Abst ract. Dynamic three point bending fracture tests performed in a modified Hopkinson Bar are used to obtain material fracture properties (such as the fracture-initiation toughness) at high strain rates. This work presents a three-dimensional munerical analysis of the aforementioned tests, performed by the Finite Element Method, implemented in the commercial code ABAQUS. The relationship between the Stres.-; Intensity Factor and the Crack Mouth Opening Displacement was examined and compared with the bidimensional one. The results indicate that the use of 2D plane strain solutions can tmderestimate the fracture properties.
INTRODU CTIONIntegrity of mechanical and structural components subjected to dynamic loading requires to know the material fracture behaviour at high strain rates. Considering mode I, several dynamic fracture parameters may be defined in relation to the crack propagation regime: the dynamic fracture-initiation toughness, K 1 d , represents the value of the Stress Intensity Factor, SIF, at which a crack starts to propagate. Other dynamic fracture parameters are KID (dynamic fracture propagation toughness) and K 1A (crack arrest toughness) . These three material fracture parameters are used in design but the first is of special significance because it rates the effective propagation of a crack within the structural element subjected to impulsive load.In contrast to the determination of static fracture toughness, K1c, the methodology for dynamic fracture initiation toughness, Kid, is not yet standardized. The instrumented Charpy test has been widely used to evaluate the dynamic fracture properties of materials, but the maximum loading rate (Stress Intensity Factor loading rate, K 1 ) achieved during the test is about K 1 = 10 5 M Pa,jmf s.
Descriptions have been published [1 , 2, 3, 4] of special arrangements of the Split Hopkinson PressureBar (SH PB) for dynamic bending tests at higher strain rates. This Hopkinson tests permit higher strain rates than those reached by instrumented Charpy bnpact tests. The system consists of a striker bar (called projectile or impactor), an inpu t pressure bar with a modified shape edge, a supporting device and the recording equipment. The cracked specimen is placed between the input bar and the supporting device and is loaded to fracture by means of a concentrated transverse force applied at its midspan. The projectile, moving at velocity Vo, strikes the input bar, generating a longitudinal strain compressive pulse, c:;(t) , that propagates along the bar. This pulse can be recorded by strain gages on its outer surface. Once the pulse reaches the right edge of the bar, part of its energy is directly transmitted to the specimen and to the supporting device, while t he remaining energy is reflected back to the input bar, but now, as a tensile pulse, C:r(t). The reflected pulse is recorded by the strain gages. Assuming the theory of onedimensional elastic wave propagation, the load exerted on the specimen, P;(t) , and t he displacement of the edge of the bar initia...