All software and codes listed below are available through the High Performance Computing Collaboratory (HPC2) supercomputers or individual desktop, depending on the system requirement. The list below includes a few select software related to multiscale materials modeling that aim to characterize structure-property relations at different levels.
The MSU Autonomous Vehicle Simulator, frequently referred to as MAVS, is CAVS’ premier tool for simulating autonomous vehicle activity in realistic digital terrain. With CAVS’ historical emphasis on high-performance computing and simulation, the MAVS software strengthens our exploration into autonomous vehicle technology.
The following software can provide descriptions of the quantum behavior of atoms and molecules (principally the ground state at 0 Kelvin temperature) using either the Density Functional Theory (DFT) or Hatree-Fock theory. Often, these methods are considered first principles (or ab initio) calculation, implying that in principle there is no need for model parameters other than physical constants. With its high prediction capability, the results can be used to supplement experimental study by providing data that is hard to probe experimentally. Due to its high computational complexity, the system size is limited tens to hundreds of atoms. Obtainable properties includes potential energy surface, zero point energy, elastic constants and geometric structure at 0 Kelvin.
Nano/micro scale material modeling can be carried out using the following molecular dynamics codes and tools to ascertain properties at the atomistic scale by solving Newton’s equations of motion. These simulations generally use interatomic potentials, or force fields, developed using properties obtained from both electronic scale calculations and experiments. The simulations then feed these results into higher scale models, such as dislocation dynamics at the microscale, or continuum models at the macroscale. The bulk of existing research at the atomistic scale focuses on informing continuum models for multiscale modeling of metal and polymer material systems. The system limit is millions to billions of atoms and the time is up to hundreds nano seconds, depending on the computational power. Obtainable properties include energy with respect to different conditions such temperature, pressure, defect in material, grain boundary, and free surface.
Based on the constitutive model, the following software provides macroscale information by using subscale information as cause-effect relations by means of the internal state variables. The concerns here are model calibration, model validation and experimental stress-strain curves. Model calibration is related to correlating constitutive model constants with experimental data from homogeneous stress states, like uniaxial compression. Model validation is related to comparing predictive results with experimental results that arise from heterogeneous stress states, like a notch tensile test. Experimental stress-strain curves can include different strain rates, temperatures, and stress states, like compression, tension and torsion.