Simultaneous efficacy of thermal and biological factors on thermo-bio-convection of nano-encapsulated phase change materials and oxytactic microorganisms: Analysis utilizing machine learning technique

Thermo-bioconvection is a state-of-the-art type of bioconvection that merges biological and thermal factors which would have many applications in biological waste processing, biomedical engineering, thermal energy storage, and bioreactor design. This article proposes a novel type of NEPCMs containing both nanoparticles and motile (oxytactic) microorganisms to study thermo-bioconvection flow in an inclined C-shaped enclosure with a corrugated wavy heater. Governing differential equations are solved via the finite element method. An ANN-based MLP is developed using 101 datasets, with 15 % for validation, 70 % for training, and 15 % for testing. The model, trained with the Levenberg–Marquardt algorithm, uses input parameters to predict the average Nusselt (Nu_ave) and Sherwood (Sh_ave) numbers. Results show that heat and mass transport rates are enhanced by increasing Rayleigh (Ra), bio-convection Rayleigh (Ra_b), and Peclet (Pe). With increasing Le, the Nu_ave decreases slightly, while the Sh_ave increases significantly. A more homogeneous distribution of microorganisms occurs at higher Pe and Le, and lower Ra_b. The optimal heat and mass transfer rates were obtained at inclination angle δ = 90°. Findings also show that the implemented neural network can accurately predict the average Sherwood and Nusselt numbers.