Microstructural effects on fracture behavior of particulate composites: Investigation of toughening mechanisms using optical and boundary element methods

Abstract

Particulate polymer composites are used in a variety of engineering applications. These are generally two phase materials with polymeric phase reinforced by a filler phase to improve overall mechanical, thermal and/or dielectric functionalities. From a mechanical perspective, polymers when filled with stiffer particulates generally show enhanced elastic properties and creep resistance. Achieving similar improvement in failure characteristics has not been consistent due to a lack of thorough understanding of microstructural and loading rate effects. This dissertation addresses a few of these issues by studying effects of filler particle size, filler size distribution and filler-matrix adhesion strength on fracture behavior under quasi-static and dynamic loading conditions.

Glass-filled epoxy composites consisting of solid spherical particles are studied in this research. Spherical particles of mean dia.(D) 7 micrometer to 200 micrometer are used to reinforce epoxy matrix at a constant volume fraction (Vf = 10%) and two different filler-matrix strengths, weak and strong. Optical interferometry in conjunction with high-speed photography is used to quantify crack growth and deformation histories during impact loading. Although elastic characteristics remain unaffected by microstructural variations, significant differences in fracture behaviors are seen. Both weakly and strongly bonded particles in the matrix show higher values of steady-state dynamic fracture toughness, KIss, relative to unfilled material. Filler particle size affects KIss significantly when particles are weakly bonded to the matrix but not when bonded strongly. Weakly bonded fillers result in consistently higher KIss values compared to strongly bonded counterparts. A particle size of 35 micrometer appears to be the optimum at the chosen Vf. The KIss of two inter-mixed particle sizes (each of 5% Vf) is bounded by the KIss values of the composite with corresponding single particle size. Fracture surface micromeasurements show that fracture toughness cannot be correlated with average fracture surface roughness Ra as in neat polymers. Therefore, a model for calculating fracture induced roughness, Raf, a component of Ra representative of the fracture process, is proposed. A linear relationship between macroscopically measured fracture toughness KIss and microstructure dependent quantity Raf/D^0.5 is demonstrated.

Crack front deflection, attraction, twisting and blunting are some of the micromechanisms responsible for the observed fracture characteristics in particle reinforced composites. To gain fundamental insight into the problem, optical and boundary element studies on how a growing crack front interacts with an isolated inclusion or an inclusion-cluster are carried out. A symmetric Galerkin boundary element method is implemented in conjunction with quarter-point crack tip element and maximum tangential stress fracture criteria for simulating crack growth. Both experiments and computations support the observation that weakly bonded inclusions in the matrix attract a propagating crack front while strongly bonded inclusions repel the same. The former increases the crack path tortuosity and hence increases dynamic fracture toughness.