Impact of structural characteristics on thermal conductivity of foam structures revealed with machine learning

The foam-like structures have been shown to facilitate thermal transport efficiency by forming 3D-network heat conduction pathways, which shows broad applications in thermal management. However, the design of foam structures with desired thermal transport property is tricky because of the complexity of the structures and their relationship with the property. Here, we combine high-throughput numerical simulations and machine learning methods to systematically investigate the quantitative relationship between the structural characteristics and thermal transport properties of a variety of foam structures. The so-called explicit jump immersed interface approach that solves the transport equations by combing the fast Fourier transform (FFT) and BiCGStab methods is used to perform high-throughput calculations of over 700 different foam structures. The high accuracy and efficiency of the approach ensure the high quality of the characterization of the thermal transport properties, which also provides excellent database for building quantitative relationship by using machine learning. We select 69 out of 131 descriptors to exhaustively describe the structural characteristics, which are used as inputs of 6 machine learning models, including the DTR, GBR, GPR, RFR, SVR, and ANN, with the ANN outperforms the rest. We find that the volume fraction of struts and faces, as well as the porosity, though playing a predominant role, are not sufficient to completely determine the thermal conductivity of the foam structures. Incorporating additional structural parameters is essential for substantially improving the accuracy of thermal conductivity predictions, highlighting the necessity of employing machine learning models in the meticulous design of the microstructures. The promising way for design of highly conductive foams provided in this work may facilitate the thermal management in a broad of applications such as solar receivers and batteries, preservation of biological tissues, and so on.