–glycol based nanofluids with different types of nanoparticles using four input parameters: nanoparticle type, size, concentration, and temperature of measurement. The model was trained and validated using 470 experimental measurements. The ANN model demonstrated high accuracy in predicting the viscosity of nanofluids. The obtained statistical error metrics between the measured and predicted values of viscosity were found to be very low. MAPE values were equal to 1.19% and 2.33% for training and testing respectively. The developed model can help researchers to better understand EG-based nanofluids viscosity behavior, and this could be considered as a good step forward to help researchers design new nanofluids with enhanced properties. To make the model more accessible for engineers and researchers, a user-friendly web application was developed using Angular and Django, allowing users to input parameters and obtain viscosity predictions without dealing with complex code. The web application offers multiple output options, including figures, tables, and Excel files. This multidisciplinary research study combines web technology, data science, and fluid mechanics to provide a valuable tool to predict nanofluids’ viscosity for different input parameters.

Graphical abstract