mesh sensitvity in cracked model (fracture of plate)

[jjamali]

How to do mesh sensitivity study on models with crack?
I am modeling a cracked sample,it is a rectangle with a crack starting in one side and extending to the center point of the rectangle plate. The plate is fixed at one side and stretched at the other side. The crack experinces opening mode I. I am modeling with ls-dyna. The question is I don't know how to find the proper size for my mesh. The area close to the crack have finer mesh. I want to know if the mesh size in my model is fine enough so I make meshes into half and compare the results with previous meshing. The problem is by making crack area elements even finer the material gets softer and the stress values drop dramatically, so the stress values from two different mesh size do not converge . Any thought?
Thank you,
Jam

 

[ninjaz]

Use H-element method . It will automatically reduces the element size at critical area until the solution converges..

 

[jjamali]

Hi,
I have not heard of H elements in LS-DYNA, Do you mean h adaptive solution method? I know solidworks simulation part has that possibility to automatically change the mesh size, but there is no similar option in LSDYNA as far as I know.

 

[ZeroExperience]

I saw this webinar on crack propagation from MSC Software. They may be able to help:

http://www.youtube.com/watch?v=iuSCTTtXDw0

 

[jjamali]

Nice! That's a great video. Thanks,however, It focuses on the failure simulation using MSC software not LS-Dyna and nothing about mesh sensitivity.
Still looking for more information on mesh sensitivity in cracked models.