Learning Forum

[Saurabh]

I have a plastic material (Stanyl TW271F) for planetary gears in the gear train analysis. I am trying to get stress distribution when the sun gear is driven at 6 N-m moment applied through joint load at its center.

I have a mesh with quadratic element order, adaptive meshing on, fast transition, resolution 6 settings with Nodes = 908055 & elements = 534399. 

I have all frictional contacts with Adjust to touch. 

I am getting extremely high values of stress at results and the output graph plotted for Equivalent stress vs Equivalent total strain does not match the input curve of the material.

I am trying to upload images here but I am unable to do so. (I will try to insert corresponding images in comments below this if it allows me to do so later).

Can someone help me troubleshoot what am I doing incorrect? 

Thank you in advance.

[Saurabh]

This is an overview of my model and problem: 

[Sean Harvey]

Hi Saurabh,

A couple of points. 

 

• I would not use the adjust to touch in this application as it is closing the gap between the parts and changes the physics of the problem. It is as if you put a bit of filler between the contacts. If you have convergence without it on, then I would suggest you use contact stabilization damping. Start with small values and increase. You don't want too high a value as the energy will go into the damping and not the elastic deformation of the structure.
• Turn on the nodal projected normal from contact for the contact detection method 
• For the mesh, you can used sizing or split the model and just get the fine mesh closer to the contact regions and roots of the gear teeth where you see the high stresses.   You will need a nice smooth fine mesh at the tooth to tooth contact. I can't tell but from what you shared in the image your mesh is too coarse and you are most likely getting extrapolation error. 
• Please see lesson 2 from this Free course. How to Evaluate Stress and Yielding in Plasticity | Ansys Courses  This explains the extrapolation you may see when the results don't match the stress-strain curve.

Circle back and let us know how you make out.

Regards,
Sean

[Saurabh]

Thankyou for your response. Appreciate it. I adjusted the physical geomtry of the gears to avoid any gaps and ran the model with add offset (zero). It seems to have worked. I watched the free coarse you referred to, I will keep these points in mind for my future analysis. Thankyou.