Learning Forum

[cunyi.ji]

Hello,

I am learning Ansys Mechanical. I have been doing a very simple simulation to learn about Harmonic Response MSUP. The geometry is the following:



The load is a harmonic torsional moment applied in the X direction:



The following surface is fixed:



The natural frequencies and modes found in Modal Analysis are:

1 and 2: 13305 Hz, bending mode
3: 22755 Hz, torsion mode
4: 34697 Hz, tension-compresion mode
5 and 6: 38814 Hz, another bending mode

The average directional deformation in Y axis of frequency response is the following:



The average directional deformation in Z axis of frequency response is the following:



The resonance happens at the torsional mode, which makes sense, since I applied a harmonic torsional moment in the X axis. But why is the directional deformation in Z axis about twice the one in the Y axis when there is polar symmetry?

The direction deformation in Z and Y axis seems to be the same when plotting the distribution in the object:




What is causing this? The mesh has around 500.000 nodes, which I think is overkill for this simple model, so poor mesh shouldn't be the problem. Also tried a coarser mesh, same result, so fine mesh isn't the problem neither.

Thanks.

Regards,

[dlooman]

If you scoped the frequency response to the face, the expected average value is zero for both UY and UZ.  Your contour display of UY and UZ at 50,000 Hz shows that.  When you display a zero quantity some numerical noise becomes significant.  That's why your frequency response values are random.  (Note how small they are.)  If you display UY (tangential) in a cylindrical system the average will be non-zero.

[cunyi.ji]

Thank you. So the negative and positive directional deformation was cancelling out. I thought it was taking the average absolute value.

[dlooman]

Right.  The positive and negative values were canceling out.  You weren't getting the average amplitude.  You could have requested "Total Deformation" which is an unsigned quantity.