Learning Forum

[Tonmoy]

Hello everyone!

I am trying to understand the response of a structure on a soil domain by SSI analysis and my model needs to utilize the DRM technique. LS-Dyna just warns by printing some messages similar to the one provided below (from d3hsp):

*** Warning 50134 (MPP+134)
     Tracked node also belongs to reference segments
     connectivity. This should not cause numerical problems.
       tied interface #           = 1
       tracked node #             = 1
       segment #                  = 179
       segment nodes:
      22 463 442 1
 Summary of warning messages for interface # = 1
   number of warning messages                = 4418

In that case the solver reports the following errors in the d3hsp01:

At line 253932730:
 *** Error 30264 (INI+264)
     SSI interface ID: 1 not found in restart file

At line 253932733:
 *** Error 30271 (INI+271)
     # of nodes 4418 in SSI contact interface id 1
     do not match stored # of nodes: 890217136

I want to understand how to troubleshoot these errors and whether they are related to the manner in which the *INTERFACE_SSI was defined. I know about the implementation and theory of effective seismic input that LS-Dyna employs in the background according to Bielak's 1984 article.

Thanks in advance for your valuable insights and time.

 

 

- Tonmoy.

[Ushnish Basu]

The surface for INTERFACE_SSI_STATIC should match that of INTERFACE_SSI. If you use OFFSET in one, use it in the other too.

Also, there is no need to specify a single-element DRM layer - the effective seismic input in LS-DYNA follows Bielak's earlier papers, not the later ones. The only requirement is that the PML should be at least two elements away from the INTERFACE_SSI

Possibly, even in the dynamic case, the meshes on either side are not separated (you should see duplicate nodes on the interface when you highlight it in LSPP)

 

[Tonmoy]

Thank you very much, Dr. Basu, for the insights. Now I can understand what we were doing wrong. I’ll let you know if there are any more questions.

Have a good rest of the day!

 

- Tonmoy.

 

 

[Tonmoy]

Dr. Basu,

I have tried to initialize the gravity on the structure and not the soil, but it fails. So, I want to start this analysis from scratch. Thank you again for your time and help.

 

- Tonmoy.

 

[Ushnish Basu]

The PML boundary has to be placed at least two elements away from the INTERFACE_SSI, where the LOAD_SEISMIC is applied. It's up to you to decide where to put the INTERFACE_SSI - if the foundation pad is not sliding on the soil, you can even place it at the pad-soil interface. Else, you can define the INTERFACE_SSI at a small distance from the pad, so that it encloses any non-linearity near the pad.

[Tonmoy]

Dr. Basu,

Thank you for taking the time to respond to my query. I greatly appreciate your assistance.

Now, the problem is that if I do not include the soil part for initialization in the static phase, the restart transient analysis terminates issuing a warning that the nodes of the soil domain interface have 0 forces.

I want to understand if there is any way to circumvent this issue.

 

- Tonmoy.

 

 

[Ushnish Basu]

Tonmoy,

Please specify only one INTERFACE_SSI_STATIC, and only one INTERFACE_SSI - it is not meaningful to specify more for this problem. 

[Tonmoy]

Thanks a lot Dr. Basu for all your help. That'll solve the issue.