Xform Roll --- You can either roll or unroll about an axis.

CAD-CAM guys don't know much about cam controlled machines -- including me. I'm sure that you don't have trouble cutting a 2d profile. I suspect that the problem is getting the cam profile in the first place. I don't know the rules to get motion from a cam, but I might be able to help if you could explain the rules.

Just make a toolpath that does what you want to do, and don't worry about the angles. The postprocessed toolpath will give you everything. In order to make a 5 axis toolpath, you may need to make a ruled surface and cut it with SWARF5AX.

The GEAR c-hook will give you the correct involutes, but the OD and ID have to be changed. If you send me the involute spline specs, I can give you the OD and ID. Basically, the tooth height is like a gear that has twice as many teeth (smaller addendum and dedendum). There are some delicate clearance issues too, so it helps to know the class of fit (and type of fit also).

Oops. Sorry, the solid cam was the reverse of what it appeared to be from the 'lines' level. The same rule applies: Drive the follower center with Rolldie. You may have to offset the solid surface to get the follower center curve.

For barrel cam slots, use the Rolldie SLOT option, and create geometry at the slot centerline. Then the slot wall will look parallel to the follower.

Starting with 2 lines and 2 arcs, break one of the lines on the CL. Create spline from curves, starting at the break. Break spline into 26 pieces.

Starting with 2 lines and 2 arcs, break the line at the center point, and create spline from curves. Break into 26 segments.

Set CPLANE to be the same as GVIEW. If you send me your model, I can help you orient your part on the sine plate.

Yes, Rolldie is the way to go, but you need to use the follower centerline for geometry. Unroll the geometry, and offset by the follower radius to find the follower centerline. You can make the Rolldie toolpath from the unrolled geometry -- no need to roll it back up before making the toolpath. You can email me if you have more trouble.

You need splines when a shape cannot be made out of lines and arcs. Many splines are made from a bunch of points which are produced by mathematical formulas. For instance, cams, airfoils, involutes, etc. Other splines are made from solid shapes, such as auto bodies. These are actually more difficult, because the original model has to be digitized first, as a series of points, then read into Mastercam as a spline.

There's no image.

Is the helix giving you a lot of points? You should only need the first and last point. Maybe the feedrate is jamming because it has to process too many points.

I'm sending you a part that shows the rotations. The final machining view is TOP. I rotated the part instead of views. I think it makes the action clearer. Before you start, create a point at the center of the tooling ball. You'll need it after you get the part rotated. You don't need any particular machining origin, because the hole is only relative to the ball. But if you make the ball center the machining origin, the offset to the hole is easier. I hope you can leave some stock so you don't have to start the hole on an angled surface.

You need to explain more about your rotary and tilt tables. The problem here is that you the part needs to tilt about an axis which is not parallel to a part edge, so it will have to be clamped at the 20.483 angle on the tilt table. I'll add more once I find out what kind of a tilt table you have.

The previous post on knurling (8-27) was very informative, complete with video link. I didn't know they made tapered knurls. Do a search.

My first response was very poor. When a surface is untrimmed, the flattened boundary matches the surface. However, when the surface is trimmed, the flattened surface is the flattened untrimmed surface, and you have to map the boundaries by first making edge curves, converting them to surface curves, and them mapping them from the 3d surface to the flat surface. For these surfaces, it might be much easier to create edge curves in the curved direction, and creating a ruled surface. The flattened ruled surface will have the proper boundaries. When you flatten the ruled surface, have the arrow go in the direction of the rulings. I'm sending you one flattened surface.

Since these parts are only plane-cylinder-plane parts, you only need to flatten the cylinders. If you have a right-angle cut of the cylinders, it's just the edge length. Then add the planes separaely. Otherwise, create a single surface which includes the three sections.

A lollypop cutter might save you from having to tilt the table. The perpendicular groove would have a little undercut, which would be OK with the cutter if you were careful with entry and exit. Of course, tilting the table would work also -- whatever's easier.

That's a lot of taper for a knurl. If you did the large end and small end diameters independently, there would be a different number of knurl points on the circumference. I think you need to mill it, rather than knurl it. You really need to get ahold of a mating part. I hope you post the final solution.

You could add 'n' to the FMT statement for R. You may have to add another FS line and number just for the R value. Just copy the FS line that R refers to now, and give it a new number, and add the 'n'(without the quotes). 'n' means non-modal.

I sent you a new file showing the toolbit and the tilt of the toolbit.

Sure.

A point in the right side view at 0 degrees in Cartesian coordinates will appear at -90 degrees in the top view (assuming +CCW rotation). This can cause a lot of mismatches. Without seeing your part and methods, this is my best guess as to where the mismatch is coming from.

Sorry, the pitch diameter is .500. Somewhere, I pushed the wrong button.