Hello,
l have some modeling problem and asking about help. See attached file (Creo 2.0)
l´m modeling slot on circle surface with a problem in transition between flat part and curved part.
Red marked surface doesn´t looke fine ... l think something is wrong in my modeling procces (but what?).
How to model it in order to use round feater on inner edge?
Is my technique correct? (using warp feater)
Thx for guidance
Regards
Milan Bonka
You need to adjust a few controls inside sweep feature.
Have a look on the attached image.
Hope this helps.
Hi, I have had a little go at doing this feature. I created a sketch plane above the cylinder and sketched a trajectory, this is then projected onto the cylinder. Use a sweep on the end to create the slot profile and project this along your trajectory. I had some issues which you may need to address too, if the included angle is too great it will project around the other side of the cylinder too so don't take it right to the edge. I have attached a couple of screenshots to show what I mean. You shouldn't need to patch any rads in. If you included angle is too great then this may not work (show as about 45 deg. in my example)
In this case, the cut geometry will actually work; because a projected curve curve used as the trajectory, the sweep section's horizontal/vertical orientation will be controlled using the "Normal to surface" option (the surface onto which the trajectory was projected). This means that the cutter axis will always go through the stock axis. And since the section plane will stay normal to the trajectory, such sweep will faithfully represent a cut made by an endmill that moves along the stock axis while the stock is turning.
Because of the limitation of the projection curve only working in the 1/2 of the cylindrical surface, using wrapped sketches to generate the trajectory curve might be preferable.
Actually, that's totally incorrect. The difference is that a cutting tool is a side-cutting device, and will remove MORE material that your cut, sometimes a LOT more depending on the sweep trajectory (the tighter the curve radii the bigger difference). Take the trajectory and pattern a series of points along it at, say, .010" intervals. Then pattern cylindrical cuts of the cutting tool size at every point and you'll see what I mean.
This is a known issue, and there is at least one thread dedicated to it. 
You are totally right Frank. For some reason, I always wrongly thought that for a case where the axis of the endmill is normal to the cylinder it is cutting, the "Normal to trajectory option" produced correct results. I recall a while ago, I wanted to see which option of helical sweep section control (normal to trajectory vs. through axis of revolution) would produce a more faithful representation of an endmill cutting a helical path. For this, I used the very same pattern of cylindrical cuts you suggested (in green below). I saw that the "through axis of revolution" option (in white) was not producing correct geometry, while while the "normal to trajectory" option (in gold, being highlighted below) seemed spot-on:
Today I "dug" deeper into the issue and sure enough, the extra tip gouging is easily revealed:
It's an easy mistake to make, I did it at first, blindly believing creo did what I WANTED, instead of doing what I TOLD it to do. And the tighter the curves on the trajectory, the more the actual cut is different. Antonius did a lot of great work to create a true cut on, I believe, a globoidal cam? It's because the section of any swept cut is PLANAR, and only cuts as a plane, whereas a cutting tool is cylindrical, and cuts in 3D, if that makes sense.
What we NEED is a "solid body" cut like (but better than), uh, SW has..... 
