ICF10B

NUMERICAL PREDICTIONS OF DUCTILE FAILURE Damage Predictions According to Rice and Tracey Numerical simulations have been performed using the flow stress of the material, which has been determined experimentally, [2]. The flow stress has been defined as a function of the plastic strain and strain rate. Three-dimensional models have been used to model 1/8 part of each notched specimen, taking into account symmetry conditions. Also a smooth round tensile specimen of 5 mm diameter has been simulated. A preliminary analysis showed that the influence of mass inertia effects is negligible and therefore no dynamic effects have been incorporated into the analysis. A standard plasticity algorithm has been used in the finite element calculations, together with a von Mises yield criterion and isotropic strain hardening. The Rice and Tracey model for ductile damage evolution has been incorporated. The model describes the growth of a void in a plastically deforming material. Assuming an initial void radius R0 the growth has been given by: ln . exp( / ) R R ART h e p p 0 0283 d F HG I KJ= ⋅ ⋅ z σ σ ε ε (2) in which ART is a constant equal to 3/2, σh the hydrostatic stress and σe the equivalent von Mises stress. Some results of finite element calculations at low strain rates will be discussed first. Figure 4 shows local values of the relative hydrostatic stress (triaxiality) σh / σe . These values have been evaluated at the position of maximum damage, for the R = 1 mm and R = 0.45 mm radius specimen this is at the centre of the notch tip and for the R = 5 mm and smooth specimen at the centre of the minimum cross section. For the specimen with radius R = 1 mm is not quite clear if the initiation position is at the centre of the specimen or at the notch tip. This will be investigated experimentally in the future. As a criterion to stop the numerical simulations the failure strain according to Eqn. 1 has been used. Numerically, the strain from area reduction at the notch equals the experimental fracture strain when the calculation stops. 0 0.4 0.8 1.2 0 0.5 1 1.5 True Strain [-] Triaxiality [-] R = 5 mm smooth R = 0.4 mm R = 1 mm 0 0.4 0.8 1.2 0 0.5 1 1.5 2 Plastic Equivalent Strain [-] Triaxiality [-] R = 5 mm smooth R = 0.4 mm R = 1 mm Figure 4: Triaxialities for several geometries at low strain rates. Notched geometries have been indicated by specifying the notch radius R, round tensile specimen is indicated by “smooth”. Left figure (a) uses strain from area reduction, right figure (b) uses local plastic strain. The calculated damage evolution according to Eqn. 2 has been shown in Figure 5 at identical positions in the specimen compared to the results shown in Figure 4. Taking into account the relation for ductile damage, Eqn. 2, together with the calculated hydrostatic stress - plastic strain history, Figure 4, one can conclude that no accordance can be achieved for the smooth and notched R = 5 mm specimen. This applies also for the R = 1 mm and R = 0.45 mm notched specimen. This conclusion becomes clear from Figure 5.

RkJQdWJsaXNoZXIy MjM0NDE=