We present a novel algorithm for generating polymer entanglement networks and apply it to polypropylene/polyamide interfaces strengthened with graft copolymers. Our guides for the generation of the polymer network are the configurational distribution functions derived from a self-consistent mean-field lattice theory of the interface. Entanglement points are placed along the contour of each chain at equal distances, corresponding to the experimentally measured molecular weight between entanglements. The spatial distribution of entanglement points in the direction perpendicular to the surface follows the statistical weights of the mean-field theory. Our initial guess for the positions of the entanglement points of each chain parallel to the interface obeys Gaussian statistics. The polymer network created by this procedure is not in equilibrium. Overstretched strands are relaxed by a Monte Carlo method involving moves that preserve the placement in the directions perpendicular to the interface. The equilibrated network is our starting point for the microscopic simulation of fracture phenomena, caused by the application of tensile stress perpendicular to the interface.