Hi,
I'm trying to measure and demonstrate the indentation on the contact points between the spheres and the V grooves.
- the red plate is fixed.
- all the spheres bonded to the holes at the blue part.
- the blue part puled with 20N of force, and get get self aligned with the V grooves.
I tried several times to solve this simulation, but all the times I was getting errors like this one.
also I'm attached zip file with all the parts and the assembly.
I did a test today with cube on a flat surface and it work just fine.
only with round objects there is an issue,
I'm start thinking that there is a bug in solidworks about that.
Any solution?
You can't do this. You have a singularity at the point of contact of the sphere on the flat surface of the vee. This is similar to the problem I submitted (supposedly fixed in 2016) relating to bearing contact stresses. FEA doesn't like point contact.
There are few choices of how to get any meaningful results. One is to put a small flat on the sphere precisely such that it makes a flat contact with both vee surfaces.
James Riddell, thank you for your answer,
are you sure there is no way to do it in solidworks 2014?
if 0 surface at the contact points equal to infinite pressure, maybe there is a way to set some tolerance?.
according to the specs the surface texture that is less than 1.0 microinch (25 nanometers) Ra.
do you have any idea how apply that kind of roughness to the surface of the sphere?
First, what is your goal? It is easy to use Roark's to figure bearing stresses (or this website: http://http://www.amesweb.info/HertzianContact/HertzianContact.aspx) as long as you know your forces. You can get the forces out of SW easily enough if you break the problem down.
If you are more concerned with the other parts of the assembly then just add some flats to the spheres and run it that way.
I had an issue (submitted) and I had even tried getting super-fine mesh in the contact area, it doesn't even come close to theory or reality.
I'm aware to that website, and yeah, I can divide the force by 6 (2 contacts for each sphere) and get some estimate for the dent.
But at the end I want to make a compression of different angles of V grooves and even to try cylindrical grooves.
Hello Sharon,
Unfortunately, I do not have SW2014 at the moment to provide you with the files, however, I solved your simulation issue.
Please apply and consider the following changes into your simulation to increase the accuracy or results and overcome the issue.
1) Use Curvature based mesh (increase the mesh density as well)
2) Under definition of the contact sets between spherical faces and other flat faces, under advanced option: use Node to Surface contact type
3) In the simulation properties: a) use soft spring b) check: more accurate ...(slower) c) use direct sparse solver
Please let me know if it solves the issue.
Cheers,
Siavash
Siavash Khajehhasani, thank you for your replay,
You rally gave me some hope
I try to follow after your instructions , but I can't find this option.
2) Under definition of the contact sets between spherical faces and other flat faces, under advanced option: use Node to Surface contact type
Other than that I set all the other settings, and took some screenshot for your verification.
I tried to run the simulation and got 3 errors (also in the screenshots).
That #2 option is under the definition of your non-penetrating contact set(s) that you have to put between the ball and the vee.
Sharon,
For now, the response from James should work for you (thanks James).
I try to instal SW2014 by this evening and send you the solution; if I could not, I will provide you at least with some screen shots.
Cheers,
- Siavash
James & Siavash, thank you so much for you help
But now instead of crushing with error it's run forever end ever, I let it run several times, but after it reach 5% it's stay there even after 4 hours.
4 hours is more than the time that take to machine a prototype.
from my perspective the simulation suppose to save you time not to waste your time.
should I let it run over 4 hours, or clearly something wrong with the simulation?
Hi Sharon,
Please find attached your simulation case with an extra simulation tab provided by me "Simulation by Siavash", which all of the recommendations mentioned before are applied in it. Enjoy!
It did not not even take 2 minutes to obtain results on my machine!!
Please let me know if you still have any issue.
Cheers,
-Siavash
Sorry for the delay in my response, (tight week)
Siavash, Thank you soo much!
I succeed to run your simulation!
but other than that there is something that look a bit incorrect on the simulation specially because in solidworks the prats are 100% in dimension, something that in the reality is in a certain tolerances..
As you can see on one side of the Vee groove the stress is higher than the other side and even located on different height.
from my point of view it's look like that the parts in solidworks don't slide to the center and lock in position as happen in the reality.
Again. thank you soo much for your assistance.
What are the stresses reported in the FEA vs. calculated stresses. I'll make a $1 bet that they are wrong in the FEA.
Hi Sharon,
Please find attached your simulation case with an extra simulation tab provided by me "Simulation by Siavash", which all of the recommendations mentioned before are applied in it. Enjoy!
It did not not even take 2 minutes to obtain results on my machine!!
Please let me know if you still have any issue.
Cheers,
-Siavash