Abstract:
Tribology mainly studies the friction, lubrication and wear behaviors and mechanisms between interfaces with relative motion or with relative motion trend, which involves the fields of physics, chemistry, material science, etc. There are many factors affecting the interface friction performance, mainly including interface properties, load, service environment and service temperature, etc. Therefore, it is very necessary to employ the computer simulation technologies, such as density functional theory (DFT), molecular dynamics (MD) and finite element analysis (FEA) to investigate the friction, because one need to comprehensively consider various factors to understand friction. Among them, the DFT method is an important mean to understand friction properties and explain friction experimental phenomena. Based on DFT calculations, our research group had proposed a new strategy to realize super lubrication, which was the pressure-induced super lubrication. However, the model construction and data post-processing wasted a lot of time in the process of friction calculations with DFT. Therefore, it was very necessary to propose a high-throughput DFT calculation method which can automatically calculate the tribological properties of solid interface. Although a high-throughput DFT calculation method for solid interface tribological performance had been proposed by Paolo Restuccia et al, this method can only work under zero load. The friction performance in the wide load range can better reflect the actual situation and explain the experimental phenomena. In this study, we built a high-throughput DFT computing method, which can automatically calculate solid interface tribological performance under any loads. This method not only realized the high-throughput calculations, which can calculate more than 1 000 systems in parallel, but also realized automation, which can automatically construct computing models, submit and manage computing tasks, and intelligently analyze computing data and finally output the potential energy surface (PES), friction force and friction coefficient of the calculated interface systems automatically. In this study, the graphene/graphene sliding system was taken as an example to verify the reliability of the method. And the graphene/graphene sliding system friction performances calculated by this automation method agreed with those calculated by traditional method, which indicated that the automation method was reliable. This method can greatly saved the time spent on studying the solid interface tribological performance through the first-principles methods. Although the high-throughput calculation method realized the automation and high-throughput calculation of solid interface tribological properties, there were still some limitations. Firstly, because it was difficult to introduce temperature directly during DFT calculations, the influences of temperature on friction performance were not considered in this method. Secondly, it should be noted that this method did not consider the exchange of atoms between interfaces and interfacial tribo-chemical reactions during friction. In addition, the influence of atmosphere environment was also not considered in this method. These aspects would be further improved in the next step.