Scholarlink

<p>This study presents a comprehensive experimental and computational investigation of laminar radial airflow between stationary parallel disks under uniform heat flux boundary conditions. A custom-fabricated spiral coil heater was employed to ensure consistent thermal loading, overcoming the flux non-uniformity inherent in earlier radial heating approaches. The experimental matrix spans five gap ratios (17.36–86.8), four heat flux levels (194.24–824.54 W/m<sup>2</sup>), and five Reynolds numbers (60–100), resulting in 100 distinct test conditions. Detailed local and average Nusselt number distributions were acquired, with results highlighting the dominant influence of geometric confinement on convective performance. A second-order response surface model, statistically validated through ANOVA, revealed the gap ratio as the most influential parameter (81.2% contribution), while heat flux exhibited nonlinear behavior with significant second-order and interaction effects. To optimize heat transfer performance, an artificial neural network (ANN) surrogate model was integrated with two metaheuristic algorithms—Teaching–Learning-Based Optimization (TLBO) and Jaya. Both algorithms converged to a global optimum of Avg. Nusselt number ≈ 48.1, with high agreement between experimental and predicted outcomes. Comparative validation against seminal works by Mochizuki et al. and Singh confirms the fidelity of the Nusselt number trends, while demonstrating methodological advancements in boundary condition uniformity, statistical modeling, and AI-driven optimization. The findings provide new insights for designing compact and efficient thermal systems using confined laminar flow.</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

Experimental investigation and hybrid AI-Based optimization of laminar convective heat transfer between stationary parallel disks

  • 期刊论文

  • 发表日期：

    2026年2月18日

查看全文
  • Rakesh Kumar Yadu,
  • Dinesh Kumar Singh

摘要

This study presents a comprehensive experimental and computational investigation of laminar radial airflow between stationary parallel disks under uniform heat flux boundary conditions. A custom-fabricated spiral coil heater was employed to ensure consistent thermal loading, overcoming the flux non-uniformity inherent in earlier radial heating approaches. The experimental matrix spans five gap ratios (17.36–86.8), four heat flux levels (194.24–824.54 W/m2), and five Reynolds numbers (60–100), resulting in 100 distinct test conditions. Detailed local and average Nusselt number distributions were acquired, with results highlighting the dominant influence of geometric confinement on convective performance. A second-order response surface model, statistically validated through ANOVA, revealed the gap ratio as the most influential parameter (81.2% contribution), while heat flux exhibited nonlinear behavior with significant second-order and interaction effects. To optimize heat transfer performance, an artificial neural network (ANN) surrogate model was integrated with two metaheuristic algorithms—Teaching–Learning-Based Optimization (TLBO) and Jaya. Both algorithms converged to a global optimum of Avg. Nusselt number ≈ 48.1, with high agreement between experimental and predicted outcomes. Comparative validation against seminal works by Mochizuki et al. and Singh confirms the fidelity of the Nusselt number trends, while demonstrating methodological advancements in boundary condition uniformity, statistical modeling, and AI-driven optimization. The findings provide new insights for designing compact and efficient thermal systems using confined laminar flow.