高孔隙率泡沫金属复合相变材料有效热导率的预测

Prediction of effective thermal conductivity of high-porosity foam metal composite phase change material

中文关键词: 有效热导率  泡沫金属  复合材料  数值模拟  热传导

英文关键词: effective thermal conductivity  foam metal  composite materials  numerical simulation  heat conduction

中文摘要:

      基于十四面体泡沫金属的单元晶胞结构，提出了一种高孔隙率开孔泡沫金属复合材料有效热导率的预测模型。热量沿热流方向在金属骨架和填充介质中的并行传递，推导得出包含结构参数f（节点边长与韧带半径之比）和d（韧带半径与韧带长度之比）的有效热导率计算式。在孔隙率0.889≤ ε ≤0.977时，针对单元晶胞的十六分之一结构，采用数值模拟的方法并结合实验数据的对比分析，校正了与有效热导率紧密相关的两个无量纲参数f和d，并以孔隙率的二阶多项式拟合出f变量的函数关系，以孔隙率的三阶多项式拟合出d变量的函数关系，从而确定出有效热导率的预测模型。通过与其他几种预测模型和实验数据的比较分析，结果表明：验证了当前预测模型具有较高的正确性和有效性，与实验值的平均相对误差ARD和相对均方根误差RMS分别为2.93%和3.50%，相对其他模型具有更高的预测精度。

英文摘要:

      Based on the unit cell structure of the tetrahedral metal foam, a predictive model of the effective thermal conductivity of the high-porosity open-cell foam metal composite was proposed. The parallel transfer of heat in the metal framework and the filling medium along the direction of heat flow was derived. The ef-fective thermal conductivity calculation formula including structural parameters f (ratio of node side length to ligament radius) and d (ratio of ligament radius to ligament length) was derived. Within the range of porosity 0.889≤ ε ≤0.976, aiming at the one-sixteenth structure of the unit cell, the method of numerical simulation and the comparative analysis of experimental data were used to correct two dimen-sionless parameters closely related to the effective thermal conductivity f and d. The second-order poly-nomial of the porosity was used to fit the functional relationship of the variable f, and the third-order pol-ynomial of the porosity was used to fit the functional relationship of the d variable, so as to determine the effective thermal conductivity prediction model. Through comparison and analysis with several other pre-diction models and experimental data, it was verified that the present prediction model has high accuracy and effectiveness. The average relative error ARD and relative root mean square error RMS from the ex-perimental value are 2.93% and 3.5%, respectively, which has higher prediction accuracy than other mod-els.