Viscosity modeling of nano-modified SAE50 engine oil using RSM and ANN methods

The rheological behavior of MWCNT (10%)-SiO2 (90%)/SAE50 nanofluid was examined. The RSM andANN models were used to predict nanofluid viscosity. The optimal values were investigated for the highest and lowest solid volume fractions (SVFs). The effects of SVF, shear rate (SR) and temperature parameters were also examined on viscosity. Experimental analysis was done in SVF range from 0.0625-1%, the temperature range from 25-50 degrees C, and the SR range from 666.5-7998 1/s. The results showed that ANN with two hidden layers had the lowest error and highest efficiency for viscosity prediction. By statistical regression analysis, the comparison of predicted ANN values with relevant experimental data has demonstrated the ability to predict the viscosity in the developed neural networks properly. By comparing the performance of ANN and RSM models obtained from the experimental data, it was found that the neural network could have a more accurate prediction than the correlation developed through RSM. In order to achieve the best performance, the optimized conditions for the maximum and minimum SVFs were reported based on the demands.

Automated summary

Through experimental analysis and comparison with mathematical models, it was found that the ANN model provides higher accuracy and efficiency in predicting viscosity for the MWCNT-SiO2/SAE50 nanofluid system. The study also demonstrated the impact of factors such as SVF, SR, and temperature on the viscosity of nanofluids.