Temporal stability and non-unique solution of reacting Eyring Powell flows over shrinking wedges using neural networks

In this paper, author presents an innovative artificial intelligence techniques based on deep learning simulation algorithms due to wide range of applications in science, robotics and engineering. The deep learning simulation algorithm using the Levenberg-Marquardt Scheme with Back Propagation Neura...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Engineering applications of artificial intelligence Ročník 141; s. 109828
Hlavní autoři: Khan, M.I., Zeeshan, A., Arain, M.B., Alqahtani, A.S., Malik, M.Y.
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier Ltd 01.02.2025
Témata:
Application of AI
Chemical reaction
Eyring powell fluid
Levenberg–marquardt algorithm/scheme
Mixed convection
Stability analysis
Wedge flow
Levenberg–marquardt algorithm/scheme
Wedge flow
Chemical reaction
Application of AI
Stability analysis
Mixed convection
Eyring powell fluid
ISSN:0952-1976
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:In this paper, author presents an innovative artificial intelligence techniques based on deep learning simulation algorithms due to wide range of applications in science, robotics and engineering. The deep learning simulation algorithm using the Levenberg-Marquardt Scheme with Back Propagation Neural Networks (LMS-BPNN) is evaluated in the flow investigation of the chemically reacting non-Newtonian fluid. An efficient similarity variable is applied to change Partial Differential Equations (PDEs) of a considered flow problem into dimensionless Ordinary Differential Equations (ODEs). An error is found to be 10−4 with function fit for scenarios 1–5, while performance in terms of mean squared error is found to be 10−10. It is seen that flow response output i.e. f′(η), θ(η) and ϕ(η) meet boundary requirements for different scenarios 1–5 with the deep learning-based technique LMS-BPNN. The dual solution is evaluated for flow response output parameter i.e. Cfx, Nux and Shx for different parameters with the proposed LMS-BPNN. The dual nature of Cfx is calculated for three cases of scenario 1, and the critical value is found to be −0.96195, −1.0581, and −1.11. The dual solution of Shx is calculated for various values of Kc=0.2,0.4,0.6, and it is found that λc=−1.1175 remains the same. The perturbation scheme is applied to the boundary layer problem to obtain the eigenvalues problem. The unsteady solution f(η,τ) converges to steady solution fo(η) for τ→∞ when γ≥0. However, an unsteady solution f(η,τ) diverges to a steady solution fo(η) for τ→∞ when γ<0. It is found that the boundary layer thickness for the second (lower branch) solution is higher than the first (upper branch) solution. This investigation is the evidence that the first (upper branch) solution is stable and reliable.
ISSN:0952-1976
DOI:10.1016/j.engappai.2024.109828