Restart of Submodel in Abaqus

[LEVanNhat]

Hi everyone,

Currently, i have a problem concerning Restart of submodel in the Abaqus software. I used the submodeling technique for my research. However its analysis is interrupted by a convergence problem. Therefore, i applied the restart technique in Abaqus for continuing this interrupted analysis. And another occurred problem in the submodel restart is that i could not apply exactly the boundary conditions of the driven nodes for the submodel in the interrupted time increment. This causes of the difference between analysis step time of submodel restart and the global analysis step time.For example: - Analysis step of global model: 20s (= step time of submodel)
- Interrupted time in submodeling analysis: 5s at increment 20

My restart file:

--------------------------------
*Heading
*Restart, read, step=1, inc=20, end step
....

*Step, name=Step-2, inc=5000
*Static
0.01, 15, 1e-6, 1.
*Boundary, submodel, step=1
DRIVEN-NODES, 1, 3
*Temperature, amplitude=TEMPERATURE
ALLNODES, 1.
*restart, write, frequency=1,overlay
...
*End Step
-------------------------------------

If i use this restart file, the boundary condition applied to submodel in restart analysis begins at t=0s of global analysis step time but not t=5s (interrupted moment of previous submodeling analysis)

I lost quite a lot of time to solve this problem but not possible. I hope you can give me a solution or suggestion to solve this problem?

Thank you in advance.

Best regards.

 

[StefCon]

Hello,
From my experience, submodels usually converged fast or not at all. Do you happen to get any warning messages or error messages? If so, they could help in figuring out what is wrong (post them here).

Good luck,

 

[LEVanNhat]

Hello,


In my submodel, I changed the constitutive models of material (crystal plasticity model and cohesive zone model) to access the microscopic behavior. In fact, my cohesive zone model is highly sensitive for convergence. Consequently, the numerical computing stop for the sub-model is normal. I started a calculation submodel. Although it runs, the boundary conditions of driven nodes (technique node-based submodeling used) are misapplied. I searched in Abaqus documentation, however it seems not to have a trick for declaring boundary conditions. Maybe I have to use a soubroutine to solve this problem.

Do you have a solution or a suggestion?

Thanks

 

[StefCon]

You seem to have quite an exotic material model there.

What do you mean "not a trick"? The CAE User guide has good instructions for defining this (38.4 Creating the submodel boundary condition). Maybe you are using some other pre-processor?

If the submodel is giving you problems you could consider using substructuring. It might be a more simple option than making a subroutine.

Good luck!

 

[LEVanNhat]

Thanks for your discussion. This problem is not yet solved. Up

 

[StefCon]

Maybe the message file can say what is wrong. Also a picture of the area for the global model mesh and submodel mesh would help to understand what you are doing.

Best regards,