基于人工神经网络分析面朝下加热池式过渡沸腾

Prediction of Transition Boiling Concerning Downward Facing Horizontal Surface in Confined Space Using Artificial Neural Network

基于不同间隙尺寸下、水平小间隙内面朝下加热自然循环池式沸腾的实验研究,利用过渡沸腾段实验数据训练出一成熟的人工神经网络(ANN)。运用该人工神经网络系统分析了壁面过热度Δtw、无量纲水平间隙δ/D、Pr、Ra对换热性能的影响。在此基础上,拟合出一用于计算水平小间隙内面朝下加热过渡沸腾Nu的经验关系式,该关系式计算值与实验值符合较好。

An artificial neural network （ANN） for predicting Nusselt number was trained successfully based on the data of transition boiling experiment, which was conducted in confined space with downward facing surface at atmospheric pressure. The effects of wall superheat, Atw, the ratio of the gap size to the diameter of heated surface, 3/D, Prandtl number and Rayleigh number on Nusselt number under transition boiling condition were analyzed based on the trained ANN. A correlation used to accurately predict the natural convection heat transfer under the present condition was obtained and it provides a reasonable agreement against the experimental data.