Wednesday, 24 February 2010

Grinding Mill DEM Modelling

Jochen Franke, of Scanalyse Pty Ltd, Australia, asks:

I am currently exploring different tools for DEM of mills. So far I am looking at Paul Cleary's work at CSIRO in Melbourne and Millsoft (Raj Rajamany at University of Utah / PERI). John Herbst / Metso also have developed the capability but this doesn't seem to be readily available. At a simpler level there is Malcolm Powell's MillTraj, and the Molycop tools. There will be a Positron Emission Particle Tracking (PEPT) workshop during the Comminution '10 conference in Cape Town, but I assume this method is probably still in its infancy?
Is anybody aware of other DEM capabilities anywhere ? Any experiences with them ? Good ? Bad ? Ugly ?

An unrelated question is : it appears that so far really only liner manufacturers have used this kind of tool to market their wares. Is anybody out there on site or in company headquarters using DEM ? If so specifically what for ?

Would be great to hear comments...


  1. eDEM by DEM solutions is another commercial tool - it even couples with CFD tools like fluent. PEPT has been around for 15 years now and we are currently studying mills with real rock, balls and slurry and tracking each of the connstituents. The ICRA workshop preceeding comminution '10 will showcase some of the capability of PEPT.

  2. Itasca's PFC3D has been successfully used to simulate milling, compute torque and required power in complex applications where the geometry is imported from CAD. 3DEC is another Itasca tool that uses polyhedral elements. See examples of both at (

    In practice though, problem set up is only a beginning. Often, one needs to add new physics such as particle break-up, phenomenological models of noise emission, wall damage, spray coating, etc. This is rather easy with PFC's FISH scripting language and the real-time access it provides to all the internal variables.

    1. how could I access this please , Stewart Fernandez - Brisbane Australia

  3. I think comminution and fine grinding is an essential problem in Australia, and that we do not have to rely on
    commertial, overseas programs in the future to investigate these processes.

    I ant to announce that I am submitting a ARC Discovery project to investigate the problem of energetics of comminution and real-time simulation of grinding mills. This a 5 years project for ARC/QEII Fellowship, wich will be in Collaboration with Dr. HongYuang Liu of the University of Tasmania

    The aim of this project is to establish a high-resolution model to investigate the macroscopic fragmentation of coal, ores and geothermal rocks, taking into account the dynamics of their microstructure. The model will be used to investigate how the input energy is converted in heat, noise and production of new fragments. This approach will help to quantify power consumption in energy-intensive activities in fine grinding, ore liberation and geothermal exploitation. The statistical mechanics of grinding will be combined with lattice-based algorithms and the latest advances of computer-games technology. The result will be the fastest multiscale simulator in the world for prototyping, testing and optimisation of grinding machines.

    Fernando Alonso-Marroquin
    School of Civil Engineering
    The University of Sydney

  4. Thanks very much for your feedback, good point about eDEM in particular, I think it is in use at JKMRC, which is a good sign.

    I would still be interested to hear from anybody who actually used a grinding mill DEM :

    what was the specific purpose of using it ?
    Did the DEM address the problem at hand, and provide the answers needed ?
    If so how well ?

    I am interested in the commercial world in particular rather than academia alone. In other words : have mining comanies or sites actually put up money to use DEM to address specific problems and got their money's worth, or is DEM still purely used for fundamental reearch ?

    Jochen Franke
    Chief Technical Officer
    Scanalyse Pty Ltd
    Perth, Australia

  5. Perhaps still a bit in the academic domain, but it might be worth your while to keep an eye on the LIGGGHTS DEM-CFD program, based on the Sandia LAMMPS code.

    It is Open Source, can be used on workstations or super computers (GPU processing under development), and is very fast (based on benchmarks with EDEM). It is a script-based program.

    You can see more about it at:

    Some examples:

    Alexander Polson
    Chief Analyst: Core Dynamics
    Pebble Bed Modular Reactor
    South Africa
    alexander dot polson at

  6. If anybody is interested in more info about LIGGGHTS, please visit our website /

    There are also training classes to be held in June and September.

    More info here:

    Christoph Kloss developer

    1. We are looking to commission a project to evaluate by simulation, the effects of changing lifer heights and weights in a SAG mill, - in this instance

      We would like to know what the resultants effects be if we

      changed the profile
      increased liner weights

      we wanted to visually evaluate the trajectory

      evaluate what would happen if, we

      increased lump feed size – or reduced it
      increased tonnage feed into the mill

      evaluate the throughput through the grates and the effect on increased discharge if we
      changed the design of the pulp discharge system from the standard to a vortex

      we would like to get an evaluation of the

      power draw

      I hope that you may be able to point me in the correct direction,

      I have twice met with Paul Cleary from the CSIRO but I think that we are now looking for a quick , reliable solution, as the CSIRO solution may take a considerable period of time

      I look forward to your response

    2. I'm not an expert on this, but maybe you should talk to John Starkey:


If you have difficulty posting a comment, please email the comment to and I will submit on your behalf