MCE Ph.D. Thesis Seminar

Material heterogeneity at some scale is common in present engineering and structural materials as a means of strength improvement, weight reduction, and performance enhancement in a great many applications such as impact and blast protection, construction, and aerospace. The experimental and computer aided simulations in this thesis seek to establish a scaling relationship between composite microstructure and shock front disruption in terms of particulate size and density through the use of multi-point heterodyne velocity interferometry. A model particulate composite has been developed to mimic the wave reflection properties of materials like Ultra High Performance Composite (UHPC) concrete and polymer bonded explosives, while also being simple to source and manufacture repeatably. Polymethyl Methacrylate (PMMA), a thermoplastic polymer, and silica glass spheres satisfy the manufacturing constraints with a shock impedance mismatch of 4.1, in between the shock impedance of UHPC concretes (~10) and polymer bonded explosives (~2).

Shock rise times are reported for composites of 30% and 40% glass spheres by volume, with glass spheres of 100, 300, 500, 700, and 1000 micron mean diameter that have been sieved. Composites with single mode as well as bi-modal bead diameter distributions are subjected to plate impact loading at an average pressure of 5 GPa. In single mode composites, a linear dependence of shock wave rise time on particle diameter is observed, with a constant of proportionality equal to the bulk shock speed in the material. Bi-mode bead diameter disruption composites were fabricated in order to achieve higher volume fractions without composite degradation. The addition of a second phase to a base 30% glass by volume composite mix results in significant increases in shock wave rise time for base mixes of 500 micron beads, while a point of maximum scattering effectiveness is observed for base mixes of 1000 micron beads. An evaluation of the discrepancies in simulation and experimental results is presented. Shock disruption mechanisms and matrix/interface damage effects are discussed as possible sources of error and potential avenues for model improvement.