Scholarlink

<p>In this study, a conjugate heat transfer with MHD effect inside a prismatic cavity containing CNT-Al<sub>2</sub>O<sub>3</sub>/H<sub>2</sub>O hybrid nanofluid has been predicted by artificial neural network (ANN). A Galerkin weighted residual approach is applied to determine the dimensionless governing equation with boundary constraints. The study investigates the significance of various dimensionless parameters, including the range of Richardson number (0.01 ≤ <i>Ri</i> ≤ 10), Reynolds number (50 ≤ <i>Re</i> ≤ 250), radiation parameter (0 ≤ <i>Rd</i> ≤ 2), Hartmann number (0 ≤ <i>Ha</i> ≤ 50), and nanoparticle volume fraction (0.01 ≤ <i>ϕ</i><sub><i>hnf</i></sub> ≤ 0.04) on temperature distribution, flow patterns, and average heat transfer rate. The empirical results indicate that <i>Nu</i><sub><i>av</i></sub> increases by: (i) 4.12% when <i>ϕ</i><sub><i>hnf</i></sub> rises from 1% to 4%, (ii) 88.34% for <i>Re</i> increasing from 50 to 250, (iii) 235.94% with <i>Rd</i> increasing from 0 to 2, (iv) 113.09% as Ri increases from 0.01 to 10. In contrast, <i>Nu</i><sub><i>av</i></sub> decreases by approximately 58.74% when <i>Ha</i> is increased from 0 to 50. Furthermore, the ANN technique uses the Levenberg- Marquardt back propagation method on 1024 datasets to find the average <i>Nu</i>, which illustrates a good correlation with the simulation results and&#xa0;an <i>R</i><sup><i>2</i></sup> of 0.99998 with an accuracy of 99%. The novelty of this work is the implication of ANN prediction on conjugate heat transfer with MHD effect inside a prismatic chamber consisting of hybrid nanofluid.</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 Prediction on MHD Conjugate Heat Transfer within a Prismatic Chamber Consisting of CNT-Al2O3/H2O Hybrid Nanofluid

  • 期刊论文

  • 发表日期：

    2025年7月15日

查看全文
  • R. M. Ziaur,
  • M. N. Hudha,
  • M. Al-Amin,
  • A. K. Azad,
  • M. F. Karim,
  • Md. Nasir Uddin,
  • M. M. Rahman,
  • M. J. H. Munshi

摘要

In this study, a conjugate heat transfer with MHD effect inside a prismatic cavity containing CNT-Al2O3/H2O hybrid nanofluid has been predicted by artificial neural network (ANN). A Galerkin weighted residual approach is applied to determine the dimensionless governing equation with boundary constraints. The study investigates the significance of various dimensionless parameters, including the range of Richardson number (0.01 ≤ Ri ≤ 10), Reynolds number (50 ≤ Re ≤ 250), radiation parameter (0 ≤ Rd ≤ 2), Hartmann number (0 ≤ Ha ≤ 50), and nanoparticle volume fraction (0.01 ≤ ϕhnf ≤ 0.04) on temperature distribution, flow patterns, and average heat transfer rate. The empirical results indicate that Nuav increases by: (i) 4.12% when ϕhnf rises from 1% to 4%, (ii) 88.34% for Re increasing from 50 to 250, (iii) 235.94% with Rd increasing from 0 to 2, (iv) 113.09% as Ri increases from 0.01 to 10. In contrast, Nuav decreases by approximately 58.74% when Ha is increased from 0 to 50. Furthermore, the ANN technique uses the Levenberg- Marquardt back propagation method on 1024 datasets to find the average Nu, which illustrates a good correlation with the simulation results and an R2 of 0.99998 with an accuracy of 99%. The novelty of this work is the implication of ANN prediction on conjugate heat transfer with MHD effect inside a prismatic chamber consisting of hybrid nanofluid.