Reducing computational time on multi-size particle system

Submitted by alberthappy on Mon, 06/15/2015 - 17:20

Hello all,

I'm simulating a particle system with multiple particle sizes (D [mm]: 0.5, 1.0, 2.0, 3.0, 4.0, 5.0). I'm wondering what kind of settings can reduce computational time? Has anyone done some research on this?

Forums:

cstoltz | Tue, 06/16/2015 - 13:37

Unfortunately, the smallest particle size tends to set the timestep and governs how fast you can run the simulation. About all I can offer in terms of advice is when inserting the particles, depending on the insertion rate, it can sometimes be difficult to create the desired particle size distribution. During a given insertion event, you want to insert the particles in order of size, with largest first. You'll have a better chance of filling small particles in around the large ones, as opposed to scattering a bunch of small particles throughout your insertion volume and then trying to find enough open space to insert a large one. You may need to break out multiple insertion statements for each particle size to make this work properly.

Regards,
Chris

alberthappy | Fri, 06/26/2015 - 16:59

Thank you, Chris. If I remember correctly, only one fix insert/ is allowed in liggghts, is that corrrect?

Fenglei Qi