Fraturamento coesivo de materiais quase-frágeis usando elementos de interface

Abstract

Crack nucleation and propagation in brittle structures has been widely analyzed through traditional theories of Fracture Mechanics. However, fracture behavior of a certain class of materials dubbed quasi-brittle cannot be reasonably assessed by such theories. Therefore, models capable of describing phenomena inherent to these materials have emerged, e. g. the Cohesive Zone Model. This model can describe the fracture process zone in a simple and adequate way, since it summarize all cracks in a main cohesive crack, embrace phenomena as stress transference between the crack faces during crack propagation and allow for the cohesive crack to develop even without preexisting macrocracks. Using the Finite Element Method, the CZM is applied as cohesive laws inside constitutive models through interface elements. These constitutive models represent the crack behavior until its complete opening, when cohesive tractions between the crack faces become zero. The cohesive laws can be represented by several curves, called softening curves, that connect traction and relative displacement. Among many possible formulations, the linear, trapezoidal and exponential curves are often used. Besides deciding on the softening curve shape, one also needs to decide which fracture modes are going to be considered. This in turn depends on which load types are acting on the crack faces. The possible modes are the tensile normal opening of the crack faces (mode I), in-plane shearing (mode II), out-of-plane shearing (mode III) or a composition of two or more of them (mixed mode). The objective of this research work is to formulate and implement analysis methods for quasibrittle materials based on the cohesive zone method in an academic software based on the Finite Element Method (FEM). In order to provide the framework for such implementations, a zerothickness two-dimensional interface finite element was first incorporated in the code, followed by both linear and exponential initially rigid softening cohesive laws. Both models are capable of dealing with fracture modes I and II, both in isolation and combined. They were applied to renowned examples dealing with quasi-brittle materials in the literature, such as DCB and SEN(B). The models were put to the test through an extensive parametric study, including number of integration points, interfacial strength and fracture toughness, initial stiffness and choice of path-following method. Results lead to the conclusion that the CZM is sensitive and susceptible to changes in both material properties, boundary conditions and loading configuration.