Critical Behavior In Dynamic Tensile Fracture Of High Purity Aluminum

The dynamic tensile failure and fracture in ductile metals is of scientific importance for many engineering projects, about which spallation is one of the typical fracture phenomena concerned. Under dynamic loading, the nucleation, growth and coalescence of microscopic voids inside the specimen will be induced due to the interaction of rarefaction waves from both free-surfaces of the impactor and the specimen, and ultimately the catastrophic fracture occurs. By means of molecular dynamic simulation, Strachan et al. (Phys. Rev. B, 63, 060103, 2001) and Seppala et al. (Phys. Rev. Lett., 93, 245503, 2004 and Phys. Rev. B, 71, 064112, 2005) recently have revealed that there is a critical behavior in the dynamic tensile fracture of ductile metals at the atomic scale, i. e., the catastrophic fracture appears only after the damage accumulated up to a critical amount. Such a concept of critical fracture in dynamic tensile spall has been developed and applied to describe the macro-scale spall experiments by Wang Yonggang (Ph. D thesis, China Academy of Engineering Physics, 2006) and two physical parameters, named as the critical linking damage Dl and the critical fracturing damage Df, are proposed. Dl indicates the critical value of damage for the onset of void coalescence, and Df the critical value for the occurrence of catastrophic fracture. With experimental measurements of free-surface velocity profiles and numerical simulations, Wang Yonggang demonstrated preliminarily that these two critical damage parameters could be regarded as the intrinsic material constants to characterize the dynamic tensile fracture behavior of ductile metals, which are not dependent upon the impact stress and the tensile strain rate.In this thesis, one-dimensional strain impact experiments were performed for the High Purity Aluminum---HPA (99.999%), and its critical behavior in dynamic tensile fracture has been investigated. By building up a set of special experiment system, the measurement of free-surface velocity profile and the "soft-recovery" of the shocked specimen have been achieved at the same time and for the same piece of sample. Characterization of damage in the recovered HPA samples has been quantitatively analyzed with a statistic counting of voids on the cut and polished sections. Based on both of the free-surface velocity measurements and the micro-damage accounting, the damage evolution behavior, especially the critical behavior of HPA in the dynamic tensile fracture has been understood systematically. The main and/or innovative points of the thesis are summarized as follows:1. With an impact velocity varying from 196.9m/s to 317.9m/s and a ratio of flyer/sample thickness of 2:4 (mm) and 3:6 (mm), the free-surface velocity profiles of the shock compressed HPA samples have been measured with Velocity Interferometer System for Any Reflector (VISAR). Based on the vibrating features of the velocity profiles, the damage behavior of HPA has been analyzed. Results indicate that the vibrating amplitude increases with increasing shock stress, and the subsequent reverberations describing the spall become more and more obvious. When the shock stress in the material is very low, the free-surface velocity profile replicates virtually the form of the compression pulse inside the sample. When the impact stress reaches a value of above 1.4GPa, the micro damage would appear, and the free-surface velocity profile changes apparently, showing a series of hash reverberation in the profile. While the impact stress reaches higher above 2.2GPa, a compressive disturbance called a "spall pulse" appears in the free-surface velocity profile, and the subsequent reverberation becomes regular again. The measured spall strength of HPA is much higher than those of commercial pure aluminum reported in many literatures, and it increases with the increasing of tensile strain rate as well.2. A quantitative analysis method for accounting the micro-voids of the shock damaged HPA samples has been established. The size distribution of micro-voids and the damage evolution of the spalled samples have been analyzed. The statistic results demonstrated that both the size and the number of voids reach their maximum value within the tensile zone going to spall. While less amount of void and damage occurred within the rest of neighborhood zones. By defining the product of the tensile stress and the time as a parameter called "Tensile Impulse", the statistic results indicate that an obvious critical behavior for the damage evolution appears with the increasing of "Tensile Impulse". When the "Tensile Impulse" is low, the damage grows slowly with a linear increment. While once the "Tensile Impulse" reaches a critical value, the damage grows rapidly and an increment as a power exponential function is observed. Our preliminary results indicate that the critical value of "Tensile Impulse" for HPA is about 0.34GPa·μs. Such a critical transition behavior has confirmed the molecular dynamic simulation by Strachan et al. (Phys. Rev. B, 63, 060103, 2001) and Seppala et al. (Phys. Rev. Lett., 93, 245503, 2004 and Phys. Rev. B, 71, 064112, 2005), and therefore for the first time it has shown this behavior by the macro-specimen experimentally. This result may also be an important complementary to the traditional void growth law given by Curran, Seaman and Shockey (Physics Reports, 147(5&6), 253-388, 1987). Not alike the present observation, they showed that the void diameter increases with the increasing of "Tensile Impulse" merely by an exponential form, but no linear increment and no critical transition. In terms of these observations, a new understanding on the dynamic tensile spall in ductile metals could be outlined.3. Theoretical analysis and numerical simulation have been completed for evaluating the damage evolution of HPA under dynamic shock loading. Given the deficiency that the model parameters of damage evolution were fitted mainly by one kind of experimental data, e. g., the free-surface velocity profile, in most previous studies, we propose here to constrain the damage evolution by both free-surface velocity data and the damage distribution accounted in the post-shock samples, with which the critical linking damage Dl and the critical fracturing damage Df of HPA have been determined. Combing with the metallographic analysis of the recovered samples, the physical meaning of the critical damage parameters Dl and Df have been discussed. Results indicate that the coalescence mechanism of voids in this material resulted mainly from the direct connecting of voids. On geometry scale, the space between two voids for onset of coalescence is about 0.74 times of void diameter, and the space between voids for occurrence of catastrophic fracture is about 0.01 times of void diameter. These values indicate that HPA has nice ductility and the void distribution in the spallation zone is very denser, and the catastrophic fracture happens only until one void closes nearly to another. Based on the simulation, the universal property of the critical damage parameters Dl and Df also have been discussed. Results indicate that the parameters determined by each shot could be applicable to the rest of other shots. So, the suggestion that the critical damage parameters could be two intrinsic material constants for describing the dynamic tensile fracture has been validated again, and they are independent on the loading conditions.