Pre-stressed modal analysis: Adding additional supports in between structural and modal solves

[MikaelME]

Dear all.

I am running a pre-stressed modal analysis. A stator for a rotating machine consist of several modules tied together with a tie-rod, which also works as a guide.

When assembling, the stator will only be clamped in one end, tied together, and then clamped in the other end. When running the analysis, i would thus like to use a fixed support first for the structural analysis, and then before the modal analysis, add a fixed support to the opposing end.

Is it possible to add the additional fixed support after the structural analysis for use in the modal analysis?

Thanks in advance, and best regards,
Mikael Mosbech Eronen

 

[Erik Panos Kostson]

Yes it should be.

If you are using WB then you need though to add command snippet for the restrain that will be done with the D apdl command.

Med venlige hilsen

EPK

 

[FEA way]

According to the Workbench documentation it can’t be done but, as it often happens, GUI limitations may be omitted by editing the code manually. Before you try with you complicated model I suggest simple test on a beam for example.

 

[MikaelME]

Thanks Erik.

I was hoping there would be a more WB friendly way of doing it, but i guess i will look into how to use the APDL command within WB.

Best regards,
Mikael

 

[Erik Panos Kostson]

No worries. For the command snippet first create a nodal named selection (called say FIXNODES) of the nodes and then use the D command, say:

CMSEL,S,FIXNODES ! This selects the nodes
D,ALL,UX,0 ! Fix X for sel. nodes
D,ALL,UY,0
D,ALL,UZ,0

ALLSEL,ALL ! Selects all nodes again

From an Fea point of view it is fine, since the static analysis generates the geometric stiffness matrix which is added to the standard global matrix and then the BC are just changed (so dof will be condensed out form the global matrix). Try it out on a simple model.

@ FEA way - please if you are not really sure about this do not provide help (unless you have a long exp. in Ansys which I cannot see in this forum) - because if advice is not correct things can go wrong!

 

[FEA way]

@Erik Panos Kostson I just said that in Workbench it can’t be done without code modifications and to try with simple example first. Nothing incorrect or uncertain here since you basically confirmed that some code lines must be added and provided them to the OP.

 

[Erik Panos Kostson]

Sure, but in general I would advice to be careful with advice (including myself, when I am not sure I will write it out and advice to seek further advice say from a person that knows the subject), and a lot of referencing, especially when not sure.

 

[FEA way]

Yoy’re right. That’s why I always try to give advices only to the safe extent not to confuse the reader. And in case of FEA software I always recommend reading the appropriate documentation chapters since they are usually very comprehensive and helpful. At least in case of Abaqus. This software made me develop the habit of using and advising the use of documentation. Ansys seems to have nice manuals too so I refer to them sometimes.