Scholarlink

<p>Nanofluid has sparked considerable interest due to its intriguing thermal, mechanical, and optical attributes. This study used Machine learning (ML) to investigate Casson nanofluid flow due to a curved stretching sheet with graphene oxide nanoparticles in engine oil base liquid under the magnetic field and variable thermal conductivity effects. Key physical phenomena including heat source/sink, Joule heating, and activation energy effects are incorporated, with Brownian and thermophoresis diffusivity impacts employed according to the Buongiorno model. The present work employs the homotopy analysis method (HAM) for simulation, with results trained using artificial neural network (ANN) applications. Using the Levenberg-Marquard Scheme-based Backpropagation Method (NN- BLMS) within feed-forward neural networks, the results obtained from the ANN are validated. Our finding shows the ANN model reliably predicts nanofluid behavior with 94% accuracy based on regression analysis. Heat transfer rate increases by 34.99% when the Eckert number rises from 0.1 to 0.7. Also, the nanofluid velocity decreases due to magnetic field effects and Lorentz force resistance. Higher heat and mass transfer rates are predicted for the Brownian motion parameter, while fluid concentration decreases with Schmidt number. ANN implementation enables faster convergence and greater efficiency compared to alternative models. In the present ANN model, the HAM method reduced computational time with a mean squared error (MSE) below 10<sup>−4</sup>. These findings are valuable for optimizing nanofluid applications in engineering and industry, particularly in heat exchangers, microfluidic cooling systems, solar thermal collectors, and biomedical drug delivery.</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

An exhaustive examination of the thermal and mechanical properties of nanofluids containing graphene oxide nanoparticles through the utilization of artificial neural network

  • 期刊论文

  • 发表日期：

    2025年7月2日

查看全文
  • Muhammad Awais,
  • Shahzad Ahmad,
  • Muhammad Sheraz Junaid,
  • Muhammad Ramzan,
  • Abdulrahman A. Almehizia,
  • Amer Alhaj Zen

摘要

Nanofluid has sparked considerable interest due to its intriguing thermal, mechanical, and optical attributes. This study used Machine learning (ML) to investigate Casson nanofluid flow due to a curved stretching sheet with graphene oxide nanoparticles in engine oil base liquid under the magnetic field and variable thermal conductivity effects. Key physical phenomena including heat source/sink, Joule heating, and activation energy effects are incorporated, with Brownian and thermophoresis diffusivity impacts employed according to the Buongiorno model. The present work employs the homotopy analysis method (HAM) for simulation, with results trained using artificial neural network (ANN) applications. Using the Levenberg-Marquard Scheme-based Backpropagation Method (NN- BLMS) within feed-forward neural networks, the results obtained from the ANN are validated. Our finding shows the ANN model reliably predicts nanofluid behavior with 94% accuracy based on regression analysis. Heat transfer rate increases by 34.99% when the Eckert number rises from 0.1 to 0.7. Also, the nanofluid velocity decreases due to magnetic field effects and Lorentz force resistance. Higher heat and mass transfer rates are predicted for the Brownian motion parameter, while fluid concentration decreases with Schmidt number. ANN implementation enables faster convergence and greater efficiency compared to alternative models. In the present ANN model, the HAM method reduced computational time with a mean squared error (MSE) below 10−4. These findings are valuable for optimizing nanofluid applications in engineering and industry, particularly in heat exchangers, microfluidic cooling systems, solar thermal collectors, and biomedical drug delivery.