Development of a model for entropy generation of water-TiO2 nanofluid flow considering nanoparticle migration within a minichannel

摘要: Abstract Thermal, frictional and total entropy generation rates are numerically investigated for the water-TiO 2 nanofluid flow in a circular minichannel considering particle migration. Then, an Artificial Neural Network (ANN) model is developed prediction of terms Reynolds number, concentration size using obtained numerical data. The examined both locally globally. migration causes non-uniform distribution thermophysical properties nanofluid. This non-uniformity becomes more intense by increasing each parameters concentration, such that mean 4% 80 nm, values increment from wall to pipe center about 32% 66% numbers 1000 2000, respectively. results indicate can have significant effects on rates, especially at great sizes high concentrations. To address contribution factors generation, Bejan number calculated. reduces but increases particles enlargement. At all concentrations, greater than 0.8 which indicates 80% generated comes heat transfer. ANN predict outputs with acceptable accuracy, MAE thermal, respectively 4.265 × 10 − 6 , 1.606 × 10 4.305 × 10 based test