Discussion Forum

Hi Stach,

Currently there are two built-in approaches in which you can model a spring with predeformation between two points. Apart from these built-in functionalities, its always possible to write your spring equations using nodes like Weak Contribution.

First, you can use the Spring-Damper node in the Multibody Dynamics interface. Note that if you are using any other physics interface e.g. Solid Mechanics, then also you can use this node by defining the connection points as Attachment nodes in order to pick in the Spring-Damper node.

Second, you can use the Spring-Damper material model in the Truss interface. Note that this method requires a physical edge where spring is to be modeled and can't be used to connect two points without a line joining the two.

There are several requests for the Spring-damper node in other structural interfaces e.g. Solid Mechanics and we are working on that.

Feel free to contact support@comsol.com for more details.

Best regards,
Pawan Soami

Hi,

Thank you very much for your quick response. My model is indeed in solid mechanics interface and I want to model the spring damper in that interface. Like you said in your first solution you said that you can use the Spring-damper node from the Multibody dynamics in the solid mechanics interface by using an attachment. However when I do that I have a Solid Mechanics and a Multibody dynamics but I cant run both (multiple moving frame specifications). How do I use the attachment made in Solid Mechanics without running both physics (solid mechanics and multibody dynamics). (Btw I am also running it enabling geometric non-linearity).

Hope to have a quick response,

Stach van Leeuwen Precision and microsystem engineering TU Delft

Hi Stach,

There could be many ways to resolve multiple moving frame issue however the change suitable in this case is to remove any domains selected in the Multibody Dynamics. The global nodes like spring-damper, joints etc. can be (should be) used in the Multibody Dynamics without having any selection in the physics in the case when attachments are defined in some other physics. Hope it helps.

Best regards,
Pawan Soami

Hi Pawan,

Thanks for your usefull tips. I figured out a working model with a predeformed spring in my porous structure. Still I have one problem and that has to do with the non linear convergence of the solution. When I add more springs in my model there is a point where the stationary solver wont converge to a solution (already at the first step). I use fully coupled stationary solver with Automatic highly non-linear as the non-linear method. Do you have any tips or suggestions on how to improve the nonlinearity in my model so that the stationary solver will find a solution meaning it converges?

Hope to hear from you soon,

Stach van Leeuwen TU Delft