Designing the best ANN topology for predicting the dynamic viscosity and rheological behavior of MWCNT-CuO (30:70)/ SAE 50 nano-lubricant

Heat transfer optimization has a special priority in all industries. Nanoparticles dramatically develop the thermophysical properties of the base fluid when they are suspended uniformly in the base fluid. In this paper, the dynamic viscosity (mu(nf)) of MWCNT-CuO (30:70)/ SAE 50 nano-lubricant is investigated. To model this parameter versus three main effective parameters, the volume fraction of nanoparticles (phi), temperature, and shear rate, an ANN is constructed and trained using the experimental data. The results show that by increasing phi from phi = 0.0625-1%, the mu(nf) is increased from 180 to 220 cP. The temperature and shear rate have massive effects on the mu(nf). Increasing the temperature from 25 degrees to 50 degrees C, lowers the mu(nf) from 480 to 120 cP. Moreover, the shear rate has a similar trend to temperature; an increase in shear rate from 50 to 600 rpm decreases the mu(nf) from around 500-125 cP. The best ANN has attained a supreme performance with MSE= 0.003756 and R-2 = 0.999 which are entirely appropriate for a fitting ANN.

Automated summary

Heat transfer optimization is of special importance across industries. The study investigates the dynamic viscosity of MWCNT-CuO (30:70)/ SAE 50 nano-lubricant and its relationship with the volume fraction of nanoparticles, temperature, and shear rate. An artificial neural network is constructed and trained using experimental data to model this parameter. The results show significant effects of the three parameters on the dynamic viscosity of the lubricant.