Skip to main content

PARticle Systems: Training on DEM simulation for industrial and scientific applications

Final Report Summary - PARDEM (PARticle Systems: Training on DEM simulation for industrial and scientific applications)

PARticle Systems: Training on DEM Simulation for Industrial and Scientific Applications PARDEM is a FP7 Marie Curie Initial Training Network comprising of four universities and nine industrial companies established to provide high quality training to a group of 16 young researchers, to work in the multidisciplinary field of computational simulation and experimental validation of granular processes (
Granular materials are estimated to constitute over 75% of all raw material feedstock to industry. They also present many challenges for innovation and fundamental science, to solve problems in areas as diverse as natural disasters and industrial material handling which can incur extensive economic losses. The Discrete Element Method (DEM) is a promising computational technique providing both visual and quantitative details of the dynamics of particle assemblies. Although the method is established in academia, immature quantitative prediction capabilities and lack of DEM experts hinder its use as an industrial engineering tool in Europe, and PARDEM sets out to remedy this situation.
The principle objectives of PARDEM were to provide network wide and individual training, to undertake knowledge transfer between academic and various industrial sectors. This was achieved by research at host sites, using secondments between partners and providing multidisciplinary and inter-sectoral training as well as complementary skills training at a range of network events. Within this overall context, each of the sixteen scientific projects that form the PARDEM network also has its own distinct scientific objectives which tie together in four scientific work packages addressing the different areas of Material Characterization, Model Validation Experimentation, Model Implementation and Novel Methodologies for DEM and Best Practice.
The research projects were constructed to fit into one of three distinct researcher profiles covering Industrial R&D Professionals, Software Developers and Fundamental Researchers, providing a range of different experience for the participants and ensuring that at the end of the project, a diverse team of professionals will result.

In terms of training, all objectives set out for PARDEM project have been achieved. A kick-off meeting and seven network events have been held, encompassing a wide range of basic and additional scientific training as well as complementary skills training as set out in the PARDEM contract documents. Since the Mid-Term Review in September 2011, all network events and training have been researcher led, giving additional skill-development opportunities to our researchers. In addition to network-wide training, all our researchers have taken advantage of many opportunities for individual training ranging from targeted scientific subjects to language training, and have developed further skills on-the-job.

All thirteen PARDEM ESRs registered for PhD degrees at one of the partner Universities, and all are either on target for completion or have completed.
Knowledge transfer has taken place within the network by a number of secondments and short visits between partners, as well as externally by dissemination to external participants at network events and by publications at conferences and in journals. The three summer events in 2011, 2012 and 2013 were all carried out in combination with established international conferences at which PARDEM researchers participated in targeted conference sessions. A website was set up early in the project, and remains under constant development to serve the need for both network interaction and external dissemination.
Arrangements have been made to ensure that the website will continue to be maintained for at least the next five years beyond the end date of PARDEM.
All the scientific objectives were achieved and reported under the four work packages.

Work Package 1 has defined a set of reference test solids and performed the material characterization tests to provide test data for DEM simulations. The underlying bulk processes (also termed “process function”) and underlying material properties (also called “product function”) were distinguished. The validation experiments in Work Package 2 provided quantitative measurements for the validation of DEM simulations. The granular processes were chosen based on the industrial relevance and their suitability for validation studies and were also aligned to partners’ expertise and facilities. These include the study of flow dynamics, stress regimes and other phenomena such as segregation in granular heaps, silos, agitated mixers, pneumatic conveying and fluidised beds. Work Package 3 focused on the identification and characterisation of simulation algorithms used for particle-particle and fluid-particle interactions in DEM simulations of validation experiments. The mathematical forms of the particle contact algorithms used and the corresponding “physics” they describe have been sorted in order of increasing complexity and the “fineness” of the length-scale.
The development of novel methods to extract relevant bulk parameters from DEM simulation results has been reported under Work Package 4, which brings together the outcomes of the other scientific work packages to describe and document the best practice validation methodologies.

In summary, PARDEM was successfully completed according to plan and the training and scientific objectives set out in the contract documents were fully achieved. PARDEM’s most significant impacts were the development of a group of young researchers equipped for a variety of leading roles in DEM development, and simulation (across a range of industry and academia) and a set of DEM and experimental results on several granular processes with a detailed understanding of the predictive capability of the technique, allowing it to be used with increased confidence in industrial and scientific applications.