A fracture mechanics algorithm, used for continuum mechanics simulations of nano-porous particle-polymer composites, is described herein. The model comprises close to a thousand ceramic particles bound together by latex polymer. These packings are generated using probabilistic methods (Monte Carlo). Pore-space arises as a function of particle shape and position coupled to the concentration and distribution of latex. Since the bridges are the weakest links in the solid state continuum, an understanding of failure behaviour is paramount for the design and optimisation of these composites. The objective of this research is to statistically characterise adhesive failure at particle-latex interfaces against cohesive failure within the latex bridges. To achieve this, a novel numerical method was developed. This method solves ordinary differential equations for vectors of force and displacement in layers through the computational packing. The model includes a scheme for non-linear elastic behaviour that evolves into a plastic flow regime. The model moreover incorporates a routine for interfacial failure between particulates and binder. Geometrical features such as solid state anfractuosity, bridge orientation, material fraction and coordination numbers are calculated from the packing output. The number of bridges straining plastically within the packing is lower than those that fracture at the interface. Fracture and failure are both related to the particle-binder coordination number. There is no evidence to suggest that decreasing the contacting sizes of binder at interfaces as well as making them thinner will lead to more plastic failure and decrease fracture. Rather, both plastic failure and fracture increase as a function of decreased contacting sizes and bulk diameters. The residual elastic modulus decreases exponentially as the number of broken connections increases.