Scholarlink

<p>Active thermal control of automotive engines is vital for improving performance, robustness and conventional coolants like water and air frequently bring insufficient heat dissipation. To counterpart this ill effect, the propose study examines and investigate the performance enrichment of an automotive radiator using Al₂O₃–ethylene glycol (EG) nanofluids concluded by collective numerical, experimental, and data-driven approach. Initially, three-dimensional radiator model was developed using AUTOCAD software and computational study was performed in ANSYS-FLUENT, using a pressure-based model and a realizable k–ε turbulence model. The Grid-independence test was directed to ensure numerical accuracy, and CFD predictions were experimentally validated through experimental radiator test rig. The CFD result for pure EG predicted an outlet temperature of 70.11&#xa0;°C that was closely agreed with the experimental value of 69.11&#xa0;°C confirming the consistency of the numerical model. The blend of Al₂O₃ nanoparticles significantly enhanced heat transfer characteristics, whereas growing nanoparticle blend leads to progressive reductions in outlet temperature. At 2% concentration, the outlet temperature reduced to 65.51&#xa0;°C, while the maximum cooling performance was achieved at 5% concentration, yielding an outlet temperature of 60.87&#xa0;°C. Although the pressure drop amplified moderately with nanoparticle loading—from 36.79&#xa0;Pa at 1% to 38.72&#xa0;Pa at 5%, the rise remained relatively small, representing a constructive trade-off between thermal enhancement and hydraulic disadvantage. The highest PEC value of 1.83 was obtained at 5% Al₂O₃ concentration, demonstrating that the enhancement in heat transfer outweighs the associated pressure drop penalty. In addition, a machine learning based predictive framework has been proposed in the present work. This outline has been developed to evaluate the radiator outlet temperature and pressure drop using nanoparticle concentration and functioning settings that helped in rapid performance prediction and parametric optimization. Overall, the combined CFD, experimental, and data-driven findings shows that Al₂O₃–EG nanofluids are a capable coolant choice for high-performance radiator cooling applications.</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

Intelligent performance prediction of nanoparticle-enhanced automotive radiator cooling using CFD and machine learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月6日

查看全文
  • Jitendra B. Satpute,
  • Dilip Kumar Jang Bahadur Saini,
  • Pooja Sharma,
  • Bipin Kumar Rai,
  • Pravin M. Thorat,
  • Gautam Kumar,
  • Kumar Dilip

摘要

Active thermal control of automotive engines is vital for improving performance, robustness and conventional coolants like water and air frequently bring insufficient heat dissipation. To counterpart this ill effect, the propose study examines and investigate the performance enrichment of an automotive radiator using Al₂O₃–ethylene glycol (EG) nanofluids concluded by collective numerical, experimental, and data-driven approach. Initially, three-dimensional radiator model was developed using AUTOCAD software and computational study was performed in ANSYS-FLUENT, using a pressure-based model and a realizable k–ε turbulence model. The Grid-independence test was directed to ensure numerical accuracy, and CFD predictions were experimentally validated through experimental radiator test rig. The CFD result for pure EG predicted an outlet temperature of 70.11 °C that was closely agreed with the experimental value of 69.11 °C confirming the consistency of the numerical model. The blend of Al₂O₃ nanoparticles significantly enhanced heat transfer characteristics, whereas growing nanoparticle blend leads to progressive reductions in outlet temperature. At 2% concentration, the outlet temperature reduced to 65.51 °C, while the maximum cooling performance was achieved at 5% concentration, yielding an outlet temperature of 60.87 °C. Although the pressure drop amplified moderately with nanoparticle loading—from 36.79 Pa at 1% to 38.72 Pa at 5%, the rise remained relatively small, representing a constructive trade-off between thermal enhancement and hydraulic disadvantage. The highest PEC value of 1.83 was obtained at 5% Al₂O₃ concentration, demonstrating that the enhancement in heat transfer outweighs the associated pressure drop penalty. In addition, a machine learning based predictive framework has been proposed in the present work. This outline has been developed to evaluate the radiator outlet temperature and pressure drop using nanoparticle concentration and functioning settings that helped in rapid performance prediction and parametric optimization. Overall, the combined CFD, experimental, and data-driven findings shows that Al₂O₃–EG nanofluids are a capable coolant choice for high-performance radiator cooling applications.