TY - JOUR
T1 - Machine-assisted simulation of heat and mass transfer through a porous Riga surface with chemical reactions and radiation effects
AU - Habib, Shazia
AU - Nasir, Saleem
AU - Khan, Zeeshan
AU - Berrouk, Abdallah S.
AU - Islam, Saeed
AU - Aamir, Asim
N1 - Publisher Copyright:
© Akadémiai Kiadó Zrt 2025.
PY - 2025
Y1 - 2025
N2 - The principal objective of the current work is to investigate the combined effects of thermophoretic diffusion and chemical reactions on the dynamics of Williamson fluid, as well as heat and mass transmission, over a continuously magnetized Riga surface. In this analysis, variable thermal properties are also considered. To solve the model dimensionless nonlinear system of equations, an innovative scheme is introduced, the Morlet wavelet neural network algorithm. This scheme validates unparalleled performance and accuracy. Moreover, the statistical procedures such as mean, standard deviation, MSE, MAD, and ENSE validate the exceptional precision of the proposed scheme. The MSE ranges from 100 to 10-10; MAD is within the interval of 10-01 and 10-10; ENSE ranges from 100 to 10-15, while fitness curves fall within 100 and 10-10 for all four cases across multiple runs. Additionally, using visual illustrations of the outcome improves the overall understanding of fundamental phenomena. This research work emphasizes the potential practical applications in fluid dynamics and encourages further research to expose new insights and advancements in the field. Combining ANN techniques with numerical simulations opens new research avenues and promotes progress in engineering and fluid dynamics.
AB - The principal objective of the current work is to investigate the combined effects of thermophoretic diffusion and chemical reactions on the dynamics of Williamson fluid, as well as heat and mass transmission, over a continuously magnetized Riga surface. In this analysis, variable thermal properties are also considered. To solve the model dimensionless nonlinear system of equations, an innovative scheme is introduced, the Morlet wavelet neural network algorithm. This scheme validates unparalleled performance and accuracy. Moreover, the statistical procedures such as mean, standard deviation, MSE, MAD, and ENSE validate the exceptional precision of the proposed scheme. The MSE ranges from 100 to 10-10; MAD is within the interval of 10-01 and 10-10; ENSE ranges from 100 to 10-15, while fitness curves fall within 100 and 10-10 for all four cases across multiple runs. Additionally, using visual illustrations of the outcome improves the overall understanding of fundamental phenomena. This research work emphasizes the potential practical applications in fluid dynamics and encourages further research to expose new insights and advancements in the field. Combining ANN techniques with numerical simulations opens new research avenues and promotes progress in engineering and fluid dynamics.
KW - Chemical reaction
KW - Morlet wavelet neural networks
KW - Radiation flux
KW - Riga plate
KW - Thermophoretic diffusion
KW - Williamson fluid
UR - https://www.scopus.com/pages/publications/105003167596
U2 - 10.1007/s10973-025-14191-y
DO - 10.1007/s10973-025-14191-y
M3 - Article
AN - SCOPUS:105003167596
SN - 1388-6150
JO - Journal of Thermal Analysis and Calorimetry
JF - Journal of Thermal Analysis and Calorimetry
M1 - 115927
ER -