Thermophysical properties of water, water and ethylene glycol mixture-based nanodiamond+Fe3O4 hybrid nanofluids: An experimental assessment and application of data-driven approaches
Thermophysical properties of water, water and ethylene glycol mixture-based nanodiamond+Fe3O4 hybrid nanofluids: An experimental assessment and application of data-driven approaches
复制标题
水、水和乙二醇混合物基纳米金刚石 Fe3O4 混合纳米流体的热物理性质：数据驱动方法的实验评估和应用
DOI：
10.1016/j.molliq.2021.117944
复制
发表时间：
2021
影响因子：
6
通讯作者：
Changhe Li
中科院分区：
文献类型：
--
作者：
Z. Said;M. Jamei;L. Syam Sundar;A. K. Pandey;A. Allouhi;Changhe Li
本文采用多元线性回归（MLR）和多元线性交互回归（MLRI）两种数据驱动模型，对水、水和乙二醇混合物基纳米金刚石+Fe3O4杂化纳米流体的热物理性质进行了实验和数值模拟研究。三种类型的基液，例如（i）水，（ii）40：60%水和乙二醇的混合物，和（iii）60：40%水和乙二醇的混合物，用于制备杂化纳米流体。对于所有基液，使用0.05%、0.1%和0.2%的增塑剂。结果表明，与水的数据相比，热导率和粘度的较高值在η = 0.2%和60° C下分别为17.76%和72.9%。同样，对于40：60%水和乙二醇混合物基纳米金刚石+Fe3O4杂化纳米流体，与基液相比，在60° C时，热导率和粘度分别提高了14.65%和79.01%。然而，在60° C下，以60：40%的水和乙二醇混合物为基础的纳米金刚石+Fe3O4杂化纳米流体的热导率和粘度最大值分别为基础流体的12.79%和50.84%。基于实验数据提出了经验关联式和数据驱动关联式。所开发的数据驱动模型产生了强大的个体关系，与相应的经验相关性相比，通过上级性能来预测所有类型的混合纳米流体的热物理性质。结果表明，对于导热系数、密度、比热和粘度的估计，MLRI（R= 0.9996）、MLR（R= 0.99989）、MLR（R = 0.999998）和MLRI（R= 0.9857）具有最佳的预测性能。
This paper aims to study the thermophysical properties of water, water and ethylene glycol mixture-based nanodiamond+ Fe 3 O 4 hybrid nanofluids experimentally and numerically using two data-driven approaches, namely, Multivariate linear regression (MLR) and Multivariate linear regression with interaction (MLRI) models. Three types of base fluids, such as (i) water,(ii) 40: 60% water and ethylene glycol mixture, and (iii) 60: 40% water and ethylene glycol mixture, were used to prepare the hybrid nanofluids. For all the base fluid, ϕ is used as 0.05%, 0.1%, and 0.2%. Results indicate that, higher values of thermal conductivity and viscosity is 17.76% and 72.9% at ϕ= 0.2% and 60° C in comparison to water data. Similarly, for 40: 60% water and ethylene glycol mixturebased nanodiamond+ Fe 3 O 4 hybrid nanofluid, the maximum thermal conductivity and viscosity enhancements are 14.65% and 79.01% at ϕ= 0.2% and at 60° C compared to the base fluid. However, the maximum thermal conductivity and viscosity enhancements for ϕ= 0.2% and at 60° C of 60: 40% water and ethylene glycol mixture-based nanodiamond+ Fe 3 O 4 hybrid nanofluid are 12.79% and 50.84% over the basefluid data. Empirical and data-driven correlations were proposed based on the experimental data. The developed data-driven models resulted in robust individual relationships to predict the thermophysical properties of all types of hybrid nanofluids by superior performance compared to corresponding empirical correlations. The reported results exhibited that for thermal conductivity, density, Specific heat, and viscosity estimation, the MLRI (R= 0.9996), MLR (R= 0.99989), MLR (R= 0.9999998), and MLRI (R= 0.9857) had the best predictive performance.