Hi
In my cylinder model, the particles are not filling the annular region. Iam just getting a packed form. I don't know what is wrong with my inputdeck. Iam attaching the input deck. log file and model.
Regards
Hari
| Attachment | Size |
|---|---|
 inputdeck.pdf | 19.94 KB |
| log.liggghts.pdf | 26.53 KB |
| part2.stl_.pdf | 81.35 KB |
cstoltz | Wed, 04/29/2015 - 15:15
Not sure what you mean.
Not sure what you mean. Looks like your particle insertion is offset from the annulus, so you have particles dropping outside the ring. Your m/m/m boundaries may give you problems here since particles are going to fall outside the ring, keep falling forever, and keep expanding your boundaries, which will slow you down. Suggest switching to fixed boundaries. Other than that, it is behaving as it looks from your simulation.
If you are wanting to drop particles to ensure that they will fall into the annular ring, I'd suggest using the insert/stream command and make yourself an annular ring STL file to extrude into the insertion region. Will avoid dropping particles on top of the center point of the ring where you don't want them. Alternatively, make a small cone on top of the center of the annulus so that when particles are dropped, they will be diverted off the center and towards the annular channel.
Finally, you are still using very large values for the modulus which is going to necessitate a very small timestep and a very slow simulation. Suggest reducing this to ~5 MPa for starters, and maybe do a little sensitivity study wrt modulus to see if you need it larger.
Regards,
Chris
dhari27 | Wed, 04/29/2015 - 19:03
Thank you chris for your
Thank you chris for your reply.
Okay i will change the boundaries, but I have a doubt regarding distribution of the different particles. In my command I used insert/rate/region to get more even distribution of all 3 particles. I used insert/stream command earlier with an annular insertion face along with the model and it was found that all the particles fell into the annular region. But using insert/stream did not give a uniform distribution of particles. the annular region seeems to be filled with 99% of smaller particle along with 1% of other 2 ones i.e the bigger ones. So is it possible to get more even distribution of paticles with insert/stream? How should I change my insert command to get equal percentage distribution of all 3 particles?
Regards
Hari
JoshuaP | Thu, 04/30/2015 - 09:29
Hi,
Hi,
is there a reason, that you have to pour it? Otherwise you could use the radii exansion method with fix insert/pack. There you get a nice particle size distribution. Besides you could turn the overlapcheck on, maybe thats the reason for losing particles, but I couldnt really figure out from your input script what the reason is.
regards
joshua
dhari27 | Thu, 04/30/2015 - 10:37
Hi, can you shed some light
Hi, can you shed some light on the radii exansion method with fix insert/pack.
Regards
Hari
JoshuaP | Thu, 04/30/2015 - 11:06
You place the particles with
You place the particles with maybe half of the real radii into the specimen. then you let the radii slowly grow to their original size. For dense packages it works very well. if you want the loosest package, you need to use fill by mass and use a region that is a bit larger than your obtained region. Then you let it settle an you get the loosest package also.
before growing the particles, set all friction values to 0, otherwise they will brace too much. After particles are grown and in steady state you can change the friction values to the real values.
dhari27 | Thu, 04/30/2015 - 12:36
Okay I will try the expansion
Okay I will try the expansion process, but what should I do to avoid the offset from the annulus? After execution the packing is formed at a distance from the annulus region.
JoshuaP | Thu, 04/30/2015 - 16:03
you should insert the
you should insert the particles directly in the observed region