Effect of Sediment Form and Form Distribution on Porosity

Data to accompany the journal article "Effect of Sediment Form and Form Distribution on Porosity: A Simulation Study Based on the Discrete Element Method" by Rettinger et al. (2022). The data set contains simulation input and results obtained by the particle packing studies using the discrete element method, as described in the article. The folder `surface_meshes` features 63 processed scans of gravel sediments from the Rhine river in Germany. The original scans have been converted to a convex surface mesh, and then simplified to feature at most 300 surface triangles. Additionally, they were oriented along their principle axes of inertia and scaled such that their size, according to the applied definition, is 1. The folder `single_form` contains the simulation results of the single shape/form studies, where all particles featured a single shape, given by one of the meshes. Three different size distributions have been used: U1 is a unimodal one with a single size class, U7 is also unimodal but with seven size classes, B50 is a bimodal one with approximately half fine and half coarse particles. Each corresponding csv file contains the obtained porosity (`porosity`) of the actual packing and of the packing using equivalent ellipsoids (`porosityEquivEllip`). Furthermore, the considered form factors according to their respective definition, as well as their inverse (`invX`), are given. See the article for the definition and the meaning of the abbreviation. The folder `form_distribution` contains the simulation results of the form distribution study, where the particle form was prescribed by flatness and elongation which were sampled from truncated normal distributions. Again two different size distributions have been used: U1 and U7. Each corresponding csv file contains the obtained porosity (`porosity`) and statistical information about the form factors, given as number-based mean (`mean`) and standard deviation (`std`). In addition to the inverse of each form factor (`invX`), the inverse of each statistical quantity has been evaluated (`1/(X)`) and is contained as column.