An efficient prediction mechanical performance of coating structures has been a constant concern since the dawn of surface engineering. However, predictive models presented by initial research are normally based on traditional solid mechanics, and thus cannot predict coating performance accurately. Also, the high computational costs that originate from the exclusive structure of surface coating systems （a big difference in the order of coating and substrate） are not well addressed by these models. To fill the needs for accurate prediction and low computational costs, a multi-axial continuum damage mechanics （CDM）-based constitutive model is introduced for the investigation of the load bearing capacity and fracture properties of coatings. Material parameters within the proposed constitutive model are determined for a typical coating （TIN） and substrate （Cu） system. An efficient numerical subroutine is developed to implement the determined constitutive model into the commercial FE solver, ABAQUS, through the user-defined subroutine, VUMAT. By changing the geometrical sizes of FE models, a series of computations are carried out to investigate （1） loading features, （2） stress distributions, and （3） failure features of the coating system. The results show that there is a critical displacement corresponding to each FE model size, and only if the applied normal loading displacement is smaller than the critical displacement, a reasonable prediction can be achieved. Finally, a 3D map of the critical displacement is generated to provide guidance for users to determine an FE model with suitable geometrical size for surface coating simulations. This paper presents an effective modelling approach for the prediction of mechanical performance of surface coatings.
