Issues with cycloidal gear made by Inventor's Curves Samples

[WHolzwarth]

Though it seems a rather simple task at first glance, I'm having some problems with it. Look at the theory (4th and 6th picture):

http://switzernet.com/people/emin-gabrielyan/100512-gear-engine/

I've added my 1st attemt to this topic: Graphics Issues in Inventor 2014, and Ray Dyson commented it already

In Inventor's Sample files, you can find Epitrochoid&Epicycloids&Cardioid.xls and Hypotrochoid&Hypocycloid&Astroid.xls. I've changed them by trial and error to the attachments, for getting the basic shapes for the gear.

Problems in the Excel files:

- First point is not calculated, in both XLS files. That's the reason for the error, that Ray has seen.

- Point numbers other than 300 seem to be not accepted, and need to be corrected by changes in Angle limits (min - max)

- Degrees are better being viewed, than radians

The Point No. - Limits combination seems to be done in an Excel Macro. Well, I don't have big experience with Macros. Perhaps someone can take a look.

Ray added a tiny straight line between the splines. It closed the gap, but added a tiny face between the splines. Looking good, but turns out to be a troublemaker.

In the end, an animation of two moving gears shall be seen. With my files, distance between axes must be set to 24 mm.

- I can drive both rotation angles by the same amount. This works

- I can drive one angle and try a tangent constraint to the other gear. Works only regional

- I can drive one angle and try a transitional constraint between the gears. Only regional

- Sometimes Dymamic Simulation can help out. But no luck at all with this sample.

😉 Come on, guys. It's interesting. And Inventor still is provided in the Suites..

Walter

[WHolzwarth · Accepted solution]

Updated version now. I've added the missing points, for getting continuous curves, and made a small IAM.

The graphics issue seems to be gone, but after deleting six side faces, a boundary patch between top and bottom of Zykloidenrad 3z-V2.ipt could not be applied. That shouldn't occur (it's not included in the feature browser).

I could drive this by sytem by constraints, but I think, that's only theory. In real life the blue part can't drive the yellow one all along 360 degrees. Driving action needs to be changed between the parts.

With only the blue part driving, system would behave as can be seen in DS environment.

Walter

[JDMather]

Rather than use formula, have you tried creating the curves using the Trace function in DS?

Looks like an interesting experiment for my students.

[WHolzwarth]

Not with this sample, Jeffrey. It only would have been more complicated, because the need of creating substitute mechanisms for the resulting curves.

But I've done similar movements in the past, see picture. Internal gear is turning CW, tool is turning CCW with 3x more angular velocity. Then you can trace point movement like this, in relation of tool versus gear.

Walter