Coarse‐Grained Simulation and Machine Learning Investigations on Tribological Property of Graphene/Nitrile Butadiene Rubber Nanocomposites

ABSTRACT The underlying mechanisms of agglomeration and dispersion for graphene nanosheets (GNS) nanofillers within the nitrile butadiene rubber (NBR) composite system have not been systematically investigated. In this study, a coarse‐grained molecular dynamics model of GNS/NBR was initially constructed. The tribological properties of GNS/NBR composites containing graphene at varying loadings and distribution states on a metallic Fe surface were examined through shear friction simulations utilizing the constant strain method. Results indicated that incorporating 3.83 wt% dispersed graphene nanosheets (DGNS) into the matrix resulted in reductions of 50.33% in the average coefficient of friction and 64% in wear rate, compared to agglomerated graphene nanosheets (AGNS)/NBR composites. To elucidate the mechanisms underlying these phenomena, analyses were conducted regarding mean square displacement and interaction energy between the GNS reinforcement and rubber matrix, as well as interactions with metallic Fe, bead motion, interfacial temperature rise, and structural evolution during dynamic friction. Finally, a machine learning data‐driven model was employed to collect mesoscopic friction information for training purposes, enabling predictions of the friction coefficient under the combined influence of external shear action and internal GNS agglomeration. This study provides theoretical insights into enhancing stability and dispersion of GNS within rubber matrices while mitigating wear failure caused by GNS filler agglomeration.