PDA

View Full Version : Unusual thread cutting feed system- circa 1888



S_J_H
10-09-2010, 11:37 PM
I have a #3 Stark lathe I'll be restoring soon.
(http://www.lathes.co.uk/stark/) And so I have been researching the various thread cutting systems that were used on this class of small precision bench lathes of yesteryear.They seem to have been very popular and were offered by Stark, Hardinge / Cataract, Rivett, Ames, Pratt&Whitney and several others.

Bear with me please..
2 types of thread cutting systems seem to have been used on this type of lathe. A change gear system that drove the top slide via a universal shaft and the "Chase" system that used a master thread, follower and tool holder which traversed along a rear mounted shaft assembly.
Here are a links with descriptions of the chase threading system if interested- http://www.lathes.co.uk/stark/page12.html
And some more good descriptions here- http://www.georgesbasement.com/FoxLathes.htm
With the chase system, a change gear setup might also be used with the master thread to allow several different thread pitches to be cut with 1 master.
Example of chase threading with change gears shown on this very fine looking Stark lathe (not sure who this belongs to?)
http://i109.photobucket.com/albums/n48/S_J_H/misc/StarkTC013-1.jpg

Which finally leads me to the subject of this thread!,
A 122 year old patent that shows a mechanical variable feed for use with the chase screw cutting system.
The patent with explanation of the system-
http://www.google.com/patents?id=PedYAAAAEBAJ&zoom=4&pg=PA1#v=onepage&q&f=false

Patent images showing the mechanically variable feed system used with the master thread-
http://i109.photobucket.com/albums/n48/S_J_H/misc/variablefeed1888patent.png
http://i109.photobucket.com/albums/n48/S_J_H/misc/variablefeed1888patent2.png

According to the patent info, the system will allow any pitch less then the master to be cut. So in theory a 10tpi master could cut 10tpi and all pitches below by the setting of the lever position.
I can envision how it all works and the mechanics seem sound, but I'm no engineer. Anybody have any thoughts?
Does anybody have knowledge of the variable feed system shown in the above patent ever being used on any lathe or other type of machine?

Steve

becksmachine
10-10-2010, 12:06 AM
Not that I am in any condition to judge at the moment, but one question jumps out at me. :D
Why would you be limited to pitches "below" the pitch of the master? Seems to me the fulcrum could be set to allow pitches both greater AND less than the master??

Dave

2ManyHobbies
10-10-2010, 12:06 AM
That's pretty neat. If I had to guess, the leadscrew and change gears method won based on fewer moving parts and repeatability (different operators on different machines can still use the same table with change gears). Wear at the rack would be accelerated over that of a leadscrew and halfnut but that could be countered with an outboard halfnut instead of a rack. The pivot point L could also be set up on a threaded rod like a cross-slide (or a stepper for CNC).

The limit though would be the total distance that could be threaded without dialing back in. With 3" of a 10 tpi master you'd wind up with 1.25" at 24 tpi and less than an inch at 32 tpi. Given not everybody needs to single point 12" at 24 tpi, but it would take considerable time to do something like recreate an acme rod that matches the distance between centers.

It is an interesting evolutionary find in machinery fossil records.

macona
10-10-2010, 12:08 AM
Thats basically how threading is done on some of the hardinge chuckers.

lazlo
10-10-2010, 12:22 AM
Thats basically how threading is done on some of the hardinge chuckers.

Hardinge Cataract, circa 1914:

http://www.lathes.co.uk/cataract/img91.gif

Hardinge HC, circa late 1940's:

http://www.lathes.co.uk/hardinge/img17.gif

Paul Alciatore
10-10-2010, 01:01 AM
The Unimat DB200 also used a thread chaser:

http://www.lathes.co.uk/unimat/page2.html

logic
10-10-2010, 01:22 AM
Here's a G. Boley watchmaker lathe that I acquired a couple years ago that uses the thread chaser method. I took the photos when I received it and then broke it down in parts for restoration. I then had to move so the parts have been sitting in boxes since but I really need to get this put back together.

Patrick

http://www.viperlogic.com/photos/boley_lathe/lathe_front.jpg

http://www.viperlogic.com/photos/boley_lathe/lathe_gears.jpg

S_J_H
10-10-2010, 01:27 AM
Not that I am in any condition to judge at the moment, but one question jumps out at me.
Why would you be limited to pitches "below" the pitch of the master? Seems to me the fulcrum could be set to allow pitches both greater AND less than the master??


That's pretty neat. If I had to guess, the leadscrew and change gears method won based on fewer moving parts and repeatability (different operators on different machines can still use the same table with change gears). Wear at the rack would be accelerated over that of a leadscrew and halfnut but that could be countered with an outboard halfnut instead of a rack. The pivot point L could also be set up on a threaded rod like a cross-slide (or a stepper for CNC).

