Crack propagation in polymeric composites

Abstract

The velocities of rapidly moving cracks in polymethylmethacrylate, an epoxy resin, a rubber modified epoxy resin, coupled and decoupled glass bead filled epoxies and randomly oriented glass fiber reinforced epoxies were measured with a crack propagation gage that was electrolissecally plated on the surfaces of the materials. When fracture was initiated from a natural crack it was found that the velocity conformed to Mott's equation a ̇ 2 = a ̇ T 2[1− (a 0/a)] , while fracture initiated from a blunted notch resulted in a velocity that conformed to Dulany and Brace's equation a ̇ 2 = a ̇ T 2[1− (a 0/a) 2] . A general energy balance was used to show how one could develop these two equations as bounds to the velocity of catastrophic crack propagation. The terminal crack velocity in the unfilled materials and the glass bead filled materials was a ̇ T = 0.28√E/ρ , where E/ ρ was the modulus to density ratio of the matrix phase at the macroscopic strain rate of the fracture test. The proportionality constant of 0.28 was independent of matrix type, temperature and degree of adhesion. Cracks in the rubber reinforced epoxies always tended to become blunt, resulting in breaking loads that were higher than that expected for materials possessing a natural crack. In addition, the average terminal velocity was less than 0.28√ E/ ρ, indicating the retardation effects of the rubber particles. These facts were used to explain the higher fracture toughness of these composites. Fracture surface roughness was primarily a function of crack extension and breaking stress and was less sensitive to crack velocity. An empirical modification of the Mott energy balance was used to qualitatively explain this behavior.