We use two codes to perform discrete element modeling (DEM) at TotalSIM US. The open source code CFDEM-coupling supported by DCS Computing (http://cfdem.com/) and STAR-CCM+ Version 11.06 developed by CD-Adapco (http://www.cd-adapco.com/products/star-ccm). 

To date, TotalSim has used CFDEM on two projects, A and B below and STAR-CCM+ on one project (B).  We are currently conducting a comparison between CFDEM and STAR-CCM+ predictions for the heat transfer and particle spatial distribution for a single heated air jet fluidizing a bed of cooler particles in a small, well-defined compartment.  The strengths and limitations of these software can be found in the following review file.

  1. Simulation of packing powder catalyst into a microchannel reactor using sonication
  2. Modeling of a large-scale, continuous peanut roaster

You may contact Dave Hesse at tel (001)-(614-506-5939) or email dhesse@totalsim.us if you have detailed questions.

 

A select set of simple animations of LIGGGHTS and CFDEM simulations are described below and available from TotalSim.

This is a simple simulation of uniform size polyurethane balls being instantaneously released from a cylinder onto a flat surface. The characteristic angle that the spill makes with respect to the surface once all the particles come to rest is known as the angle of repose and is used to adjust material parameters such as Young’s modulus, coefficient of friction and coefficient of restitution.  

This test is similar to Test #1 except the particle sizes are randomly sampled from a normal distribution and have the material properties of beans initially contained in a polyurethane cylinder. As the cylinder is lifted the beans spread onto a flat surface composed of galvanized steel.

Three discrete sizes of particles are randomly selected and dropped into a hopper. Note as the particles pass through the hopper the agitation of the mixture leads to the “Brazilian Nut Phenomena” whereby larger particles gradually rise to the top and smaller particles fill in the space at the bottom. Eventually a combination of particle sizes meet in the neck of the funnel leading to plugging and no more particles pass through the hopper.  

A drum containing a collection of particles randomly sampled from a size distribution function spins clockwise about its axis. The particles adhere to the drum due to gravity and cohesive forces and then at a critical elevation roll down along the free surface of particles to the bottom of the drum. The model values for cohesive energy density and coefficient or rolling friction were adjusted to get the bed dimensions that agree with experimental data.  

A simulation of two augers running independently in parallel such as would be common in a grain eleveator used for transporting grain to a processing unit.  

Example of an agitated bed

A single, high-flow, elevated-temperature air jet points directly down into a bed of particles randomly selected in size. The flow displaces a portion of the particles and heats them via convective heat transfer. The particles transfer heat between one-another via conduction at the point of contact.