V&V/UQ and Experimental Physics
The purpose of the V&V/UQ program is to provide a platform for computational model verification, validation, and propagation of uncertainties through the computational framework. Our emphases will be on quantifying the predictive ability of our multiscale simulations through deterministic verification strategies, coupled with stochastic and physically-based statistical techniques, for validation and UQ. The key component in our validation plan is a series of carefully co-designed simulations and experiments (with quantified uncertainties) to enable meaningful and rigorous comparison of simulations with experiments. Our computations will be data-driven, with morphological information and calibration data provided to our micro-continuum model, leading to a straightforward means for statistical validation.
Material Characterization and Calibration Experiments
Characterization of a virgin material composition will be performed in order to construct an initial computational domain for the data-driven simulations. Results from the imaging studies will be used in conjunction with multi-point statistical reconstruction techniques to define a Representative Unit Cell (RUC) that has the same statistics as the original material microstructure. When performing the reconstruction, we will focus on the type, shape, and size of particles/crystals, and build the multi-dimensional probability density function (PDF) of various geometric metrics (morphological pointers). We will focus on the n-point volume and surface correlation functions, since accurate modeling of surface physics and chemistry is essential.
A Planetary Ball Mill is used for high-energy ball milling (HEBM) of the powder. The rotational speed of the mill is adjustable. A special locking device and jar lid allows one to evacuate the vial and perform grinding operations in an inert atmosphere. During HEBM the particles of the mixture are subjected to mechanical impacts with the balls and the wall of the jar with a force sufficient for breakdown of brittle and plastic deformation of ductile components. Brittle particles are milled to finer particles, whereas plastic components (usually, metals) are subjected to multiple flattening procedures, forming layered composites with the layer thickness decreasing as the milling time increases. Small fragments of brittle components are often found inside the particles of plastic reagents. HEBM not only decreases the particle size of reagents, but also increases their contact area between the reagents. The mechanically induced contact surfaces are cleaned from oxide films and the defects of the crystalline structure are accumulated on theses boundaries. It has been demonstrated that HEBM can be used to obtain nano-structural reactive composites with the size of structural components (phases) of the order of 10–100 nm. This type of treatment leads to significant changes in the properties of reactive mixtures, for instance, to a significant decrease in the self-ignition temperature by hundreds of degrees.
Asay Shear Impact
Reaction initiation of pressed samples by mechanical impact is performed through combined compression/shear tests, used previously with explosives at Los Alamos. The monolithic samples are placed in a windowed sample holder, as shown in the figure. A projectile accelerated by a light gas gun travels toward the sample holder and impacts a plunger supported in the sample holder. The light gas gun used in these experiments propels a projectile to over 1000 m/s. The plunger then impinges the sample, producing strain in the material. The impacting face of the plunger can use different geometries ranging from a flat face the width of the sample, producing pure compressive strains, to a mixture of shear and compression with a radii-shape plunger to nearly pure shear with a narrow plunger.
Reverse Taylor Impact
A Taylor impact experiment is where a material of interest is launched into a fixed wall. A reverse Taylor experiment is where a sample is fixed, and is impacted with a flyer plate. We plan to adopt and modify these types of experiments to provide data for shock assisted combustion synthesis.
University of Notre Dame,
Lead Team Member
University of Notre Dame,
Professor, Purdue University,