Clustered x polyhedral shape: why is it’s important to consider particle shape

In real life, no particle is a perfect sphere. Instead, we see particles with different shapes, sometimes too complex to describe with simple geometries. It is well known that real shaped particles differ from spheres in packing density, shear dilation and interlocking, linear and rotational modes of transport, and material strength.

But why do most DEM codes use a combination of spheres glued to each other to approximate a particle shape?

The answer is simple – it’s easier.

Representation of particle shape as single polyhedral particle in Rocky DEM, and glued spheres in standard DEM code.
Representation of particle shape as single polyhedral particle in Rocky DEM, and glued spheres in standard DEM code.

Simulations with glued spheres run fast and are easy for a DEM code to handle because:

• The shape can be described with a single variable, so memory consumption by a single particle is low.
• The contact detection is simple and done through a well-known analytical formula. Furthermore, contact takes place at a single point that is easy to determine. Given that contact detection is a difficult and arguably the most time-consuming step of a DEM calculation, running an analysis with spheres provides a big advantage in this regard.

In cases where the shape is not important, like those involving a fine powder bed where particles don’t roll, or rocks where the shape is approximately spheroidal, one can account for non-sphericity with some rolling resistance. But in cases where it’s important to capture the correct rotational behavior of particles, which in turn is driven by moment of inertia and hence the particle shape, getting the accurate particle shape is critical.

But if particle shape is important, are accurate physics captured with glued spheres?

No. The issues with glued spheres are multifold. The major ones are:

1) Accuracy of shape:  Regardless of how small the sphere is, sharp edges cannot be captured. Decreasing the sphere size to capture unique shape features would lead to an increase in particle count and a decrease in time-step, thus slowing down the simulation. This partly defeats the motivation of running with spheres.

For example, in the chip seasoning process at Pepsico, standard DEM codes with glued spheres could not determine the accurate particle shape with any number of spheres. But with Rocky DEM, engineers could model the seasoning and filling process accurately.

2) Artificial friction: Gluing spheres introduces an artificial source of friction, as can be seen from the particle shapes figure above. Also, some shapes may aid rotations and not restrict them, so using rolling resistance would be especially detrimental for these applications.

As an example, capturing tumbling behavior is critical to the pharmaceutical tablet coating process. Typically, the tablet shape assists rolling. If approximated with glued spheres, the artificial surface bumpiness leads to poor contact dynamics. In addition, this approach would require a large number of small spheres to reproduce the accurate shape, which would reduce the time step and increase computational burden. In addition, the calibration process would be more complicated with a number of friction coefficients to be adjusted.

In contrast, when run when the exact polyhedral shape using Rocky DEM, one doesn’t need to input any rolling resistance parameter as the particle shape accurately accounts for the rolling behavior making the calibration straightforward. Naturally, the other issues like higher particle counts and low time steps do not present themselves when running with exact polyhedral shapes.

Watch how Rocky DEM successfully modeled a tablet coating operation at Bristol Meyer Squibb (BMS).

3) Breakage: In Rocky breakage models, mass and volume are conserved. But when a particle made with glued spheres is broken, volume conservation is not guaranteed. Furthermore, the strength and breakage propensity of a shaped particle is quite different from that of a sphere that is accounted for by breakage models in Rocky DEM. With a 3D polyhedral shape, the mass and size of fragments account for stresses at the contact points, and these results aren’t predetermined by the size and number of spheres around the contact points.

The Rocky advantage

As a high-fidelity DEM software, Rocky developers recognized the need to capture accurate particle shapes. Thanks to Rocky’s extra effort in writing very fast and accurate codes for contact detection, force models including those breakage and fluid drag, and memory distribution in hardware architecture, Rocky users can model any shape they wish!

In fact, the shapes can be created from the myriad of generic particle shapes from Rocky’s UI (figure below). And if these don’t work, you can import your custom geometry as a particle and run the case.

Figure of default 3D templates in Rocky and examples of particle shapes using these templates
Default 3D templates in Rocky and examples of particle shapes using these templates.

And with the ability to run large counts of shaped particles with our multi-GPU technology, simulating the system with real particle shapes and size is possible with Rocky.

Finally, Rocky is not restricted to just custom 3D particles. With Rocky, you can also model fibers and shells, both rigid or flexible.

“The highlight of the Rocky software is certainly its speed and realistic particle shapes. For us, being able to simulate the exact number of particles in the exact same size distributions and shape representations that we have in our experimental studies is the primary benefit to using Rocky DEM.”
Preetanshu Pandey, Ph.D.
Principal Scientist at BMS

“One of the advantages of using Rocky DEM is the ability to model more realistic curved/concave shapes, an important feature of many of our products. Using simulation models with simplified (i.e. flat) shapes can lead to omission of important physical interactions between uniquely-shaped materials, such as snack chips, and the processing equipment.”
Chris Koh, Ph.D.
Director and Global R&D Fellow at Pepsico

“Rocky’s non-sphere, full-polygon particle function was for our company the deciding factor for DEM software selection. Additionally, we appreciate that we can use GPUs to process DEM simulations with accurate and high-speed results. And by inputting several simulation files into the Rocky Scheduler tool, we can easily calculate many simulations as a series process, saving us even more time and effort.”
Yoichi Narita
Researcher at NSSMC

If you have an application where particle shape is critical, contact us today.


Saurabh Sarkar

Applications Engineer, Rocky DEM

Dr. Saurabh Sarkar is an Applications Engineer for the Rocky DEM Business Unit. Prior to joining ESSS, Dr. Sarkar worked as an Adjunct Faculty at Rutgers University and an on-site Consultant at Sunovion Pharmaceuticals where he supported drug formulation and process development activities. He obtained his Ph.D. in Pharmaceutics from the University of Connecticut where his focus was understanding and optimization of different pharmaceutical unit operations using DEM and CFD tools in projects with multiple industrial and government collaborators. He is a Senior Member of the AIChE and serves as an expert reviewer for several journals.

Leave a comment


Get Fresh Updates on Email





We'll never share your email address, and you can opt out at any time, we promise.