Learning Forum

[eceron]

Hello, I am running a cold forging simulation in LS-DYNA using the implicit solver, and I am experiencing a sudden solver termination before the end time.
The workpiece is modeled using the EFG (Element-Free Galerkin) formulation (elform 42). At a certain stage of the forming stroke the solver stops abruptly due to negative volume error. I would like advice on which aspects of my model setup should be improved or reconsidered in order to solve the issue. I can post here some of the keywords I use:  

[Pedram Samadian]

Hi,

Do you have tetrahedral elements in your model? Are you doing any mass scaling in your model? How is the contact pressure/force at the interface? Did you scale the contact stiffness in the contact card? What solver are you using? Did you try the latest released solvers available at the ftp website?

I look forward to your response.

Thanks,

Pedram

[eceron]

Hello Pedram

yes I use tetrahedral elements in my solid mesh. For the tools (rigid body) I use shell elements. I am not using mass scaling. I have tried different contact stifnesses with no luck (I have though not checked the contact pressure. You are right I should do that). I have tried the latestet solver release 16.2

[Pedram Samadian]

Hi,

Thanks for the information. Do you see any weird response at the contact interface? What type of contact are you using? Have you tried *CONTACT_FORMING_SURFACE_TO_SURFACE_MORTAR?

Moreover, Are there any specific reasons you set IAT1-3 to 1 in *CONTROL_REMESHING_EFG? These are higher than the recommended values in the manual. Lower values may help prevent mesh distortion and negative element volume error.

Please also share a screenshot of your SECTION_SOLID_EFG card.

Thanks,

Pedram

[eceron]

At the contact between the tool and the billet I see some distorted elements, which do not remesh. Here is a picture

Here a screen shot of the contact card.

I also tried the MORTAR option with no luck.

and here the section card

Regarding the IAT values, I found an old example from LSTC where they used IAT=1. I will try with lower values. Thanks.

[Dennis Chen]

There's not enough information here to troubleshoot just from those keywords... you can turn on implicit to explicit switch and change the d3plot to allow a much shorter dt around the time the error occured.    This may allow you to visualize the issues of where this happened.  

implicit to explicit switch can be activated in in control_implicit_general with imflag = -lcid or =4 with control_implicit_auto.   database_binary_d3plot can be done with a load curve to control output frequency.  

hope this helps in some way. 

 

[eceron]

Hello Dennis

thanks for the advice. By the way, I saw your video on EFG. Very nice.