Learning Forum

[VeraL]

Hi,

I am posting because I have a problem with my modal boxes.

First, I've got a static structural box where my equipment is pressured and where the screws have bolt pretension. Then, I link a modal box so that my static structural is my pre-stress environment.

Then, I copy my two boxes and I rotate my equipment of 45°. Then I perform my static structural and my modal. But, I don't have the same frequencies between my two modal boxes (without rotation and with 45° rotation).

When I copied, I didn't change anything. Something must be changing when I rotate my equipment but I can't find what.

I have two joints in my models but the local coordinate systems are the same.

Does someone have an idea?

 

Thank you,

[peteroznewman]

Remote Points scoped to a face have the point created at the centroid of the face and the coordinates of the point are saved in Global coordinates.

After a 45 degree rotation of the part, the face ends up elsewhere, but the remote point does not move. It remains at the original location that was recorded in Global coordinates. You have to repair every remote point to fix this problem.

This is true for implicit remote points such as for a Remote Displacement or Remote Force, not just explicitly created Remote Points.

[VeraL]

Thank you Peter.

I don't have any remote displacement or remote force. I re-selected my faces in my joints but it didn't change anything.

In my static structural analysis, I only have fixed support, pressure and bold pretension...

Do you have an other idea?

[peteroznewman]

I did a quick test on a bolt pretension load to see if that had any defect in the direction of motion of an adjustment load in 45 degree rotatated geometry compared with the original geometry and it behaved correctly in both cases.

Check that all bodies have the correct material assignment. I don't expect that would have changed.

If you can't find what changed, all you can do is start with a new Static Structural and rebuild the entire model with the rotated geometry and add a new Modal analysis to that new model.

Why did you need to rotate the geometry in the first place?  Is there an alternative where you don't rotate the geometry?

[VeraL]

Ok I will try to start again but it will be quite long.

I need to rotate the geometry because I will perform an harmonic response and the axis of excitation given had changed.

[peteroznewman]

I assume you are rotating the gometry in SpaceClaim or DesignModeler.

Instead of that, use Mechanical, use Part Transform and rotate the model there. Maybe you get a better result.

[VeraL]

It worked with part transform. Thank you !
I've got a SR open with ANSYS, if they give me the solution, I will put it here.