Error: Can't output Elems with invalid ids! #17158
-
|
Hello MOOSE users, My model has Phase Field+ Tensor mechanics kernels and I am using the combined module. What I figured out so far is that it is some how related to the adaptive mesh I have. For example, if I chose to update my mesh every 6 iterations, the error occurs around t = 5e-6 sec. But, if I update every 18 iterations it lasts up to t = 5e-4 sec. I have attached some part of my input file. P.S. My MOOSE version is kinda old! |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 6 replies
-
|
@sapitts Have you seen this before with Tensor Mechanics? |
Beta Was this translation helpful? Give feedback.
-
|
I have not seen this error before, but based on some discussions I've had with @laagesen and @dschwen I believe there is an issue with plasticity stateful materials and mesh adaptivity: we don't currently have the infrastructure to handle moving over stateful properties (e.g. the old plastic state from the previous timestep) to a modified mesh. Daniel or Larry, would one of you be willing to clarify/ expand on this issue please? It could be that you see the difference in the "allowable" simulation time is a function of both the onset of plasticity and how close to that onset the mesh is modified through adaptivity. I suspect that as soon as the simulation goes to look for the old plastic state material property information on a mesh that has been modified by adaptivity. If you delay adaptivity, the simulation should be able to run longer. |
Beta Was this translation helpful? Give feedback.
-
|
You could use initial adaptivity. I'm pretty sure that'll solve your issue here. |
Beta Was this translation helpful? Give feedback.
-
|
Take a look at this input file for example: https://github.com/idaholab/moose/blob/next/test/tests/adaptivity/initial_adapt/initial_adapt.i . But as Stephanie said, incremental formulations in tensor mechanics with adaptivity are still under the works. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks hungry, It helped a lot. Do you know how can I define multiple boxes instead of one? |
Beta Was this translation helpful? Give feedback.
-
|
Combo marker is for that exact purpose: https://mooseframework.inl.gov/source/markers/ComboMarker.html |
Beta Was this translation helpful? Give feedback.
-
|
Thank you , yep found it 10 min ago :) |
Beta Was this translation helpful? Give feedback.
I have not seen this error before, but based on some discussions I've had with @laagesen and @dschwen I believe there is an issue with plasticity stateful materials and mesh adaptivity: we don't currently have the infrastructure to handle moving over stateful properties (e.g. the old plastic state from the previous timestep) to a modified mesh. Daniel or Larry, would one of you be willing to clarify/ expand on this issue please?
It could be that you see the difference in the "allowable" simulation time is a function of both the onset of plasticity and how close to that onset the mesh is modified through adaptivity. I suspect that as soon as the simulation goes to look for the old plastic …