An Open Data Pilot for the validation of Discrete Element Models

Granular materials consist of an agglomeration of discrete particles. In nature, they occur frequently, for example as river pebbles, sands or food grains. But also many technical applications deal with granulates, e.g. pills in the pharmaceutical industry. The simulation of granular materials has become increasingly important. For this, the particle based Discrete Element Method (DEM) is a widely used approach to predict e.g. the mechanical behaviour of such materials. The validation of DEM models is an important and challenging task, which will be addressed in the project An Open Data Pilot for the validation of Discrete Element Models. From the considered granular material, many different types of data are necessary, such as real grain shape data, e.g. 3D scans, single-particle tests and measurements of the bulk material behaviour in principal experiments. This data is then used for the simulation. In the next step, the particle shape in the simulation is modelled and material parameters are calibrated to match the simulation results to the experiments. In a new approach, it is envisioned to integrate information of real grain shape into a DEM particle shape representation, which is as simple as possible, leading to fast and robust simulation methods. Also, material modelling, especially particle contact friction, is expected to play a key role in the simulation process and will be focussed on in the project. Scientific papers on DEM model validation, usually do not contain data on the above mentioned aspects in full detail. Thus, for each investigated material, researchers have to go through the whole process of data collection and validation from the beginning. In this project, all data needed for validation, developed algorithms and results will be made fully available for other researchers (and possible other interested parties), in a so-called Open Research Data Pilot. Scientists, who cannot conduct experiments due to financial restrictions, can use the data or the validated DEM model directly for their work. The sole use of freely available open source software in the whole project will make this even easier. Thus, it is expected that sharing data, as well as developed methods and results in full detail, will directly promote research in the field of granular materials.

Granular materials consist of an agglomeration of discrete particles. In nature, they occur frequently, for example as river pebbles, sands or food grains. But also, many technical applications deal with granulates, e.g. pills in the pharmaceutical industry. The simulation of granular materials has become increasingly important. For this, the particle based Discrete Element Method (DEM) is a widely used approach to predict e.g. the mechanical behaviour of such materials. The validation of DEM models is a challenging task, which was addressed in the project "An Open Data Pilot for the validation of Discrete Element Models" at the example of different types of railway ballast. First, the shape of single ballast stones was analysed: using 3D scans so-called shape descriptors were evaluated. Second, a DEM particle shape representation was developed: in a new approach not the particle's actual shape (complex structure with many edges and corners) was approximated but the particle's shape descriptors. This allowed the construction of simple particle shape representations for the DEM simulations, which are highly computational efficient. Third, measurements of the single ballast stones' properties regarding friction and elasticity were carried out. Finally, these results were incorporated into a new calibration methodology for the DEM model ensuring that the simulation results match the experimental results as closely as possible in order to make the models reliably usable for other applications. All data generated in this project was made fully available for other researchers (and possible other interested parties), in a so-called Open Research Data Pilot. Scientists, who cannot conduct experiments due to financial restrictions, can use the data or the validated DEM model directly for their work. The sole use of freely available open-source software in the whole project makes this even easier. Thus, it is expected that sharing data, as well as developed methods and results, will directly promote research in the field of granular materials.