Scholarlink

<p>The present study investigates the influence of an inclined magnetic field and Hall current on the free convective flow of a Jeffrey fluid within a channel under Navier-slip conditions. The governing nonlinear equations are transformed into a dimensionless form and solved using a hybrid computational framework that combines the spectral quasi-linearization method (SQLM) with a Levenberg-Marquardt-based artificial neural network (BANN-LM). The novelty of the work lies in integrating a high-accuracy numerical solver with a data-driven ANN model to efficiently predict velocity, cross-flow, and temperature profiles under coupled physical effects. The reliability of the developed BANN-LM framework is verified through a validation process that incorporates performance indices including mean squared error, regression analysis, and error histogram evaluation. Graphical illustrations present the key findings, highlighting the influence of thermophysical parameters. The velocity and cross-flow velocity increase with larger values of the inclination angle, Jeffrey parameter, and magnetic strength. Conversely, the temperature rises with a higher Hall current and Jeffrey parameter but decreases as the inclination angle and magnetic parameter intensify. The proposed framework provides a robust and computationally efficient approach for modeling complex Non-Newtonian magnetohydrodynamic flows.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Artificial neural network-based computational modeling of Jeffrey fluid in sloping channel with Navier-slip boundaries using Levenberg–Marquardt algorithm

  • 期刊论文

  • 发表日期：

    2026年7月1日

查看全文
  • Ravi Mahla,
  • Kaladhar Kolla

摘要

The present study investigates the influence of an inclined magnetic field and Hall current on the free convective flow of a Jeffrey fluid within a channel under Navier-slip conditions. The governing nonlinear equations are transformed into a dimensionless form and solved using a hybrid computational framework that combines the spectral quasi-linearization method (SQLM) with a Levenberg-Marquardt-based artificial neural network (BANN-LM). The novelty of the work lies in integrating a high-accuracy numerical solver with a data-driven ANN model to efficiently predict velocity, cross-flow, and temperature profiles under coupled physical effects. The reliability of the developed BANN-LM framework is verified through a validation process that incorporates performance indices including mean squared error, regression analysis, and error histogram evaluation. Graphical illustrations present the key findings, highlighting the influence of thermophysical parameters. The velocity and cross-flow velocity increase with larger values of the inclination angle, Jeffrey parameter, and magnetic strength. Conversely, the temperature rises with a higher Hall current and Jeffrey parameter but decreases as the inclination angle and magnetic parameter intensify. The proposed framework provides a robust and computationally efficient approach for modeling complex Non-Newtonian magnetohydrodynamic flows.