Global and local microstructural weak links for spall damage were investigated using 3-D characterization in polycrystalline (PC) and multicrystalline (MC) copper samples, respectively. All samples were shocked via flyer-target plate experiments using a laser drive at low pressures (2–6 GPa). The flyer plates measured approximately 500 μm thick and 8 mm in diameter and the target plates measured approximately 1000 μm thick and 10 mm in diameter. Electron Backscattering Diffraction (EBSD) and optical microscopy were used to determine to presence of voids and relate them to the surrounding microstructure. Statistics on the strength of grain boundaries (GBs) was conducted by analyzing PC samples and collecting the misorientation across GBs with damage present, and it was found that a misorientation range of 25–50° is favorable for damage. Statistics were also taken of copper PC samples that had undergone different heat treatments and it was found that although the 25–50° range is less dominant, it is still favorable for damage nucleation. Removal of initial plastic strain via heat treatments and an increase in Σ3 CSL boundaries, indicative of strong annealing twins, also led to an increased amount of transgranular damage. 3-D X-ray tomography data were used to investigate the shape of the voids present in untreated, as received and heat treated samples. It was found that the as received sample contained a higher amount of “disk”, or, “sheet-like” voids indicative of intergranular damage, whereas the heat treated samples had a higher fraction of spherical shaped voids, indicative of transgranular damage. MC samples were used to study microstructural weak links for spall damage because the overall grain size is much larger than the average void size, making it possible to determine which GBs nucleated damage. Simulations and experimental analysis of damage sites with large volumes indicate that high Taylor factor mismatches with respect to the crystallographic grain GB normal is the primary cause for the nucleation of damage at a GB interface and a low Taylor factor along the shock direction in either grain drives void growth perpendicular to the GB. Cases where experimental results show damage and simulation results show no damage are attributed to the presence of an intrinsic microstructural weak link, such as an incoherent twin boundary.
Modeling of microstructural effects on the creep of hardened cement paste using an experimentally informed lattice model
