Depending on how smooth running you want the groove I have seen it done using a rotary table and a set of co-ordinated generated from the CAD model. Basically a series of overlapping plunge cuts where you adjust the x-axis for each plunge. If I ever get round to it this will be the way I do part of the winding gear on my Ploughing engine.

Most of the valve shuttles that I have seen with this type of groove (two actually) have been done with 4th axis CNC.

Clamp it to a lathe faceplate with packing under one side to give the required angle. The clamping will need to be in the centre hole, you may need an intermediate plate with suitable threaded holes in it. If there are a number to be made make the intermediate plate into a jig. There will be a slight variation in depth of the groove due to the angle of the work piece. Centering will have to be done carefully as the od will not appear to be round when mounted at an angle.

Martin

Edited By Martin Connelly on 17/12/2015 09:05:48

If you don't have 4 axis cnc, or too much time on your hands with a rotary table, make it in 3 pieces, glue it together (or solder if you're fancy).

Seriously … that's how the original was done. It's the easy way.

An offset workpiece between centres will give you an elipse, which will not have a uniform depth of groove (although at 16 degrees it's probably not that much out)

You can't angle the workpiece if you want uniform depth.

Martin

If you have a lathe with no backlash in the halfnuts (good luck with that one), all you need is a mechanism to rotate the leadscrew briefly in one direction , then the other., by the correct amount.

A simple crank mechanism or eccentric mechanism as used on steam engines comes to mind as the most common directional reversing mechanism.

You could, theoretically, drill a large change gear on the end of the leadscrew and press or screw in a crankpin as such.

Then do the same with a small change gear on the quadrant. Set this up so the small gears teeth do not engage with the leadscrew gear. But join the two crank pins together with a suitable length piiece of flat bar with two holes to fit the two crank pins. Gears would have to be arranged so the link bar is at a tangent to the PCD of the crankpin on the larger gear. So as the small gear rotates, driven by the train of gears from the mandrel, its motion is translated into a small back and forth motion of the leadscrew gear by the crank and link assembly.

Calculating the exact PCDs of the two crankpins and the geartrain to drive the smaller gear is a job for wiser heads than mine. I would probably just bodge about with a sliding crank pin assembly until I accidentally got it right.

But it seems, on further reflection, that the small gear would need to run at 1:1 with the mandrel. So the tool returns to the same point once per revolution. The PCD of the crankpin on the small gear would have to give the total 'stroke" to move the crankpin on the large gear back and forth by the amount you want the leadscrew to turn, which can be determined from the lead screw pitch and the amount of "lead" on the groove.

If the amount of movement is more than one revolution of the leadscrew, the crank mechanism would have to be on the mandrel end of the gear train and suitable gears used to increase the rotary movement the correct amount.

Edited By Hopper on 17/12/2015 12:42:50

Something a bit like this?

Machine it so the work piece is concentric at the point you want the machined face.

Used this to make some inlet manifold adaptors to match some injector bodies to an inlet manifold with different, siamesed centres.

Worked OK but was a pain to make as you can see. Ideally you'd reduce the overhang more than I bothered to. And even more ideally you'd provide tailstock support.

In the trial in the bottom photo, the workpiece will be turned around so the cylindrical faces of both ends are parallel but offset.

Edited By Muzzer on 17/12/2015 13:34:23

The reason I got into model engineering was the need to make tools for making jewellery. I would tackle your ring by hand-filing the groove after laying it out with a piercing saw. Use the depth of the saw blade as guide to the depth of the groove, then use suitable needle files to curve the edges. Ti isn't hard to saw or file.

The 16 degree angle of the groove means that its actually an ellipse.For an average sized ring (say 20mm ID) the depth of the groove, measured normal to ring's axis, varies by almost 0.4mm. In other words, the difference in length of the major and minor axes is about 0.8mm. That will make the width of the groove vary very obviously if the cutting tool isn't moved that much radially. Better to do it by hand.

Frank's scheme but with a toolpost milling attachment would be even better. If you do use a fixed tool, I'd go for a bit more than 0.0001" at a time, or it could easily rub rather than cut

Hi Andrew,

To mark it out I would get a bit of pipe (copper, plastic, iron, whatever) with an ID suitable for boring out to accept the ring. Cut the end off at your required 16 degrees, insert the ring and follow the pipe end with the saw.