Simulation and accurate prediction of unsteady nonlinear convection flow by artificial neural network using numerical data DOI
Sabba Mehmood, Sajjad ur Rehman

Physics of Fluids, Год журнала: 2025, Номер 37(4)

Опубликована: Апрель 1, 2025

In the present article, time-dependent Casson fluid flow over an oscillating plate under nonlinear Oberbeck–Boussinesq approximation is considered. The problem aligns with Stokes's second problem, which used to describe motion due oscillatory surface. set of governing equations consists velocity, temperature, and mass transfer. These are first nondimensionalzed by using appropriate scaling. heat transfer process has been investigated in presence additional effects, such as Joule heating, viscous dissipation, generation source. well-known Buongiorno model employed for Brownian thermal migration nanoparticles. irreversibility system, source also various values parameters. finite difference method obtain numerical solution problem. obtained solutions subsequently integrated into artificial neural network (ANN) model. ANN trained famous Levenberg–Marquardt (LM) backpropagation algorithm. performance assessed a range statistical methods, mean squared error, regression analysis, curve fitting, error histograms. shows high accuracy, absolute errors ranging from 10−4 10−5. impact key parameters on fluid, heat, phenomena depicted graphically numerically. absence solutal convection, i.e., when Gc=0, it observed that convection significant However, case reverse effect absent, Gr=0. both (Gr=Gc=0) plays crucial shaping thermodynamics. velocity enhanced this case. Other physical quantities skin friction coefficient well Nusselt Sherwood number different Skin decays Grashof increased. increment random nanoparticles causes reduction numbers. Augmentation entropy profile enhancement magnetic Brinkman number. Consistency between computational results friction, number, observed, confirms accuracy our We believe beneficial biomedical engineering optimizing blood environments, microfluidic devices precise manipulation, industrial cooling systems enhancing efficiency. Additionally, they contribute geophysical dynamics improving understanding buoyancy-driven flows natural systems.

Язык: Английский

Simulation and accurate prediction of unsteady nonlinear convection flow by artificial neural network using numerical data DOI
Sabba Mehmood, Sajjad ur Rehman

Physics of Fluids, Год журнала: 2025, Номер 37(4)

Опубликована: Апрель 1, 2025

In the present article, time-dependent Casson fluid flow over an oscillating plate under nonlinear Oberbeck–Boussinesq approximation is considered. The problem aligns with Stokes's second problem, which used to describe motion due oscillatory surface. set of governing equations consists velocity, temperature, and mass transfer. These are first nondimensionalzed by using appropriate scaling. heat transfer process has been investigated in presence additional effects, such as Joule heating, viscous dissipation, generation source. well-known Buongiorno model employed for Brownian thermal migration nanoparticles. irreversibility system, source also various values parameters. finite difference method obtain numerical solution problem. obtained solutions subsequently integrated into artificial neural network (ANN) model. ANN trained famous Levenberg–Marquardt (LM) backpropagation algorithm. performance assessed a range statistical methods, mean squared error, regression analysis, curve fitting, error histograms. shows high accuracy, absolute errors ranging from 10−4 10−5. impact key parameters on fluid, heat, phenomena depicted graphically numerically. absence solutal convection, i.e., when Gc=0, it observed that convection significant However, case reverse effect absent, Gr=0. both (Gr=Gc=0) plays crucial shaping thermodynamics. velocity enhanced this case. Other physical quantities skin friction coefficient well Nusselt Sherwood number different Skin decays Grashof increased. increment random nanoparticles causes reduction numbers. Augmentation entropy profile enhancement magnetic Brinkman number. Consistency between computational results friction, number, observed, confirms accuracy our We believe beneficial biomedical engineering optimizing blood environments, microfluidic devices precise manipulation, industrial cooling systems enhancing efficiency. Additionally, they contribute geophysical dynamics improving understanding buoyancy-driven flows natural systems.

Язык: Английский

Процитировано

0