
Re: Contact stresses in parallel cylinders
Bill McEachern Jan 18, 2018 12:32 PM (in response to Rahul Mula)it would be helpful if you showed the mesh and the hand calc, otherwise every one is just guessing at what the issue is. This has historically not been one of SWX 's strong suits and you need to correct the VM stress for a state of pure shear for the peak stress below the surface.

Re: Contact stresses in parallel cylinders
Rahul Mula Jan 18, 2018 1:53 PM (in response to Bill McEachern) 
Re: Contact stresses in parallel cylinders
William Radigan Jan 19, 2018 7:50 PM (in response to Bill McEachern)
Re: Contact stresses in parallel cylinders
Bill McEachern Jan 21, 2018 9:07 AM (in response to William Radigan)take a look at the VM wikipedia

Re: Contact stresses in parallel cylinders
William Radigan Jan 22, 2018 11:04 AM (in response to Bill McEachern)Thanks Bill! I appreciate your expertise on this!
Referring to this article: von Mises yield criterion  Wikipedia I think you're pointing out that:
"at the onset of yielding, the magnitude of the shear stress in pure shear is {\sqrt {3}} times lower than the tensile stress in the case of simple tension."
... And that Von Mises stress is therefore not a good failure criterion for contact stress problems, is that correct?
My question relates to the orientation of the coordinate system that is being reported when a user follows steps like what I showed in the figure above. According to: Cauchy stress tensor  Wikipedia (or here: 2017 SOLIDWORKS Help  Principal Stresses Definition )
The components {\Tau _{ij}} of the stress tensor depend on the orientation of the coordinate system at the point under consideration. However, the stress tensor itself is a physical quantity and as such, it is independent of the coordinate system chosen to represent it.
If I ask for "Tau_xy" and select a plane, do I get a "Tau_xy" that is oriented according to the global XY plane, or a Tau_xy that has been 'rotated into' my plane of interest? Or something else?
This (2017 SOLIDWORKS Help  Stress Components ) makes it sound like it does do the rotation, but it's still a little unclear to me.
TXY Shear stress in the Ydirection acting on the plane normal to Xdirection of the selected reference geometry For shear stress components, the first index indicates direction of surface normal, and the second index indicates direction of shear stress component.
Is there an additional 'correction' that needs to be applied before these shear stress values are used?

Re: Contact stresses in parallel cylinders
Bill McEachern Jan 22, 2018 12:31 PM (in response to William Radigan)If you are comparing the VM stress to a hertz hand calc you need to adjust the sim VM stress in the subsurface shear zone to account for the fact that is it pure shear in that location to compare apples to apples.

Re: Contact stresses in parallel cylinders
Mike Pogue Jan 22, 2018 2:55 PM (in response to Bill McEachern)To agree a little further with this, you ought to be able to plot max shear, rather than VM. Compare the peak max shear stress a little bit inside the surface above the point of contact and it should agree with the failure criterion in Roarke's or other sources. Assuming the analysis is correct, of course, but at least you'd be comparing applestoapples.

Re: Contact stresses in parallel cylinders
William Radigan Jan 22, 2018 4:53 PM (in response to Mike Pogue)Thanks for the help guys, and I'm sorry for being a bit slow on the uptake here. I think I ought to be able to plot "max shear" as well, but I'm not sure that I can. The closest that I can come is plotting "Stress Intensity (P1P3)" and dividing by two per the note below. Is there something that I'm missing?
And yes, I am comparing to hand calcs from Roarke and specifically to the subsurface shear (\Tau_xz). Thus far I've just used whichever shear stress component that had the largest absolute values, but Bill's original post about a 'correction' made me realize that this might be incorrect.
Referring back to 2017 SOLIDWORKS Help  Stress Components again:
(a) In some design codes and references, the Tresca equivalent stress is defined as twice the maximum shear stress which is equal to (P1 – P3), or else the stress intensity.
cc: Bill McEachern

Re: Contact stresses in parallel cylinders
Bill McEachern Jan 22, 2018 8:25 PM (in response to William Radigan)It is easy if you just correct Von Mises.

Re: Contact stresses in parallel cylinders
William Radigan Jan 23, 2018 3:10 PM (in response to Bill McEachern)I'm feeling pretty foolish here, but is this as simply as multiplying the reported Von Mises stresses by SQRT(3)? If so, is there a way to produce a plot with this 'correction'?







