Community Tip - New to the community? Learn how to post a question and get help from PTC and industry experts! X
Gurus,
I've just been assisting a colleague (who's new to Mechanica) to set up an analysis. This analysis lends itself to cyclic symmetry, albeit only two-fold. He's successfully split the surfaces (actually using volume regions I think) and created a cyclic symmetry constraint, which has correctly recognised a 180° angle.
However, when we run the analysis there is significant displacement into the symmetry 'plane', and it's all in one direction ('outwards' or 'plus material') rather than the 'S'-shape I'd consider acceptable for this constraint. It looks suspiciously as though it's not actually enforcing the constraint, or at least not as we want it.
Possibly relevant: this is a contact assembly with two components and several volume regions on each, underneath the contact location in 'onion layers'. He's selected all the relevant surfaces for each half of the symmetry constraint, and there are no obvious discontinuities in the stresses which would suggest a missed surface.
Has anyone done this before and encountered similar?
Thanks!
Solved! Go to Solution.
Jonathan,
Another colleague of mine looked at it too and, like me, right away said this is not an application for cyclic symmetry (CS) - together we have about 45 years of FEA experience. The general issue is this is not a part/assembly that is a repeating section within a cylindrical system. I’d like to say that one major rule is the number of sections must be greater than or equal to 3, but I cannot find this written anywhere. Cyclic symmetry is intended for parts that are repeated rotationally. We are not able at this point to provide a mathematical reason, but I'll try to explain as best I can. (Maybe somebody else can chime-in with a good one)
First, I am unable to get Simulate to mesh the model. Cyclically symmetric models require the "low" and "high" sides of the symmetry planes to have exactly the same geometry - surface-for-surface, line-for-line, etc. Your model does not have surface areas that are identical. This will cause havoc with the mesher. In addition, there is an extremely small surface at one end of the roller that is likely also wreaking havoc with the mesher – I’m using Creo 2 (M050). Two times the mesher wouldn't even try and the third time it ran for almost 10 minutes without creating one element. I did get eventually get Simulate to mesh the part when I removed all the additional volume regions that help with the contact, but it never gave “low-high” regions that were identical on the roller’s cut surfaces. When I tried to have it solve, Simulate crashed Creo (not surprised here).
From a cyclic symmetry point of view, constraint equations are used to “link” the two cut planes together so they deform in exactly the same way. In a typical CS part, the angular spacing, of 120° or less, of the symmetry planes keeps the part from moving in the “radial” direction except for small displacements due to stiffness that allow identical movement on each plane. When the CS cut planes lie in the same plane in space, you will have issues with the general constraint of the part. My colleague and I agree the underlying issue is with stiffness and how it doesn’t work when it’s a 180° section. We also agreed that this should just be solved as a full model. You may be interested to know that ANSYS will not solve this as a cyclic symmetry model either.
As far as hand calcs go, I would look to Roark’s and the “general case of two bodies in contact” (chapter 13, case 4). BTW – having spent 13 years at Torrington (Needle Bearing Division), 15° skew is huge – I hope this just an exaggerated example, otherwise you’ll have issues with load distribution and skew-lock!
I hope this helps,
Chris
Jonathan,
How long does it take you to mesh (autogem) the assembly outside of the analysis?
First time I tried to mesh it manually it gave me an error
Could not complete creating elements
on all of the selected entities.
Source: AutoGEM
But nothing was highlighted in the model
Second time I tried meshing it it just kept grinding away. Would think this would mesh fairly quickly.
This was done using Creo 2.
Don Anderson
Jonathan,
I may be wrong from what I can see in the model you shared, but it looks like you have applied a cyclic symmetry constraint incorrectly. The help inside Creo Simulate does not seem to provide images of examples, but a Google of the subject might help. PTC's University does have a good reference video if you have access:
http://learningexchange.ptc.com/tutorial/52/understanding-cyclic-symmetry-constraints
Cyclic symmetry is used to analyze repeated sections of a revolved part that has deformation occuring out-of-plane from each cut, but in an identically deformed shape. The slender cylinder that is cut does not look like it would repeat itself properly using the axis that was selected - the loading is not creating the same amount of strain energy on each split section. Am I missing something from the complete model that is not shown? Also, it is unusual for Simulate to allow an axis of revolution along a common element edge from each "side" of the symmetry cut planes.