The limit though would be the total distance that could be threaded without dialing back in. With 3" of a 10 tpi master you'd wind up with 1.25" at 24 tpi and less than an inch at 32 tpi. Given not everybody needs to single point 12" at 24 tpi, but it would take considerable time to do something like recreate an acme rod that matches the distance between centers.

This is how I read it and it seems to make sense the way the system is shown in the pics :confused:.


Assuming that the parts are so set that the sectional nut occupies its position of midstroke, and that the sliding rod is consequently
at right angles to the nut and chasing-bar, and that the pivot L is so set upon the T
shaped arm that the distances from L to J and from L to H are equal, then ifc will appear that for each unit of distance moved by the sectional nut in a path parallel to the axis of the lathe-spindle the pivot L, and consequently the chasing-bar, will have moved one half such distance, and that this proportion of movement will hold good throughout the travel of the sectional nut in either direction. Under such circumstances the chasing-tool controlled by the chasing-bar will have been able to chase a thread whose pitch would be one-half the pitch of the hob.

It will now be readily understood that the pivot L may be set at various points along the T-shaped arm, and thus permit of the chasing of threads of various pitches by the use of a single hob, it being understood, of course, that as the pivot L is an intermediate pivot, all threads chased will be of less pitch than the pitch of the hob.
So which is it guys?

The half nut is much longer than the master thread and slides in the stationary T holder. Unlike the short halfnut that traverses a long master thread as used in the change gear system. So it seems to me that the length of thread that can be cut should be equal to the change gear method.
What I like about this system (if it works) is that it should operate very smoothly and quietly with no gears anywhere on the lathe.
Unless I can find a nice original example of a chase threading attachment for my Stark ( highly doubtful) I am going to shop build one for a nice project and give the lathe some more usefulness.
This looks like it would be pretty nifty to build and I don't have to machine a slew of change gears then.

Robert, The Hardinge shown at the top looks to have a real nice chase type system, the bottom looks electronically controlled.
Do either of those use anything that resembles the system in the patent above?
Steve

whitis
10-10-2010, 02:46 AM
Including yours, I have seen pictures of 5 different types of lever based screw cutting lathes. I believe at least some of them were actually used. Given the date on your patent, though, this one may not have been used to any significant extent because it was a step backwards. My treddle/lineshaft (later electric) lathe was apparently built the same year as that patent and it can cut 36" long screws with a change gear system and has a spline for power feed on both axes; they had a previous model that did the power feeds with a worm gear instead of the spline. And 6 years earlier, Hendey shipped lathes with the norton quick change gear box. Other than some changes in spindle noses like 5C collet holders and cam-lock spindles and the use of cone head drives instead of v-belts/gears and ball bearings the manual engine lathe as we know it today existed then.

The history of screw cutting lathes went something like this:

1480 or earlier? master screw mounted on the same axis as the work and as they turned the master screw advanced the work past the cutter. A different master screw needed for each pitch.
1741? lever action multiplies master screw pitch you could cut a wide range of pitches with a single master but your pitch isn't exact
master screw mounted parallel to the work and moves a tool holder. Required a different master screw for each pitch.
1770 Ramsden screw cutting only lathe, rack and pinion drives cutter
1810, maudsley lathe, lead screw drives the spindle, change gears
quick change gear boxes
NC/CNC

But the actual history seems to be more convoluted than that. Every time I look up the history of screw cutting, I get a different answer. :)

Somewhere in there, there was also the chain lathe that had gears driving a chain which moved the spindle.
http://chestofbooks.com/home-improvement/woodworking/American-Lathe-Practice/The-Development-Of-The-Lathe-since-Screw-Threads-Part-4.html

in 1933 someone patented a sine bar mechanism for changing the pitch ratio between master screw and work. Possibly for fine tuning?

Here is one history, with a number of illustrations, including one lever based system.
http://homepages.ihug.com.au/~dispater/turning.htm

An Encyclopedia of the History of Technology shows one 18th century watchmaker's fusee in which the spindle includes a master lead screw and one end of a lever follows a nut on that screw. A sliding bar connects to that lever at an adjustable position and moves the cutter. Your patent is similar but changes the nut to a rack and makes it longer than the screw.

Levers based system have infinitely variable pitches which can be both a blessing and a curse; more the latter in routine use. They can lead to out of spec parts. In the days before widespread dimensional standardization and interchangeable parts, this could be useful to match parts produced on other machines but also would tend to contribute to non-interchangable parts. However, they could be used with a lot of fuss to correct pitch error on a master screw. Change gears can also be used for fairly fine pitch adjustments by inserting two gears with almost the same number of teeth.

The lever systems appear to have poorer rigidity and don't scale well to producing long screws. Depending on the pitch, the length of screw you can cut relative to the master screw length varies.

