Displacement Jumps and Overclosure Error

[CS54]

Hi, I would like to create a V-notch on my DWTT specimen (Step-1) and break it using rollers and an impactor (Step-2). Everything works well for Step-1. However, at beginning of Step-2, I have a displacement jumps warning and I get an unresolved initial node-analytical rigid surface overclosure(s) error. I use Abaqus Explicit and work with VUMAT.

Could you please help me with this?

 

[FEA way]

Can you paste the contents of those messages here ? Take a look at the locations where overclosures occur, this may give you an idea about what's wrong.

 

[CS54]

I attached the .msg file link below. Again, in the first step I applied the V notch impactor (4mm displacement applied). I have not gotten any errors in that step. When I moved to the second step, I got some jump warnings that were the same length as the displacement applied in the first step. Then, got overclosure warnings and finally got the excessive distortion error. I can give more details about this.

 

[FEA way]

Is there an amplitude controlling that boundary condition ? If not, it should be added. If yes, what does it look like ?

 

[CS54]

Yes, I applied smooth amplitude.

 

[FEA way]

Can you paste the .inp file fragment containing the complete definition of this BC in both steps with amplitude assigned to it ?

So something like this, for example:

*AMPLITUDE, NAME=Amp-1
amplitude_definition
...
*STEP, NAME=Step-1
...
*BOUNDARY, AMPLITUDE=Amp-1
BC_definition
...
*END STEP
*STEP, NAME=Step-2
...
*BOUNDARY
BC_definition
...
*END STEP

Especially the *BOUNDARY keyword for this BC in the second step is important here since that jump occurs at the beginning of the second step. Complete definition of the amplitude is also important.

 

[CS54]

I was able to fix my displacement jump issue. I realized that amplitudes were not continuous when Abaqus goes to the next step. However, I am still getting errors due to overclosure in my last step. This is because I am not having an acceptable element condition on V-Notch impacted surfaces. Elements were excessively distorted and overlap each other at the tip. I tried to use ALE meshing but it is crashing when I apply my desired V notch punch depth. I am not sure how to do this. Is there any other suggestions ?

 

[FEA way]

That’s what I wanted to check in your definition, among the others. Amplitudes can be defined with respect to step time or total time. In your case, the latter approach might be best.

Did you try refining the mesh in that region significantly ? Maybe even using a larger number of tetrahedral elements instead of rather long hexahedrons.

 

[CS54]

Sorry that I could not reply before. Please see my attached file. I use VUMAT for this simulation to run. I was wondering if the contact in STEP-1 is very stiff and it makes the simulation unstable. Would that be possible to have this kind of error in the beginning of the next step ?

Here is the INP file.

https://files.engineering.com/getfi...4-4345-b591-6f8e777e4d1c&file=32_DEPTH_11.rar

 

[FEA way]

It seems that you still should significantly refine the mesh around that wedge support. Deformations there are large and the mesh is not ready for that so some elements distort too much.

 

[CS54]

I think my problem is overclosured nodes that I have. In the first picture here end of my first step, I have some over closured nodes.

These overclosured nodes start moving freely when hit my hammer in my last step.

After that, one of the nodes hits these elements and my simulation crashes.

How do I get rid of these overclosure nodes before my last step ?

Thank you.