Scholarlink

<p>This work presents an integrated approach for analyzing and optimizing the performance of a counter-flow heat exchanger by combining experimental measurements, CFD simulations, and Artificial Neural Network (ANN) modeling. The study includes an experimental investigation of the overall heat transfer coefficient (U) for hot water mass flow rates ranging from 0.021 to 0.046&#xa0;kg&#xa0;s⁻<sup>1</sup>, cold water mass flow rates from 0.022 to 0.11&#xa0;kg&#xa0;s⁻<sup>1</sup>, and a log mean temperature difference (LMTD) between 9.7 and 20.4&#xa0;°C. A steady-state CFD simulation was performed in ANSYS Fluent and validated against the experimental data, showing a maximum deviation of ± 7%, confirming model accuracy. Additionally, an ANN model was developed to predict U-values from operating parameters, achieving prediction errors within ± 5% and correlation coefficients above 0.97 across training, validation, and testing datasets. The integration of these methods establishes a robust framework for heat exchanger design optimization, real-time performance monitoring, and predictive maintenance, enhancing both research and industrial applications.</p> Graphical abstract <p></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 numerical simulation of a counter flow heat exchanger: artificial neural networks

  • 期刊论文

  • 发表日期：

    2026年2月18日

查看全文
  • Santhosh Kumar Dubba,
  • P. Ramana Sai,
  • E. Tarakeswara Rao,
  • Kelli Durgaprasad,
  • Roopsandeep Bammidi,
  • Turali Narayana,
  • Talabaktula S. Viswanadham

摘要

This work presents an integrated approach for analyzing and optimizing the performance of a counter-flow heat exchanger by combining experimental measurements, CFD simulations, and Artificial Neural Network (ANN) modeling. The study includes an experimental investigation of the overall heat transfer coefficient (U) for hot water mass flow rates ranging from 0.021 to 0.046 kg s⁻1, cold water mass flow rates from 0.022 to 0.11 kg s⁻1, and a log mean temperature difference (LMTD) between 9.7 and 20.4 °C. A steady-state CFD simulation was performed in ANSYS Fluent and validated against the experimental data, showing a maximum deviation of ± 7%, confirming model accuracy. Additionally, an ANN model was developed to predict U-values from operating parameters, achieving prediction errors within ± 5% and correlation coefficients above 0.97 across training, validation, and testing datasets. The integration of these methods establishes a robust framework for heat exchanger design optimization, real-time performance monitoring, and predictive maintenance, enhancing both research and industrial applications.

Graphical abstract