Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
contact modeling and time dependent analysis
Posted 14.04.2013, 15:03 GMT-4 Structural Mechanics Version 4.3a 4 Replies
Please login with a confirmed email address before reporting spam
I am trying to model a time-dependent vibration problem (resonant vibration) with contact and I am having trouble solving the model. The contact is between the shaker and the object. I want to apply a prescribed displacement at the tip of the shaker and also to avoid the shaker penetrating the object. I saw a couple of posts discussing about contact in transient problems in the forum.
www.comsol.com/community/forums/general/thread/71/
www.comsol.com/community/forums/general/thread/30262/
where the advice was primarily to avoid augmented lagrangian method and solve using pure penalty method. I am running 4.3a and these settings seem obsolete now? The comsol structural mechanics manual suggests that "as long as inertial effects are negligible, it is possible to apply contact for time-dependent problems". In my case, I guess inertial effects are indeed important since it is a resonant excitation.
Can someone suggest how I can approach my situation? Also, when I modeled the same set up for a stationary analysis with constant prescribed displacement, the model works fine. It doesnt work for time-dependent analysis.
Thanks
Jyani
Please login with a confirmed email address before reporting spam
contact problems are highly non linear and long to solve, are you sure you "need" the contact ? isn't it enough to do a standard continuity and monitor the forces at the interface, and then decide if the part might separate or not ?
Normally one are bolting parts so hard to the shaker that they do not get free by the load case, because else it's impossible to interpret the measured data, or ?
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Thank you for your response. The shaker sits on the center of the bar and is not bolted to the bar but rather held in place with a pneumatic preload. the bar is clamped on both ends. So, the over all loading is sinusoidal vibrations with static preload. So yes, as long as there is sufficient preload, the contact doesn't break.
By continuity, do you mean apply an identity pair at the interface and set the prescribed displacement BC to the shaker? or is there an explicit continuity condition that we can specify in comsol?
What I originally had was that I made a union (In the finalize step instead of assembly) of the shaker and the bar and applied a prescribed domain displacement to the shaker. it seemed to work, but that model didnt seem practical so I wanted to attempt a contact model.
Jyani
Please login with a confirmed email address before reporting spam
I would rather go for your first approach, as adding contact is complex w.r.t. solving, but should also work, so its possible.
I would say, either the shaker is resting on the bar with a preload or you "just" apply a weight and a force/pressure
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
will try that.