含TiO₂纳米颗粒的纳米流体热导率的测量与预测

Measurement and Prediction of Thermal Conductivity of Nanofluids Containing TiO₂ Nanoparticles.

作者信息

Verma Kamalesh, Agarwal Ravi, Duchaniya R K, Singh Ramvir

出版信息

J Nanosci Nanotechnol. 2017 Feb;17(2):1068-075. doi: 10.1166/jnn.2017.12584.

DOI:10.1166/jnn.2017.12584

Abstract

In this paper, experimental results on thermal conductivity of TiO₂-Water, TiO₂-Ethylene Glycol and TiO₂-Engine Oil nanofluids in concentration range between 0.25 to 2 volume percent and temperature range of 10–70 °C is presented. Results show that thermal conductivity of nanofluids increases with increasing concentration and lower enhancement is observed with temperature as compared with concentration. Maximum enhancements are 9.3%, 15% and 8.9% for 2% volume of TiO₂ nanoparticles at 70 °C for TiO₂-Water, TiO₂-Ethylene Glycol and TiO₂-Engine Oil nanofluids respectively. 4-input and 1-output Artificial Neural Network (ANN) approach is used to predict thermal conductivity of nanofluids. Experimental results obtained were compared with some theoretical models and ANN approach. It is observed that ANN approach gives better predictions and is in good agreement with experimental results.

摘要

本文给出了二氧化钛-水、二氧化钛-乙二醇和二氧化钛-发动机油纳米流体在0.25%至2%（体积分数）浓度范围以及10-70°C温度范围内的热导率实验结果。结果表明，纳米流体的热导率随浓度增加而增大，与浓度相比，温度升高时热导率的增强幅度较小。在70°C时，对于二氧化钛-水、二氧化钛-乙二醇和二氧化钛-发动机油纳米流体，2%（体积分数）的二氧化钛纳米颗粒对应的最大增强率分别为9.3%、15%和8.9%。采用4输入1输出的人工神经网络（ANN）方法预测纳米流体的热导率。将获得的实验结果与一些理论模型和ANN方法进行了比较。结果表明，ANN方法能给出更好的预测结果，且与实验结果吻合良好。