Rigid block contact with balls

Hello,

I am currently working on a project that involves modeling interactions between PFC rigid blocks in the shape of axis aligned boxes with PFC balls (using PFC 6.00.30). The goal of the project is essentially to study the stability of a retaining wall composed of rigid blocks that is supporting a backfill soil (concept illustrated in the following figure).

However, when I am filling the space behind the retaining wall with enough balls such that the wall begins to topple, I start running into stability issues in my model where balls start penetrating into rigid blocks without having the contact being detected. This in turn causes the model to blow up once the ball’s centroid reaches the rigid block’s edge (see following gif).
Stability_Issue

Thinking this may have been caused by some contact detection/creation issues, I proceeded to change the following:

  • Set tolerances for balls and rigid blocks to zero (helped reduce the number of times balls penetrated into rigid blocks)
  • Changed the proximity distance (caused more instability)
  • Reduced the timestep drastically (alleviated but did not eliminate the issue)

To further investigate potential contact issues between balls and rigid blocks, I created a very simple model consisting of two rigid blocks supported by a wall and had a ball resting right on top of the two rigid blocks (see below). Although the model was simple, it was unstable. Trying to run it without setting the tolerances to zero or without using a very small timestep gave an error message which read “The distance is getting larger - bad news”. After setting the tolerances to zero, I was able to get the model to run although it was still unstable.
RBlock_Ball

The code used for the simple model with two rigid blocks and one ball is attached along with the code used for the retaining wall.
2RB_1Ball.txt (929 Bytes)
SIMPLE_PFC_CODE.txt (1.7 KB)

Has anyone run into similar issues with rigid block/ball contacts? Do you have have potential solutions in mind that could help?

Thank you!

Hello @romaric.desbrousses,

It looks like there was a bug in PFC 6.0 that gave, “The distance is getting larger - bad news” error. It seems to have been fixed in PFC 7.0, since you can run the 2RB_1Ball.txt file without issues when you comment out the ball and rblock tolerance commands. In this case the best thing to do would be to upgrade to a newer version of PFC.

Hi!

Thank you for your reply.

I was able to run 2RB_1Ball.txt once the ball and rblock tolerances were set to zero. Not doing so resulted in PFC giving me the error message.

Even then, while the model did run, I still got an unstable behavior where the ball which was placed on top of the contact point between the two rblocks started to roll to the right and then sunk into the rblock (see the last GIF in my first post) which caused the model to explode.

Is that behavior also a result of a bug in PFC 6.0? Or is that an issue specific to ball-rblock contacts?

Upgrading to a newer version of PFC may be an option I’d be willing to explore, but I’d like to make sure the issue would not happen with it.

Yeah, you shouldn’t need to set the tolerances to zero, doing so creates the unstable behavior you describe. You can download PFC 7.0 and since your model is below the demo limits you can test it. https://www.itascacg.com/software/PFC

1 Like