[struct] mpc 184 e cp.

[volaff]

Hello everyone, I have two plates (shell) one on the other but at a distance of 2 mm.

to simulate the fact that, in reality, the plates are a whole, I had thought to bind them rigidly by mating the gdl of the knots overlapping (i.e. same coordinate x,y but z different).

to do this, I read in the ansys guide that the cp command to work properly, must "work" at coincidence knots.
for this reason I thought to create the copy of the nodes of the two plates in the half-way that separates them, and to join the nodes of each plate with the node copied through an mpc 184 element (rigid beam) and to couple the dofs of the nodes coincident by cp command.

However, the following error appears to me:

_
node 217900 of mpc184 rigid link/beam element 499454 is part of a set
of coupling degrees of freedom (cp). this is currently not allowed

someone knows how to solve the thing or, if it exists, a command that pairs the dofs of the knots without having to create rigid elements between the knots!?

thanks to everyone for the attention.

 

[Onda]

Just use a single mpc between the two knots, without affecting the cp.
using it with beam extension that binds knots on six degrees of freedom.

 

[volaff]

thank you so much wave but I didn't understand well, you could be clearer please.

fly

 

[Onda]

connect the single nodes of the two plates with a single mpc184. in that way you rigidly connected the plates between them.
you do not need to make a double mpc and a knot in the center joined by a cp.
but what kind of union do you want to simulate like this? not a bonding, since we have already talked about it for a long time.
Welding?
or do you want to neglect the stiffness of the bonding?

 

[volaff]

I actually wanted to simulate the installation of the frame of the test machine and the panel.

they are so "accepted" that we can consider the two as a "unique piece" avoiding, therefore, the insertion of rivets and lightening the model.

of this panel model has been made a 2d beam analysis which, at present, is better than the one before mine.

simulating the rivet as a beam4 element, for example, the preload of the bolt and the fact that the contact between the plates is not punctual.

I hope it was clear.

thanks for the attention.

fly

 

[Onda]

so if you make a single mpc184 element for pair of knots (of the two plates) it is fine.
In this way the knots are rigidly connected to each other.

 

[volaff]

Thank you very much.

Unfortunately I think the .db file of ansys has corrupted.... I am blaspheming in Turkish.

I hope I'm wrong. You don't know how I could possibly recover it?

 

[Onda]

I don't know.
Don't you make copies of work progress?
I honestly don't remember the structure of ansys files.
you should have a launch file, different from the database, which is created at the time of analysis and which can then be read and imported into a new database, but I don't remember well.

 

[volaff]

Thank you very much.

we made 567890 thousand modifications... And I don't know where the hippo happened.

I have a "good" file but it has to be changed quite "heavy".

I cross my fingers!

Thank you again!

 

[eih64]

instead of the springs put the mpc:
et, 184,0,0