Supplementary Material to "Unravelling densification during sintering by multiscale modelling of grain motion"

Supplementary material to the paper "Unravelling densification during sintering by multiscale modelling of grain motion" ( formerly named "Investigating densification during sintering with molecular dynamics and phase-field simulations") by Marco Seiz, Henrik Hierl, Britta Nestler. This contains videos of simulation data as well as a simple python implementation for the rigid-body motion model determined within the paper. The initial particle packings employed in the paper are also contained for this, along with a simple reader for usage in the python implementation. Finally, some more detailed analysis of the raw data concerning the coordination number is presented. A binder for quick trialling is available at https://mybinder.org/v2/gh/https%3A%2F%2Fgit.scc.kit.edu%2Fxt5201%2Fsupmat-densification-md-pf/HEAD with the base repository being located at https://git.scc.kit.edu/xt5201/supmat-densification-md-pf/-/tree/master/ . {advmu,md}_blocks.webm: A simulation of 16 grains in a chain calculated with both model ADV-mu and the MD-inspired (MDi) model developed in the paper. Yellow indicates the solid grains, dark indigo the vapour, red lines any interfaces. Model ADV-mu eliminates the pores in a step-wise manner from the outside in, suggesting inhomogeneous densification. Model MDi eliminates them almost at the same time, suggesting homogeneous densification. full.webm: A simulation in a 400^3 voxel simulation box with particles of initial size R=12 cells = 12nm. The mesh visualized here is based on the largest local phase-field excepting the vapour phase-field. This allows introducing an etching-like effect with an appropriate choice of contour level l, with l=0.6 employed here to show grain boundaries and higher order junctions. Macroscopic densification is quite evident here. Note that the early part of the simulation was written out at a higher frequency, so there's an apparent jump in densification rate; it actually decreases monotonically. fracture.webm: Same data as above, but after "fracturing" along a (011) plane. This allows the inner part of the green body to be visualized, showing the polyhedralization of the grain structure. contact-solver.ipynb: Shows how to use the python implementation of the rigid-body motion model solver coded in helpers.py. packs.zip: packs/*dat: The initial position and sizes (uniform) of the particles, can be used as input for the above. coordination-extras.ipynb: Analysis of the raw data of the contact detection for a simulation. data.zip: data/* : Data used in the above notebook.The research was funded by the Deutsche Forschungsgemeinschaft under grant number NE 822/31-1, with most of the simulations being performed on the HAWK supercomputer at the High Performance Computing Center Stuttgart. The parallelization and code optimization were funded via by KNMFi within the programme MSE (P3T1) no. 43.31.01