ACCEPTED SOLUTION
Accepted Solutions
Yes a cylindrical csy (created in simulate mode) is a way to do this. Always attempt a Free Body Diagram. You seem to be missing the loads/constraints at the output shafts. (R in my image).
Also, the pinion load may be statically redundant if using 2 pinion gears as is typical. The load may be higher on one pinion gear than another based on manufacturing tolerances and wear. I would not assume full halving of stress and load is happening due to the 2nd pinion gear.
Also see this for my quiz on mechanizing a differential.
Yes a cylindrical csy (created in simulate mode) is a way to do this. Always attempt a Free Body Diagram. You seem to be missing the loads/constraints at the output shafts. (R in my image).
Also, the pinion load may be statically redundant if using 2 pinion gears as is typical. The load may be higher on one pinion gear than another based on manufacturing tolerances and wear. I would not assume full halving of stress and load is happening due to the 2nd pinion gear.
Also see this for my quiz on mechanizing a differential.
When using cylindrical CSY, Theta is the rotation.
Degree of freedom 1 is R (radial), DOF 2 is Theta (rotation about Z) and DOF 3 is Z translation.
Leave those other rotations free. (DOF 4,5,6)
My take on this model is attached.
Analysis1 has bolt preload but no finite friction on the contacts (CREO limitation) To run this you need to turn friction to none on the four contact interfaces (Interface2,4,5,6)
Analysis2 does not have bolt preload but has finite friction. The interface frictions might need to be turned back on. I used a Creo parameter named MU_STATIC to control the friction from a parameter. Instead of typing 0.2, type MU_STATIC. To change the value use TOOLS - PARAMETERS.
Some parts were de-featured and element size increased to reduce element count (to about 50k elements)
Volume regions were used to separate key contact areas better and to reduce contact searches.
Bolt preload was modeled as volume prismatic preload using the portion of the bolt within the clearance hole.
Some light springs were added to prevent unconstrained parts. (roll pin and pinion bolt)
The 3 gear loads were merged into one load defined by three components. Also surface regions were added to the gear to spread the load over portions of 3 gear teeth.
The ring gear was made a free interface to the housing and then a contact interface defined at the bolt face to absorb the bolt preload.
I think the axial force that opposes the helix angle will be taken by the end of the housing not the pinion bolt so I only held the pinion bolt in Theta. (The light spring was used to stop any small movement)
Your solver memory was set to 512MB. This should be about 1/2 machine ram up to 16384MB.
I used 4000MB. Analysis1 took about 40 minutes. Analysis2 has not finished yet.
Yes, finite friction does not work well. However this worked. It took about 2 hours to run.
Note that results are very similar to Analysis1, however the contact pressure shows how we are missing the contact pressure due to bolt preload. My opinion is that the finite friction is not needed. I might compare friction None to friction Infinite but doubt that makes much difference either.
The shear stress at the pin is about 130MPa by hand calculation. The model has the pinions further from the housing than reality so bending stress is artificially introduced to the pin. This is to keep St Venant's principal working as we get to the housing by having the constraint further away. For more correct stress on the pin, the pinion gears would need to be included, with contacts to the pin, or the constraint must be replaced by a bearing load and then constraining at the gear tooth instead of loading. Or a third option is to have full free body loads and applying inertia relief constraint.
Sorry about using Creo5. When I get time I will try to demonstrate how this was done so it can be replicated in Creo4.
I don't know of a way to convert from Creo 5 to Creo 4. Also there are just so many steps taken to describe. I can give images... sorry too many words to explain. I will let you ask the questions where it is unclear.
