A Neural Regression Model for Predicting Thermal Conductivity of CNT Nanofluids with Multiple Base Fluids

• Published in: Journal of thermal science 2021-11, Vol.30 (6), p.1908-1916
• Main Authors: Zou, Hanying, Chen, Cheng, Zha, Muxi, Zhou, Kangneng, Xiao, Ruoxiu, Feng, Yanhui, Qiu, Lin, Zhang, Xinxin, Wang, Zhiliang
• Format: Article
• Language: English
• Subjects: Carbon nanotubes; Classical and Continuum Physics; Engineering Fluid Dynamics; Engineering Thermodynamics; Ethylene glycol; Heat and Mass Transfer; Heat conductivity; Heat transfer; Machine learning; Nanofluids; Physics; Physics and Astronomy; Regression models; Thermal conductivity
• ISSN: 1003-2169; 1993-033X
• DOI: 10.1007/s11630-021-1497-1

High thermal conductivity of carbon nanotube nanofluids ( k nf ) has received great attention. However, the current researches are limited by experimental conditions and lack a comprehensive understanding of k nf variation law. In view of proposition of data-driven methods in recent years, using experimental data to drive prediction is an effective way to obtain k nf , which could clarify variation law of k nf and thus greatly save experimental and time costs. This work proposed a neural regression model for predicting k nf . It took into account four influencing factors, including carbon nanotube diameter, volume fraction, temperature and base fluid thermal conductivity ( k f ). Where, four conventional fluids with k f , including R113, water, ethylene glycol and ethylene glycol-water mixed liquid were considered as base fluid considers. By training this model, it can predict k nf with different factors. Also, change law of four influencing factors considered on the k nf enhancement has discussed and the correlation between different influencing factors and k nf enhancement is presented. Finally, compared with nine common machine learning methods, the proposed neural regression model shown the highest accuracy among these.