Heat transfer and erosion control of nanofluid flow with tube inserts: Discrete phase model and adaptive neuro-fuzzy inference system approach

The current research explores the heat transfer and erosion phenomena in a tube employing newly designed inserts. Seven different geometries for the inserts have been developed into three dimensions. The fluid used is water with iron oxide nanoparticles and the flow regime is turbulent. The modeling technique involves a finite volume method with the Eulerian-Lagrangian framework incorporating a particle-scale erosion model. The outcomes indicate that raising the velocity and volume fraction of nanoadditives enhance both the heat transfer and erosion. In comparison to the scenario without fins, the heat transfer can be increased by approximately 8.07, 6.22, 18.03, 53.72, 45.21, 14.74 and 13.08 for the considered cases G1, G2, G3, G4, G5, G6 and G7, respectively. The specific geometric design of the inserts significantly influences the heat transfer and erosion in the tube, with geometry G4 showing the most substantial improvements in the heat transfer (approximately 53 %) and erosion (around 100 %) compared to a smooth tube. Factors such as velocity, particle size, fluid viscosity and insert shape affect the erosion in the tube with water demonstrating higher erosion rates than oil. The introduction of iron oxide nanoparticles leads to 2-3 times increment in the tube wall erosion. The erosion estimation on the tube wall has been performed using an adaptive neuro-fuzzy inference system artificial intelligence model with R2 = 0.981.