Scholarlink

<p>The growing demand for efficient heat transfer and energy conversion systems require advanced working fluids with superior thermal and electrical transport properties. In this study, investigation on the thermophysical behaviour of Al<sub>2</sub>O<sub>3</sub>/MWCNT–DI water hybrid nanofluids through comprehensive experimental characterization and predictive modelling was carried out. Hybrid nanofluids of 0.05–0.25&#xa0;vol% were synthesized via a two-step method, and viscosity, electrical conductivity, and thermal conductivity were measured over temperatures ranging from 15 to 60&#xa0;°C. Results show that viscosity decreases with temperature but increases with nanoparticle concentration, while both electrical and thermal conductivities rise significantly with temperature and nanoparticle concentration. At 0.25&#xa0;vol% and 60&#xa0;°C, electrical conductivity exceeded 2400&#xa0;μS&#xa0;cm<sup>−1</sup>, and thermal conductivity improved by ~ 35 to 50% relative to DI water, confirming strong synergistic enhancement mechanisms. Empirical correlations developed for relative viscosity, relative electrical conductivity, and relative thermal conductivity demonstrated high predictive accuracy, with R<sup>2</sup> values between 0.938 and 0.991 and deviations within ± 2 to ± 5%. Advanced machine-learning models; Artificial Neural Network (ANN), Adaptive Neuro-Fuzzy Inference Systems- Fuzzy C-Means (ANFIS-FCM), and Particle Swarm Optimization (PSO)-ANFIS, were used to train the experimental datasets to model nonlinear interactions between temperature, concentration, and thermophysical responses. Among these, the ANFIS-FCM, yielded the highest accuracy coupled with the lowest error values across all the studied properties, e.g. Viscosity (ANFIS-FCM5: RMSE = 0.0140, MAE = 0.0114, MAPE = 1.3505%, U = 0.0153). Overall, the combined experimental and intelligent-modelling framework provides robust predictive capability for optimizing Al<sub>2</sub>O<sub>3</sub>/MWCNT hybrid nanofluids in thermal management, cooling systems, and electrohydrodynamic applications, supporting their integration into next-generation energy and process engineering technologies.</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 characterization and predictive modelling of Al2O3/MWCNT hybrid nanofluid thermophysical properties using ANN, ANFIS, FCM and hybrid techniques

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月26日

查看全文
  • Modaser Momin,
  • Emmanuel O. Atofarati,
  • Stephen Oladipo,
  • Victor O. Adogbeji,
  • Luke Ajuka,
  • Solomon O. Giwa,
  • Mohsen Sharifpur,
  • Josua P. Meyer

摘要

The growing demand for efficient heat transfer and energy conversion systems require advanced working fluids with superior thermal and electrical transport properties. In this study, investigation on the thermophysical behaviour of Al2O3/MWCNT–DI water hybrid nanofluids through comprehensive experimental characterization and predictive modelling was carried out. Hybrid nanofluids of 0.05–0.25 vol% were synthesized via a two-step method, and viscosity, electrical conductivity, and thermal conductivity were measured over temperatures ranging from 15 to 60 °C. Results show that viscosity decreases with temperature but increases with nanoparticle concentration, while both electrical and thermal conductivities rise significantly with temperature and nanoparticle concentration. At 0.25 vol% and 60 °C, electrical conductivity exceeded 2400 μS cm−1, and thermal conductivity improved by ~ 35 to 50% relative to DI water, confirming strong synergistic enhancement mechanisms. Empirical correlations developed for relative viscosity, relative electrical conductivity, and relative thermal conductivity demonstrated high predictive accuracy, with R2 values between 0.938 and 0.991 and deviations within ± 2 to ± 5%. Advanced machine-learning models; Artificial Neural Network (ANN), Adaptive Neuro-Fuzzy Inference Systems- Fuzzy C-Means (ANFIS-FCM), and Particle Swarm Optimization (PSO)-ANFIS, were used to train the experimental datasets to model nonlinear interactions between temperature, concentration, and thermophysical responses. Among these, the ANFIS-FCM, yielded the highest accuracy coupled with the lowest error values across all the studied properties, e.g. Viscosity (ANFIS-FCM5: RMSE = 0.0140, MAE = 0.0114, MAPE = 1.3505%, U = 0.0153). Overall, the combined experimental and intelligent-modelling framework provides robust predictive capability for optimizing Al2O3/MWCNT hybrid nanofluids in thermal management, cooling systems, and electrohydrodynamic applications, supporting their integration into next-generation energy and process engineering technologies.

Graphical abstract