Velocity prediction of Cu/water nanofluid convective flow in a circular tube: Learning CFD data by differential evolution algorithm based fuzzy inference system (DEFIS)

The ability of the artificial intelligence (AI) of the differential evolution algorithm with a fuzzy inference system (DEFIS) in the prediction of fluid flow hydrodynamic is investigated. The DEFIS learns the CFD results in the convective flow of Cu/water nanofluid in a circular tube. The ANSYS-FLUENT CFD package is used. The thermalchemical equilibrium conditions exist between the Cu nanoparticles and water. The homogenous single-phase CFD model can be considered by this assumption. The CFD simulations are done for several inlet velocities (i. e., 0.6, 0.7, 0.8, 0.9, and1). The DEFIS takes the x, and y coordinates and the inlet velocity as inputs to predict the nanofluid velocity at z-direction an output. The best intelligence of the DEFIS is tried by adjusting the input numbers, population numbers, and cross-over. The results displayed that for the best intelligence of the DEFIS (i.e., input number = 3, population number = 5, and cross-over =0.7), there is high compatibility between the results of the CFD and the DEFIS. At this condition, the DEFIS shows the ability to predict nanofluid velocity related to any inlet velocity independent of the CFD.

Automated summary

The ability of DEFIS in predicting nanofluid velocity related to any inlet velocity was investigated through the differential evolution algorithm with a fuzzy inference system. The results demonstrated high compatibility between DEFIS predictions and CFD outcomes, showing its effectiveness in fluid flow hydrodynamic prediction.