BTW - this looks like a roller bearing skewing problem .... Hertz can handle this situation with a hand calculation.
Chris
Don,
We had a similar meshing issue - inexplicably, suppressing the cyclic constraint allowed it to mesh, and after resuming it still meshed!
Chris,
I've used cyclic symmetry plenty of times before and I hope I have a good understanding of what it does - why do you say this is applied incorrectly? The model and loading has two-fold rotational symmetry as far as I can see.
The axis of revolution along a common edge shouldn't be a problem - in principle this is no different to analysing a 'wedge' of a solid shaft, except that in that case there would be no need to explicitly specify the axis, as Mechanica would infer it from the two selected 'sides'.
You're spot on, it is a roller bearing skew problem. My colleague started with Hertzian calcs but found they only seemed to give sensible numbers for very large skew angles, ~25-30° and larger - at small but non-zero angles the results didn't seem believable.
Do you have a good resource for skew-cylinder Hertzian formulae?
Thanks,
Jonathan
Jonathan,
Another colleague of mine looked at it too and, like me, right away said this is not an application for cyclic symmetry (CS) - together we have about 45 years of FEA experience. The general issue is this is not a part/assembly that is a repeating section within a cylindrical system. I’d like to say that one major rule is the number of sections must be greater than or equal to 3, but I cannot find this written anywhere. Cyclic symmetry is intended for parts that are repeated rotationally. We are not able at this point to provide a mathematical reason, but I'll try to explain as best I can. (Maybe somebody else can chime-in with a good one)
First, I am unable to get Simulate to mesh the model. Cyclically symmetric models require the "low" and "high" sides of the symmetry planes to have exactly the same geometry - surface-for-surface, line-for-line, etc. Your model does not have surface areas that are identical. This will cause havoc with the mesher. In addition, there is an extremely small surface at one end of the roller that is likely also wreaking havoc with the mesher – I’m using Creo 2 (M050). Two times the mesher wouldn't even try and the third time it ran for almost 10 minutes without creating one element. I did get eventually get Simulate to mesh the part when I removed all the additional volume regions that help with the contact, but it never gave “low-high” regions that were identical on the roller’s cut surfaces. When I tried to have it solve, Simulate crashed Creo (not surprised here).
From a cyclic symmetry point of view, constraint equations are used to “link” the two cut planes together so they deform in exactly the same way. In a typical CS part, the angular spacing, of 120° or less, of the symmetry planes keeps the part from moving in the “radial” direction except for small displacements due to stiffness that allow identical movement on each plane. When the CS cut planes lie in the same plane in space, you will have issues with the general constraint of the part. My colleague and I agree the underlying issue is with stiffness and how it doesn’t work when it’s a 180° section. We also agreed that this should just be solved as a full model. You may be interested to know that ANSYS will not solve this as a cyclic symmetry model either.
As far as hand calcs go, I would look to Roark’s and the “general case of two bodies in contact” (chapter 13, case 4). BTW – having spent 13 years at Torrington (Needle Bearing Division), 15° skew is huge – I hope this just an exaggerated example, otherwise you’ll have issues with load distribution and skew-lock!
I hope this helps,
Chris
Thanks Chris.
I hadn't spotted that the surfaces weren't identical - I thought they should have been, but as mentioned it isn't my model. Per my reply above I also found meshing errors, but for some reason it meshed happily when the cyclic symmetry was suppressed.
Your point about constraint with co-planar cut surface does make sense to me, and sounds like as good a reason as any to only use cyclic symmetry with 3 or more sections.
The 15° skew was a test case, mostly so that we could clearly see what the model was doing - we also figured that if it works at 0° and at 15° then there's a reasonable chance it will work at any angle in between!
We've got the full model running, anyway, and that's also proved used to investigate tilt (which is clearly not cyclically symmetrical). It's just become fairly slow... overnight we've run sensitivity studies on both skew and tilt angle (on two machines) and both have only reached step 6 so far!
Thanks again for the feedback and explanations.
Jonathan