Hello!
I am trying to use servo walls in order to apply a constant confining force on some particles. The force is not so large, only about 2-5N. I am using periodic boundaries for the compressed particles but since I do not my confining mesh in the x and y directions I guess its ok (right?). I should also mention that between mesh and the flowing particles is a single layer of a particle lattice which is rigidly bonded.
The problem is when I start to compress my particles I get very odd force readings (via f_cad3[3]) which fluctuates from very large to negative values
Example:
Step Atoms KinEng CPU cad3[9] cad3[3]
42000 9075 0.00032148222 0 0.068 0
43000 9075 0.00034883258 3.726357 0.068061014 -2.067412
44000 9075 0.00031647133 7.3952589 0.068107596 1.9471037
45000 9075 0.00034295138 11.041145 0.068158262 0
46000 9075 0.00037995199 14.94019 0.068227062 -27.008279
47000 9075 0.0003807783 18.712951 0.068256712 0
48000 9075 0.00030951824 22.419078 0.068277169 93.275607
49000 9075 0.00030917772 26.39754 0.068296661 12.067375
50000 9075 0.00029063274 30.221007 0.068289211 2.4643935
my mesh implementation is as follows:
fix cad3 all mesh/surface/stress/servo file plane_struc2.stl type 1 scale 1 com 0.125 0.05 0.068 ctrlPV force axis 0 0 -1 target_val 3 vel_max 0.2 ratio 0.001 mode auto
I have tried:
1. decreasing my timesteps (min. 1e-5)
2. decreasing mesh resolution (only 4-8 faces)
3. making an unstructured mesh
4. changing vel max and ratio
all of which produce the same erroneous results. I have had some success however when I decrease the flow domain and mesh size to about half that of the example. I would also like to mention that the no particles pass through the mesh and none disappear. My liggghts version is 3.6.0.
I would really appreciate it if someone could at least tell me why my force readings are like this or how I should stabilize mesh.
I'd be glad to provide more information if needed.
Thanks in advance.
kfcui