mini-lathe worm cutting

As you cut deeper with a form tool, the area of the tool edge trying to cut metal goes up very fast. This puts an enormous load on the gears in the machine.

You have to rough it out by cutting one flank of the worm to partial depth, and then cutting the other flank to the same depth. Hog out any metal in between. Then repeat a little deeper. This way you have less of the tool cutting at any pass.

Once it is roughed out, do a few finishing cuts with the tool in full contact, but at this stage your infeed should be in the one thou range or less.

Paul K. Dickman

Mike,

Cutting 6.4 TPI with a 16 TPI lead screw means that the lead screw is geared up from the lathe spindle, ie. the lead screw turns at a higher RPM than the spindle during the threading operation.

This gearing-up puts large stresses into the change gears of the threading gear train, and light-duty gears may very well fail.

What's sometimes done is drive the lead screw directly with a gear motor at a speed that produces the desired rpm at the lathe spindle... the reverse of the normal function. This has the advantage that the gear train is used as a reducing gear train, the way it was meant to be used.

Absent a suitable gear motor a hand crank could be fitted to the lead screw directly and used as the driving force.

I also built a dividing head to suit my mill, and because I have a threading gear box I decided to use 8 TPI (the coarsest thread it will cut) and produce a worm gear to match. But this is another (humorous) story.

Let us know what you decide on and how it goes.

Wolfgang

On Wed, 14 Jan 2009 23:30:36 +0000, Michael Gray wrote in rec.crafts.metalworking and uk.rec.models.engineering :

Also consider A B C D = 55 35 80 50, for ratio 6.364.

As wfhabicher pointed out, driving the leadscrew rather than the spindle will have a more favorable mechanical advantage. The web page

about 2/3 along has a section labeled ULTRA-FINE POWER FEED that shows one way of attaching a motor to the leadscrew, if you expect to drive it very often. Just adapt the idea to a removable handle.

is a C program you can compile on linux with gcc and is the program I used to find the 6.364 combination mentioned above. The command "./gears 6.36-6.38" displays "6.364 55 35 80 50". If you still have an intact

30-tooth gear, run the program via "./gears +30 6.36-6.38" or change line 51 ("int teeth[N1] = 20,20,35,40,45,50,55,57,60,65,80,80") to reflect the tooth counts in your set of change gears. Apparently the Cummins model has some that the HF model doesn't or didn't.

I have just cut a small batch of the worms for a George Thomas Myford Super 7 dividing head (.157" pitch) which is the same as yours (almost) I cut them on a Hardinge HLV at about 200 RPM using the retracting topslide to pull the tool back after each pass. The first cuts were .005" up to .020" deep then ..002" to about .035" deep then .

0005" until the end, each cut was passed down 3 times to remove spring in the tool. The whole job takes 1 hour and is fraut with problems due to the depth of cut (about .150" in my case). I set the top slide to 2 degrees less than the PA of the worm so it shaved on one side, cut on the end and the other side, high speed tools with suitable clearance and they still dig in, using anything less than top class gear you will never make a good one (I had 5 failures before I was getting it OK) If you need futher help PM me Peter

Or almost a 4 mm pitch (3.988 mm)

Nick

fx?uUÑn7|¿¯?¸6 ÉV??? Ò6H?ÔEQ$èã?\Ýâ??ç?òõåIN´?îxËÝ?ÙÕ] ß8Ðú?®o¯®|}5ôú?}Å{Itæ?~~«Õ??¬?µåM)¹m^ºWõõ?3?§·?Ço//??ïóQä2?H¡g«ëg?î?nÝ%bßÇ\?é)õ= ® ÃùõêùÍ Þ/þ÷¦¶ñè`¹|ôúW'äãd{ùó?w?EÛ])Ãíåå4M«Û}×8üÝ¡ ­¢×)ÞMévâ]£¥¸¥

Glad to see you drop in Nick.

Wes

Here is my AWK program for printing exhaustive change gear possibilities:

You'll have to add the mini-lathe configuration.

Here is the discussion thread:

Thanks to all and sundry for their help and advice. I have a 110vac gear motor that I'll put on the leadscrew and have a go with that, using my favourite old light dimmer as a speed controller. Pics will appear later at a Photobucket address. again, many thanks, Mike in BC