I have been able to use the granular model effectively to simulate monodisperse and binary mixtures of spherical particles. I especially like that I can use read_data to set the initial position/velocity/etc of each individual particle. I now would like to simulate mixtures of particles with more complex shapes - e.g. a mixture of cylindrical particles.
I have tried using multispheres with a few particles "glued" together, but the multisphere approach seems less developed in some respects. It does not seem to be possible to use read_data to set initial conditions, for example. However, I know that I could create very complex shaped particles with the multisphere approach.
Before I continue down this path, I am curious if this is really the best method for me. I cannot find much in the way of documentation or examples but it seems that hybrid granular bond/gran could be another option that might work. There might even be something else that I am completely unaware of.
Thank you for any recommendations.
Ron
jachueire | Fri, 10/23/2020 - 11:14
Simulating cylinders and other shapes
Hello
Take a look at the 'superquadric' particles.
The following links may help you out
https://www.cfdem.com/media/DEM/docu/superquadric_simulations.html
https://link.springer.com/article/10.1007/s40571-016-0131-6
GONCALVES Joao
RonMexico | Thu, 10/29/2020 - 18:12
That paper may eventually be
That paper may eventually be where I am heading. For now, I am sticking with "glued" spherical molecules in a hybrid approach. The molecules are treated as rigid bodies. Thank you for the suggestion.
EmersonSouza | Fri, 06/23/2023 - 17:18
Rigid particles multisphere alternative
Hello RonMexico
How have you managed your problem without the use of multisphere particles?
I'm interested in an alternative from multisphere because with its fix one cannot have heat transfer. Now, I'm using bond/gran, but my particles do not behave as rigid bodies.
Thank you for your attention,
Emerson