Learning Forum

[Oleg Makarov]

Hello! I am trying to use GISSMO model with Smooth Particle Galerkin method in Ansys Workbench 2023R1 and want to clarify some points:

1. Am I correctly understand that the GISSMO model has to be used with a separate plasticity model? In this case the plasticity onset is defined by yield criterion (for example Mises criterion) from the plasticity model (for example MAT_024) and the failure (element or bond deletion) is defined by the curve inside the GISSMO model, right?
2. The GISSMO model uses two different curves -  equivalent plastic strain to localization (e_loc, ECRIT) and equivalent plastic strain to failure (e_f, LCSDG). I am not a good English speaker, so which parameter is bigger e_f or e_loc? As I understand e_f > e_loc, right?
3. What is the "driving quantity" for damage? It is noted in "DTYP" parameter description in Keyword manual? Am I correctly understand that I can use not the plastic strain but any history variable, even from UMAT, to calculate damage (in this case Equivalent plastic e_p in the Damage equation will be replaced with the History variable?)
4. According to Keyword Manual, the LCSDG parameter (e_f curve) can not be equal 0, which means that It has to be defined anyway. Also, the damage is accumulated depending on the ratio of plastic strain to failure strain defined by LCSDG. But in ANSYS workbench I don't see this parameter in the MAT_ADD_GISSMO material model. I can only define ECRIT or DCRIT which are not the criteria for material failure (as I understand). How do LS-Dyna calculate Damage in that case if no e_f (LCSDG) is defined?
5. Do I really need all these parameters or I can just use DTYP=0 and define critical damage inside SECTION_SOLID_SPG card as FS? In that case material failure will occur? I am asking because in Keyword Manual DTYP=0 denoted as "no failure" option. 

Thank you in advance!

[Ram Gopisetti]

Hi, i might answer in few of the points 

1) yes 

2) what do you mean by term "bigger", both parameters are different in terms of defining the plastic stain to initate the failure locally or globally,  i need to know more context on this. 

3) yes, but you have to have the defined history variable to be called into this parameter. 

4) you need to add it from the additional tablular data to define the LCSDG into the MAT_ADD_DAMAGE in the LSDYNA WB.

5) i hope i have answered this in your another discussion on difference between IDAM and MAT_ADD_DAMAGE, use DTYP=1 for failure and is supported in WB LSDYNA mat cards. 

cheers, Ram 

[Oleg Makarov]

2) When I say "bigger" I mean that one type of strain has greater value than another. For example e_failure > e_localization. Is it correct?

3) So I can define this history variable from UMAT, right?