Learning Forum

[Oleg Makarov]

I am trying to model uniaxial compression test of cylindrical specimen using SPG method in LS-Dyna R13.1 in Workbench. I use MAT_024 to define yield stress and MAT_ADD_GISSMO with very small strain to failure (LCSDG) to model "brittle" fracture. IN SECTION_SOLID_SPG I set IDAM = "Programm Controlled" and Self-Contact Indicator (ISC) equal to YES (1). So I hoped that destroyed particles will interact with each other even after specimen crushed.

But when the specimen was crushed the two halves started to move towards each other and I see the penetration of the material (see figures). Therefore the particles don't interact with each other.

Could someone explain why is that so and how can I fix this?

Maybe I have to define glocal contact in the whole model?

[Andreas Koutras]

Hi Oleg,

See the following:

"FEM-based single surface contact can be used for SPG parts as long as there is no new surface created due to material failure around self-contact areas. If failure happens everywhere even 'inside' the domain, e.g. in impact and penetration analysis, we currently have two options of self-contact to use:
a. *CONTACT_SPG, ISELF. Support SMP/MPP. More efficient.
b. *SECTION_SOLID_SPG, ISC. Support SMP only. More robust but slower."

I hope this helps.

[Oleg Makarov]

Thank you for your answer!

1. As mentioned before I've already tried to use option b (SECTION_SOLID_SPG, ISC) to include self-contact handling as you can see in the figure below. But it is not working! This is my exact question: why is it so and how can I fix it?
2. How can I include CONTACT_SPG, ISELF if I use ANSYS Workbench?

[Andreas Koutras]

If *CONTACT_SPG is not available through the WB GUI, you can copy the *CONTACT_SPG keyword into a keyword snippet as follows.

[Oleg Makarov]

I added *CONTACT_SPG card to my simulation using the keyword snippet as you advised, but crushed parts are still penetrated when I use penalty factor = 1.0 (see figures below). As I understand the penalty factor is the fraction of material elastic modulus, right? A good picture I can obtain only when I use penalty factor = 1000 but in that case the initial stress inside the SPG body becomes very high. If I set penalty factor = 1e06 my sample crushes at the very beginning of the simulation. Therefore I still don't understand what is an appropriate value for penalty factor and how to adjust it?