Discussion Forum

Dear Adrián,

You can couple these two physics in a simple way by making sure that the truss and the solid physics both use the same dependent variable. This way they are coupled.

An other option would be to set a prescribed displacement and use the u and v from the other physics if you do not want them to have the same dependent variables.

Regards,

Thanks a lot!,

I will try it next week :)

Hi guys,

I just tried that thing with an extremely simple structure (one truss and a bi-dimensional "L", and it seems to work.

See attachments:

I used "Pinned" at the left point of the truss, and a prescribed displacement at the right, and it seems to work well (please, confirm it ;-) )

Now, the real thing I want to apply is not working, but I guess I am really close.

In my real case, that "prescribed displacement" must come from another model.

I have a variable called "up" that is calculated in the PDE. It is one of its solutions. Now I want to use that "up", because that value is the actual displacement that the truss makes on the structure :).

I attach also something called "GoodL.png". You will understand the problem there. What I did to use my variable "up" from the solution of the PDE on model 1, was just to define, in model 2 (the 2D structure) that my prescribed displacement is "mod1.up". I would say this is the way to do it.

I hope you can understand the problem and you can help me again.

Thanks a lot in advance for any help you can give me!

edit: could it has something to do with that I am putting everything together in the same Study and in the same Time Dependent Solver?

I also must add that the solution of the PDE comes from Time Dependent solver. So... I actually need to apply a time dependent variable as an input to my second simulation (i.e. a profile of displacement vs. time)

Dear Adrián,

the problem is not completely clear to me, you mean that mod1.up is not used in the displacement?
Please try to use g.up, g is from the pde form. As this is in the same model, mod1 is not needed if I am correct.
If that is not successful, you might want to upload the model.

Best regards,
Frank

Hi, thanks for the reply.

It does not work.

Where you see truss -> Prescribed displacement is where I should use the displacement that I get from the PDE, "up"

If I write "g.up, or mod1.g.up" in the model2, (see other attachment), whatever I write there appears in orange and it says "unknown variable..."

Could it maybe have something to do with that I have only one study?

Thanks

Dear Adrián,

probably your mod1.up is known over the domain (line in model1), therefore comsol doesn't know which value to take for your prescribed displacement.
So if you create a model coupling, eg. a maximum operator over your domains in model 1 and then call for "mod1.maxop1(up)" in model 2. That should do something.

Best regards,
Frank

This does not work... :(, I will keep on trying something.

Thanks a lot for all your replies, were VERY useful!

Another question, please. How can I define a contact between the truss and the 2D structure?

I can define the contact on Definitions, but when I open a "contact" in the Solid Physics, it says the part of the contact that corresponds to the truss appears as "not applicable".

Do you know something about that?

thanks

Does anyone has an idea about this of the contact?? (last post)

Thanks a lot!

Hi,

Contact analysis in COMSOL is currently only implemented between solid boundaries.

You could try implementing something approximate by yourself. You could for example use forces with a very strong nonlinear dependence on the distance between the objects. Such a formulation will inevitably introduce some over-closure between the objects, as well as unphysical forces before they go into contact. With properly chosen values, the solution will still probably be within acceptable engineering accuracy.

Regards,
Henrik

Thanks a lot, I will work on that. In case of problems I will post here again,
Thanks!