Work

Machine learning integrated simulation of thermal convective flow dynamics of radiative nanomaterial using Buongiorno model

Year: 2026
Type: article
Abstract: Present study deals with numerical and predictive analysis of Maxwell fluid flow over a permeable surface. Various physical phenomena are taken under consideration such as double diffusion, first order chemical reaction and heat source in heat and mass equations. Proposed model resolved with help of Lie transformations. Transformed ODEs were resolved using the well-known numerical algorithm Bvp4c in MATLAB, and datasets were created for application of machine learning algorithm (i.e. Levenberg-Marquardt algorithm). Furthermore, the uniformity and stability of the approach were ensured. For verification, the results were compared to previously published work. Velocity, thermal-mass profiles were evaluated via ANN models. Tables and graphs were utilized to display the numerical results. Considered profiles versus double diffusion and porous parameter were investigated. Outcomes showed that radiation as well as heat source augmented flow's heat production. It is also interesting to notice that porosity and Dufour variables have a significant influence on the flow of skin frictions, Sherwood, and Nusselt numbers. Smaller values of MSE indicated that there are very small errors between actual and predicted results and accuracy of proposed ANN model verified. Similarly, we obtained a regression coefficient equal to one, which also indicated linear correlation between inputs and targeted values. Nonetheless, current results have applications in fields of heat exchangers, steal industries, petroleum reservoirs, nuclear waste disposal and cooling applications among others.
Language: English
Open Access status: gold
Estimated APC paid: $2,090
OpenAlex ID: W7237219489
FWCI:
Cites: 33
Cited by:
Related to: 10