Learning Forum

[lsdyna928]

I have a simulation that involves shell elements and SPH particles. According to the documentation, setting SHLTHK=0 under CONTROL_CONTACT means that the shell thickness is not considered in contact calculations. I used the same setting and ran the simulation. However, the messag file indicates that many SPH particles have initial contact.I do not understand how initial contact can occur if the shell thickness is ignored. I should mention that the coordinates used for geometry/mesh generation were defined such that there was no penetration before starting the simulation. Could someone help me understand why these initial contacts are being detected and how to resolve this issue when the shell thickness is ignored in the contact definition?

Thank you in advance.

[Ram Gopisetti]

Hi, 

Its tough to comment on this aspect with out the cotact type and version of dyna, Even with SHLTHK=0, initial contact can still be detected because:

1. SPH particles have their own radius — contact is checked from the particle’s kernel radius, not just the shell midplane
2. Contact segment offsets may still apply depending on the contact type used
3. The shell midplane itself could be coincident or very close to SPH particle centers, triggering penetration warnings regardless of thickness offset

You can try the following fixes 

• Check SOFT=1 or SOFT=2 penalty formulation for SPH-shell contact
• Use IGNORE=1 or IGNORE=2 in *CONTACT to bypass initial penetration
• Verify SPH particle radius (h value) isn’t overlapping the shell midplane

Cheers, Ram 

[lsdyna928]

Thank you for your comment.The contact type is Automatic Node to surface . Based on the particle distribution, the shell midplane does not coincide with either the SPH particle centers or their radii. However when the shell thickness is considered, the shell overlaps with the particle radius. Since SHLTHK = 0, I assumed that the shell thickness would not affect the contact calculations. Therefore, I positioned the shell elements midplane close to the SPH particle radius.I added a schematic showing the positions of the parts below. Could my assumption be incorrect?

Also, if I use IGNORE as you suggested in the contact definition, it will ignore the initial penetrations. Could this lead to inaccurate results?

[tgerling]

Hi, 

it is recommend to specify the particle contact thickness in the contact definition (Card 3, SST) to capture the particle volume more accurately.

[lsdyna928]

Hi tgerling,

Thank you for your comment. Regarding your suggestion that adding contact thickness would help capture the particle volume, I wanted to ask whether you mean that the contact thickness should be set equal to the particle radius.

 

[tgerling]

Hi lsdyna928,

yes, try setting SST equal or less to the particle radius.