I use LS-Dyna as the solver for my analysis. My model has two parts connecting with some spotwelds. I defined a contact for them, but it doesn't seem to work.
Hi
Contact usually implies nonlinear behavior, i.e. that the system stiffness varies with time as the system goes in/out of contact. Modal analysis usually is a linear system analysis.
Recent development marries the two, http://web-code-aster.org/V2/spip.php?article750&debut_articles=40 However, this, to the best of my knowledge, is unusual.
So without knowing much abour LS-Dyna. You need a steady state situation, perhaps with upstream nonlinear analysis, before you can freeze the system state as locally linear and compute your modes.
Sincerely
Claes
Dear Hoa Lia, dear Claes Richard Fredo, The conventinal modal analysis, uses the superposition principle to put together the effects of different modes. And since superposition is vallid basically for linear systems and contact is a reason of nonlinearity, you can not use ordinary modal analysis to arrive at the responses of problems involved in contact. I would rather also add that there is no change of actual stiffness in your problem; contact causes abrupt changes of the status of the system depending on the response and status of the system, and this is the source and reason of the nonlineaity, causing the inapplicability of ordinary modal superposition. Have a nice day.
Dear Aram
True, superposition requires linearity of some sort.
Locally linear is one way to account for things like linear and non-linear prestressing. If you want to delve deeper, one can split the model and use substructured modes for linear sections and model smaller, non-linear, parts differently, etc.
Did you take a look at the link regarding shock non-linearities? Some progress is being made and the work on shock non-linearities did at least impress me quite a bit.
Sincerely
Claes
LS-Dyna is capable of doing nonlinear analysis in which you should be able to define contact pairs. But modal analysis is essentially based linear systems. Defining them in a linear-based modal analysis seems contradictory. However, some software can do it by "linearizing" the contact features. In that case, the contact is already bonded (meaning no separation any time) and therefore the system matrix is constant. I doubt this approach will serve the purpose of our vibration analysis.
- Badges
- Science topic
- Similar topics
- Filing