Thread milling vs single point threading

[Conrad_R_Hoffman]

I checked and the book I found useful was Precision Machine Design by Slocum. It's a full size brown book about an inch thick. If you can find an affordable copy, it's quite worth it if you enjoy this sort of thing. I've seen the Vogel book sell for as much as $1000 because it's hard to come by, but there's an affordable copy on Abe Books right now; I doubt it lasts 5 minutes.

I have a VFD on my ancient surface grinder and find it very useful for changing the apparent hardness of the wheel!

[Wolfgang]

Methinks most of you know that on modern high-strength fasteners the threads are ROLLED.

And this is performed by two methods I am aware of:

1) With 2 or 3 rollers the work is squeezed and the threads are formed. Bicycle spoke threads are a case in point, some being formed with a very simple tool.

2) The work is rolled between 2 flat dies which proceed across the work in appropriate relationship with the axial feed of the work piece.

With mass production using process 1) the roll dies are quite large. And with production using process 2) the rolls have an infinite radius.

Question: How do these methods affect the shape of the thread flanks? Provided of course that the dies' threads have straight sides.

I suppose it is possible that the dies' shape may be "tweaked" to ensure straight thread flanks, but I really do not know much about this process.

Any enlightenment would be appreciated. w

[Conrad_R_Hoffman]

Not a clue how the geometry comes out, but the big Kinefac machines are impressive- http://www.kinefac.com/mc-300.html Note the size of the guy next to the machine, though they make little ones too. The process goes all the way down to watch sized parts- http://www.habegger-sa.com/english/produits_en.htm (click on thread rolling dies to the left)

I'm guessing the results are the same from an infinite diameter grinding wheel as an infinite diameter thread roll, save for material spring-back. I'd also guess they can grind the rolls to give whatever geometry thread is desired.

The question to me is, what's the desirable flank shape and does it match the nut? I work with ultrasonic leadscrews where wear is a factor and good contact is desirable. OTOH, our parts are so small that controlling geometry to any great degree is near to impossible.

[pete]

Ron's original question has gotten way off track so I hope he doesn't mind where this has gone.

I also suspect that theoretical "perfect" thread profile only becomes important enough on ultra high accuracy items. Micrometer threads (maybe) high end optics and lasers then almost for sure. For most common day to day items and thread fits it can probably be ignored, or judged still good enough to work? Rolled threads are known to be stronger since it's a lot like a cold forging process and the method is obviously a lot quicker. Maybe one of the built in side benefits of today's use of ball screws instead of ACME or square threads is it's likely easier to produce a more perfect and continuous helical radius than one with straight thread flanks around the screws helical path? Those perfect thread flanks are still only one part of the problem though. Moore Tools seemed to be much more concerned with advancing or retarding screw pitch inaccuracy's along the screw or within the nut that seemed to be unavoidable with any manufacturing method they tried. Moore went so far as lapping there screws and nuts to average out those errors.

In the past I've checked a few websites of the bigger names who produce ACME, ball screws and nuts. Nook, SKF etc. You start getting into some really big dollars and actually harder to find guarantees of less than .001"-.002" accuracy's per foot of screw length. A few months ago I ran across this video, https://www.youtube.com/watch?v=AQ5PnydOBtY&t=79s So screw straightness is another problem. There's also quite a bit of fairly easy to find calculations and engineering data for what's called screw whip if the rpms get high enough.