Rigid body motion error when restarting a terminated FSI simulation

[lingdeer]

Hi,

I am running 2 way FSI on ANSYS CFX, simulating blood flow in deformable vessels. My simulation went well for 3 cycles until at the end of the 3 cycles it is terminated with the following error:

License server will time out in 16 minutes if not reconnected before.
ANSYSLI_SERVERS:
FLEXlm Servers:

WARNING:
License server lost for 50 minutes.
Please save your work.
License server will time out in 10 minutes if not reconnected before.
ANSYSLI_SERVERS:
FLEXlm Servers:

WARNING:
License server lost for 56 minutes.
Please save your work.
License server will time out in 4 minutes if not reconnected before.
ANSYSLI_SERVERS:
FLEXlm Servers:

ERROR:
Feature ANS_SOLVER timed out.
ANSYSLI_SERVERS:
FLEXlm Servers:
Exiting...


I has been using time step of 0.01s and each cycle is 1s. This error showed up when the 2.99s time step is done and working on the 3s time step.
Since the simulation is terminated, no db file is written.
I tried to restart the simulation with the .rdb file and continue from the 299_full.trn files written, using procedures described in the ansys manual, and I have been successful in doing that so I don't think it's the matter of my restart procedure.
When I start restart my simulation at 2.99s, once it start solving 3s, the following error showed up:

*** ERROR *** CP = 57.855 TIME= 17:40:21
The value of UZ at node 6247 is 150158011. It is greater than the
current limit of 1000000. This generally indicates rigid body motion
as a result of an unconstrained model. Verify that your model is
properly constrained.

*** ERROR *** CP = 57.855 TIME= 17:40:21
*** MESSAGE CONTINUATION ---- DIAGNOSTIC INFORMATION ***
Rigid body motion can also occur when net section yielding has
occurred resulting in large displacements for small increments of load
or when buckling has occurred. You can plot the time history curve
for node 6247 in the UZ direction to check for stiffness (slope of the
curve) approaching zero.


And there are more nodes with similar error. Since the simulation ran well for the past 3 cycles, I don't think the model was not probably constrained. Also, time step 2.99s is completed so I m not starting from a partially solved/non-converged point.

I had similar problem before, the way I got around it is to restart the simulation at earlier time step, and as long the license goes well, I don't see the simulation stop at the same time point it stopped before. However, in the current analysis, it has ran for 3 cycles and I do not want to waste the time I spent on these 300 time steps.

Does anyone have suggestions on that? I don't know much about ANSYS, so would definitely appreciate if somebody can give me detailed step on how to work that out.

Thanks in advance!

 

[nizzle1234]

Ridged body motion error is always an constraint issue. Probably not what you wanted to hear, but that is what it is. Hope this narrows down your bug search.