Microstructure level modelling for properties prediction of WC–Co cemented carbides

Abstract

A random distribution micromodel was developed in order to predict the properties of WC–Co cemented carbides. MATLAB and VC++ were hybrid programmed to extract the necessary information for modelling the microstructure from scanning electron microscopy images. By analysing the distribution regularity of the microstructural parameters, the randomness characterisation of the microstructure of WC–Co cemented carbides was accomplished using probability density function. The parameterised model based on ‘random method’ was developed, in which the random distribution characteristics of microstructural topology parameters, such as average grain diameter, major axis, minor axis, centroid and grain orientation, were considered, with Co volume fraction freely controlled. The representative volume element (RVE) size was determined using the moving window method, with average grain diameter distribution as the evaluation criterion. It was found that the RVE should contain ∼120 grains. The RVE was directly imported in the finite element software ABAQUS, and the finite element simulation of microindentation tests and uniaxial tensile tests were accomplished to calculate the hardness and elastic properties of cemented carbides. The predicted results are in good agreement with the experimental dates, which reveals that the microstructure level finite element model can effectively predict the properties of WC–Co cemented carbides.