Machine learning analysis of thermo-bioconvection in a micropolar hybrid nanofluid-filled square cavity with oxytactic microorganisms
Bioconvection within nanofluids originates as microbes associated with nanoparticles (NPs), triggering convective motion. This mechanism holds significance in the development of nanotechnology, biotechnological fields, and ecological engineering. Therefore, this work utilizes numerical simulations t...
שמור ב:
| Main Authors: | , , , , , |
|---|---|
| פורמט: | Article |
| שפה: | אנגלית |
| יצא לאור: |
De Gruyter
2025-07-01
|
| סדרה: | Nanotechnology Reviews |
| נושאים: | |
| גישה מקוונת: | https://doi.org/10.1515/ntrev-2025-0177 |
| תגים: |
הוספת תג
אין תגיות, היה/י הראשונ/ה לתייג את הרשומה!
|
| סיכום: | Bioconvection within nanofluids originates as microbes associated with nanoparticles (NPs), triggering convective motion. This mechanism holds significance in the development of nanotechnology, biotechnological fields, and ecological engineering. Therefore, this work utilizes numerical simulations to investigate the thermo-bioconvection of a magnetohydrodynamic micropolar hybrid nanofluid (HNF), enriched with motile oxytactic microorganisms, within a lid-driven square enclosure, driving innovation in energy-efficient fluid technologies. The enclosure’s horizontal walls are adiabatic, while its left and right walls are heated (T
h) and cooled (T
c), respectively. Moreover, the upper wall moves steadily in the x-direction with velocity u
0. Also, a uniform magnetic field (B
0) applied along the flow direction. The TiO2-GO/H2O HNF flow is modeled as a steady, laminar, 2D, incompressible, electrically conducting, and homogeneous viscous fluid neglecting the joule heating and viscous dissipation effects. Numerical simulations utilize finite-difference approach to discretize the ensuing equations and boundary conditions, solving the resulting algebraic system iteratively via successive over-relaxation, under-relaxation, and Gauss–Seidel techniques utilizing in-house MATLAB codes. In addition, a machine learning approach was employed to accurately predict fluid transport properties using a multilayer artificial neural network structured with a feed-forward backpropagation model and optimized using the Levenberg–Marquardt algorithm. The results reveal that increasing the Reynolds number amplifies inertial forces, accelerating flow velocity and steepening temperature gradients, ultimately elevating Nuavg by 67.64%, while Shavg and Nnavg increase by 3.34 and 8.42%, respectively. Moreover, advancing Hartmann number strengthens Lorentz forces and compressing flow and reduces Nuavg by 3.25%. Furthermore, as Richardson number increases from 0.1 to 10, Nuavg increases by 14.03%, while Shavg and Nnavg decrease by 2.79 and 3.50%, respectively. Thus, the findings are useful for researchers working in nano-bioconvective systems, bio-microsystems, microbial fuel cells, and bioconvection applications. |
|---|---|
| ISSN: | 2191-9097 |