0

I'm analysing some dynamic models using Ansys LS-DYNA explicit dynamic models.

My model only has beam elements (BEAM161), concentrated masses (MASS166) and springs/dampers (COMBI165).

The problem is that the length of the springs/dampers is affecting the results, which shouldn't happen: according to the manual, you define COMBI165's stiffness or damping coefficient, which is multiplied by the relative displacement/velocity between the nodes to obtain the reaction force.

For my particular problem, the length of the spring/damper is not conditioning the time-step, so that is not the source of the difference.

AFS Rodrigues
  • 1
  • I'm not sure that I understand what "length of the springs/dampers" means. Are you saying that you have an element connecting two nodes a distance Ly apart, and the value you choose for Ly affects the results? Is it possible to connect one end of the element to a beam node and simply ground the other? – duffymo Jun 24 '15 at 17:38
  • Yes, that's exactly it. I'm modelling a railway: the beam is the rail, the spring/dampers are the track supports. In this case the spring/dampers then connect to masses, who connect to other spring dampers that are then grounded. But I've tried just a beam connected to springs that connect to the ground, and the length still makes a difference. – AFS Rodrigues Jun 25 '15 at 10:07
  • I don't have LS Dyna available to me. How do input spring stiffness? Do you enter a force/length stiffness value? Or is it a rod like specification like modulus*area/length? – duffymo Jun 25 '15 at 12:14
  • I define the material: ET,2,COMBI165, choose the translation option: KEYOPT,2,1,0, define the material as discrete, linear elastic: TB,DISC,2,,,0, define the stiffness: TBDAT,1,Kp, and add an empty real constant list to avoid warnings: R,2,,,,,,, – AFS Rodrigues Jun 25 '15 at 14:17
  • To be clearer: I use Ansys' Linear Elastic Spring Model (it is not exclusive to LS-DYNA). According to the help: "This model provides a translational or rotational elastic spring located between two nodes. Input the spring elastic stiffness with TB,DISCRETE,,,,0 (...) and TBDATA,1,KE (Elastic stiffness (force/displacement) or (moment/rotation))" – AFS Rodrigues Jun 25 '15 at 14:22
  • Yes, this is possible in all FEA codes that I've used. It usually means that you enter a force/unit displacement value in that's added directly to that stiffness matrix entry for that degree of freedom. I don't see what length has to do with it. When you say "it makes a difference", do you mean that it changes the eigenvalues of the system? The dynamic response for the same forcing function? – duffymo Jun 25 '15 at 15:18
  • It changes the dynamic response for the same forcing function, yes. I'm not sure how LS-DYNA assembles the springs into the stiffness matrix, but yes, that would be the idea (except when the spring connects two nodes with active degrees of freedom, but is should still be quite straightforward). The fact that I'm using the Ansys LS-DYNA module creates an additional layer of separation that makes it difficult to identify the problem. – AFS Rodrigues Jun 26 '15 at 12:45
  • I would boil it down to a simple, self-contained example that you can calculate a closed form solution for. Maybe a single beam between two nodes with spring endpoints. Calculate natural frequencies by hand and then with LS Dyna. Send the results to LS Dyna and ask them to comment. I don't have the deep knowledge of their product to help you. – duffymo Jun 26 '15 at 13:12
  • Thanks, I will do that. – AFS Rodrigues Jun 27 '15 at 13:01

0 Answers0