Home   >   Solutions   >   Materials Modeling   >   Projects & Simulations

Multiphysics Simulations Supersonic nozzle flow against a flat plate, showing density (grey) and vorticity (colors). Read More >

Rocstar Simulation Suite Cutaway of a joint slot in the Space Shuttle solid motor showing inhibitor fluid-structure interaction. Read More >

Projects & Simulations

Granular Materials Modeling

The Illinois Rocstar IMSim suite, described here, consists of a variety of packages and submodules for modeling the organization, statistics, and response of heterogeneous granular materials. This page provides examples and further details of some of the uses of IMSim and it’s modules. A standalone subset of IMSim technologies, termed Tomoprop (for “Tomographic Property Prediction”) forms the leading steps in most mesoscale material analyses, and will be described first.



The Tomoprop suite consists of a pipeline of several modules and utilities that combine to facilitate the mesoscale geometry description and prediction of the thermal and mechanical properties of a material. The order of use and relationship between these modules is diagrammed in the figure to the left. Voxel data from tomographic scanning are fitted and binned by Shape3D  or a Rocpack  pack is produced, and then n-point statistics are produced by Stat3D. Prop3D  then produces estimates of thermomechanical properties of the system from the statistics.

Materials are characterized using microcomputed tomography (MicroCT) to obtain a detailed image set, producing raw voxel data.  The MicroCT has a unique, high resolution, three-dimensional imaging ability that allows for analysis and visualization of the internal, fine three-dimensional structures (i.e., microstructures) of intact samples, which is not possible with typical surface analysis tools. These raw data are then processed using a particle segmentation code, either commercial tools, or in-house Illinois Rocstar techniques using the Insight Segmentation and Registration Toolkit (ITK ).

Illinois Rocstar uses the Xradia MicroCT machine at the University of Illinois at Urbana-Champaign Beckman Institute. The Xradia machine repeatedly images the sample while it is rotated through a series of predetermined angles. These two-dimensional projections are processed by the MicroCT driving software to create a three-dimensional reconstruction of the sample. Sample images from tomograpy datasets are shown below for two surrogate systems: (1) polydispersed glass beads in a simple matrix and (2) mustard seeds and rice in a silicone holder. Many other proprietary and sensitive material systems have been imaged and processed as well. Energetic material systems such as the Ammonium Perchlorate (AP) – HTPB propellant below (produced in the University of Illinois Laboratories of Professor Nick Glumac) are typical.


Three surrogate material scans (from left) Rice (green ellipsoids) and mustard seeds (blue spheres); Salt crystals; 44 and 120 micron glass beads in HTPB binder



(left) Processed scan of an AP/HTPB laboratory-produced propellant (courtesy Prof. Nick Glumac, University of Illinois); (right) Scan cutaway of aluminized AP-based propellant from the Space Shuttle booster


Illinois Rocstar is continuing to develop an integrated system for mesoscale modeling and analysis of energetic materials called IMSim (Insensitive Munitions Simulation). The goal of this system is to enable scientists and engineers to perform predictive response simulations for existing and new energetic material designs, allowing for safe assessment of IM scenarios, where we can bridge the gap between mesoscale structure and effects, and the macroscale response that is required. Consider the idealized HMX (spheres)/ HTPB binder system in the figure below. A block of the material is heated on top and bottom and simulated in two-dimensions with our CTM2D chemothermomechanical solver.


Chemothermomechanical response of HMX/HTPB after 40 s: temperature in K (top left), Eulerian strain in m/m (top right), mass fraction of δ-phase (bottom left), and damage (bottom right).

The above figures show the temperature field, the Eulerian strain, the damage parameter, and the mass fraction of the HMX δ-phase after about 40 seconds. The damage parameters for HMX were chosen in accordance with the experimental stress-strain measurements. Note that the experimental data are available for a very small range of strain. At larger strains, our model imparts a plastic-type behavior of the HMX particles, which is a typical response for energetic crystals. We observed that the maximum strain (and consequently the damage) occurs in the matrix between the particles due to the large volumetric expansion during the β → δ phase transformation.


Reaction rate fields, R in kg·s−1·m−3, for sandwich propellant combustion for case 1 (left) and case 2 (right) after 0.006 s.

CTM2D can also perform surface-burning with regressing boundaries and navier-stokes solution of the fluid domain above the burning surface. The above figures show the reaction rate fields (R) for the combustion model. In both the cases, the temperature and the reaction-rate fields can be qualitatively  compared with those presented in the literature.