Dynamic fracture in rubber toughened polymers: the influence of the fracture surface roughness

Abstract

Dynamic fracture tests have been performed with rubber toughened polymethylmethacrylate (RT-PMMA) samples. For these kinds of materials the macroscopic crack tip velocity å ≈ 0.6cr is observed to not change during propagation whatever the available dynamic energy release rate. Therefore dynamic fracture energy values GIdc, according to the crack velocity in a classical formalism, are not unique at the branching velocity (approximately 0.6cr). Otherwise the classical formalism considers the amount of created surface during propagation as a flat rectangle (the sample thickness multiplied by the crack length). Nevertheless the RT-PMMA fracture surface roughness are observed to fluctuate as a function of the dynamic energy release rate. The more (respectively less) the dynamic critical energy release rate the rough (respectively smooth) the fracture surface. The real 3D topography of the created surface has to be included in the energy balance to quantify an intrinsic material fracture energy. If not, fracture energy can be significantly underestimated. Using different types of profilometer, the precise amounts of created surfaces for different locations along the fracture were measured both before and after branching at different scales. Since the fracture surface roughness depends on the analysis scale some precautions are requested in the fracture surface analysis. A self-affine geometrical model is introduced using two parameters: the Hurst exponent and the topothesy. The multi-scale description of the fracture surface roughness by a self-affine model is shown to provide a significantly better approximation of the created surface. A new and original geometrical method is introduced to estimate self-affine parameters: the 3D surface scaling method. It is based on the estimate of the amount of created fracture surface using a routine which makes a surface triangulation. Hurst exponents are shown to be unique, χ = 0.6 ± 0.1 for the different fracture zones and measurement scales. It is shown that topothesy ratios indicate a significant difference of fracture surface roughness amplitude depending on the observation resolution when the detrending technique is not correctly introduced. Indeed, the lower the topothesy, the smoother the fracture surface.