Thermal conductivity of CNT/Al composites is considered to be an important property on composites design. However, the effect of microstructure on thermal conductivity has few been reported. In this study, a 3D multi-scale finite element model was developed to predict the effective thermal conductivity of CNT/Al composites, and the influences of CNT configuration, heat transfer direction, interfacial thermal resistance and volume fraction were investigated in detail. Five configurations of CNT were constructed including randomly arranged, evenly oriented, layered, bundled and networked. The results show that the CNT configuration plays a major role for the effective thermal conductivity of CNT/Al composites. The CNT/Al composite with layered configuration has the highest thermal conductivity, and when the volume fraction of CNTs is about 3.7 %, the CNT/Al composite displays an excellent effective thermal conductivity of 400 W/m·K, which increases about 84 % than that of the Al alloy. Moreover, the interface and thermal loading direction also have obvious effect on the thermal conductivity of the composites. The calculated results are in good agreement with experimental data existing in literature. The obtained findings in this study could provide a good theoretical basis for designing high thermal conductivity CNT/Al composites.

CNT/Al 复合材料的热导率被认为是复合材料设计的一个重要特性。然而，微观结构对热导率的影响鲜有报道。在这项研究中，开发了一个 3D 多尺度有限元模型来预测 CNT/Al 复合材料的有效热导率，并详细研究了 CNT 配置、传热方向、界面热阻和体积分数的影响。构建了五种结构的碳纳米管，包括随机排列、均匀取向、分层、成束和网络。结果表明，CNT 配置对 CNT/Al 复合材料的有效热导率起着主要作用。层状结构的 CNT/Al 复合材料具有最高的热导率，当 CNTs 的体积分数约为 3.7% 时，CNT/Al 复合材料显示出 400 W/m·K 的优异有效热导率，比铝合金提高了约 84%。此外，界面和热载荷方向对复合材料的热导率也有明显的影响。计算结果与文献中已有的实验数据吻合较好。本研究结果可为设计高导热碳纳米管/铝复合材料提供良好的理论基础。