Best way to cut helical grooves

[swirnoff]

I've been doing some research and I haven't found any ideas for how to cut helical grooves without the benefit of CNC machinery. I have 8 steel rods, approximately 10mm diameter and 75mm long and have to mill a flat bottomed, helical groove down the length. I don't have a CNC mill and will not be investing in one for what would be a VERY small run. Does anyone have any tips? I do have a rotating table with chuck, but I don't see that it will help me overly for this task.

If this sort of work requires CNC, does anyone know of small job shops that would take on this (and maybe a few other complex machining)?

Thanks.

 

[benmychree]

CNC not strictly necessary, a universal milling machine will also do the job with the dividing head geared to the table screw.

 

[RJSakowski]

IIRC, Tom Lipton, Ox Tools, did it on a manual mill with a modified rotary table. The modifications included adding an encoder to the x axis and a stepper drive to the RT.The pitch of the spiral was determined by the ratio of the stepper pulses to the encoder pulses.

 

[markba633csi]

I've seen it done on a specially modified lathe with the leadscrew driven by a motor which in turn drives the spindle
So the leadscrew is turning many more rpm than the spindle without putting too much strain on the geartrain

 

[Huub Buis]

Your helical grove is just a thread with a special form. It can be made on "any" manual lathe that can feed the required pitch.

• You can grind a form tool and "turn" that groove like any other thread.
• If you place a drill/pencil grinder/etc in the tool post, the grove can be milled.

If the groove is small and shallow, I think the 75 mm long bar only needs tail stock support. If not, a follow rest could be required.

Using a CNC lathe or CNC 4 axis router would be my choice of tool. Outsourcing to a shop of a person with the right tools in your neighborhood is a good choice.

 

[swirnoff]

Your helical grove is just a thread with a special form. It can be made on "any" manual lathe that can feed the required pitch.

• You can grind a form tool and "turn" that groove like any other thread.
• If you place a drill/pencil grinder/etc in the tool post, the grove can be milled.

If the groove is small and shallow, I think the 75 mm long bar only needs tail stock support. If not, a follow rest could be required.

Using a CNC lathe or CNC 4 axis router would be my choice of tool. Outsourcing to a shop of a person with the right tools in your neighborhood is a good choice.

Oh my goodness. What an amazingly simple idea. Still complex, but really great. Thanks so much for this.

 

[Janderso]

CNC not strictly necessary, a universal milling machine will also do the job with the dividing head geared to the table screw.

That's the old school way. I believe Keith Rucker did that on his K&T?
Technology sure has come a long way.

But then on the other hand, the Pratt & Whitney, Wright and Rolls Royce engines manufactured for WW2 production was all done without computers.

I am always in awe when I think about the quality, durability and sheer numbers of these engines produced.

Ford's Willow run factory built a B-24 every hour and six minutes. They had the Pratt & Whitney R1830.
Sorry, I got carried away.

 

[homebrewed]

One possible issue with using a lathe is the large ratio of spindle vs. leadscrew rotation. The mechanical advantage, from the change gears' point of view, is terrible because it has to drive the carriage a LOT relative to the spindle. On the other hand, reversing the drive so the LS drives the change gears produces a very good mechanical advantage for driving the spindle.

I attached a crank to my LS (for other reasons) and it wasn't all that difficult to do. But that may not be the case for your lathe. There also is a safety issue when driving the LS in the conventional way, since the crank would be flailing away -- unless it's been removed or decoupled from the LS. Mine is just removable.

The other thing is that you don't want to disconnect the half nuts during this procedure, otherwise you'd lose sync between the spindle and LS. Very similar to cutting metric threads on an imperial machine.