The one you have shown is different in that it uses a short screw and a long rack such that the rack is really your master screw.

If the fulcrum in your patent had been moved to the middle of the lever arm, it could have produced pitches above and below the master. If you did that and also made the sliding rod "D" in the first picture be the master lead screw of a change gear system, and geared down the worm from the spindle, then your lever could apply +/- variable fine tuning to thread pitches.

The lever system seems to have been more used by watchmakers.

Trivia: Whitworth invented the half nut and possibly the combined longitudinal/transferse power feed:
http://www.whitworthsociety.org/history.php?page=2

whitis
10-10-2010, 02:52 AM
The half nut is much longer than the master thread and slides in the stationary T holder. Unlike the short halfnut that traverses a long master thread as used in the change gear system. So it seems to me that the length of thread that can be cut should be equal to the change gear method.

Note even close. First, the rack bumps into the cone pulleys. Second, while you can make the rack any length you want if you overcome the previous problem, provide supports for it along its length, and allow many extra feet on the end of the lathe for the rack to protrude, the lever is short and limits the total length of the thread you can cut. And it will get very ungainly and lack rigidity if you try to also make the lever many feet long.

S_J_H
10-10-2010, 04:25 AM
Whitis, Great link! but are we on the same page?

When I speak of change gears in this thread, it refers to the small split bed lathes such as this-
http://www.lathes.co.uk/stark/img117.jpg

This type of lathe did not have a rack, lead screw or carriage and apron . They coexisted with the more common engine lathe. These lathes were offered by some lathe builders into the 50's or even newer
This family of lathe was built by many of the best known high quality lathe builders and all similar in construction.

Screw cutting was an accessory for this type of lathe. They were not used for cutting long threads.
As shown in the patent, this would be a basic layout and certainly could be built with several options as you mentioned.
The other threading system used for this type of lathe used change gears and a universal shaft driving the topslide -
http://i109.photobucket.com/albums/n48/S_J_H/misc/rivettlathe.gif

Steve

lazlo
10-10-2010, 11:23 AM
Trivia: Whitworth invented the half nut and possibly the combined longitudinal/transferse power feed:
http://www.whitworthsociety.org/history.php?page=2

I'm a little dubious about that. Maudslay's screwcutting lathes had half nuts in the latter half of the 1700's. In fact, thread pitch error adjusters were pretty common for cutting precision leadscrews in the late 1700's:

http://i164.photobucket.com/albums/u15/rtgeorge_album/threadpitchcorrection.png

http://i164.photobucket.com/albums/u15/rtgeorge_album/threadpitchcorrection2.png

Whitworth was a Journeyman in Maudslay's shop, and the text says that Whitworth invented the "Clasp nut" -- I'd imagine his patent was the application of the half nut to selectively drive the carriage.

wierdscience
10-10-2010, 11:54 AM
This type of lathe did not have a rack, lead screw or carriage and apron . They coexisted with the more common engine lathe. These lathes were offered by some lathe builders into the 50's or even newer
This family of lathe was built by many of the best known high quality lathe builders and all similar in construction.

Screw cutting was an accessory for this type of lathe. They were not used for cutting long threads.
As shown in the patent, this would be a basic layout and certainly could be built with several options as you mentioned.
The other threading system used for this type of lathe used change gears and a universal shaft driving the topslide -

Steve

Steve,some of the answer could be tradition.As an example several engine lathe mfgs still were offering flat belt lathes well into the 1940's,even though gear heads were all the rage.I count two factors for this,one the presence of line shaft shops still in service and two factory men who grew up on belt lathes and weren't ready for a change.

Another may have been the class of work being done,Watch maker and instrument maker lathes wouldn't necessarily need screw generating capability,since there aren't many long threads to be seen in either skill.

philbur
10-10-2010, 12:13 PM
All other things being equal a belt drive will produce a far better surface finish than a comparable geared head lathe, even today.

Phil:)


Steve,some of the answer could be tradition.As an example several engine lathe mfgs still were offering flat belt lathes well into the 1940's,even though gear heads were all the rage.I count two factors for this,one the presence of line shaft shops still in service and two factory men who grew up on belt lathes and weren't ready for a change.

becksmachine
10-10-2010, 01:52 PM
This is how I read it and it seems to make sense the way the system is shown in the pics :confused:.

So which is it guys?
Steve

See, I told you I was in no condition to judge. :rolleyes:

I missed the " pivot L is an INTERMEDIATE pivot" part, which would certainly preclude cutting any pitches coarser than the master. I had the anchor and the sliding bar mixed up.

Dave

wierdscience
10-10-2010, 10:37 PM
All other things being equal a belt drive will produce a far better surface finish than a comparable geared head lathe, even today.

Phil:)

Funny,I've spent a lot of time running both and haven't found that.Once power is transmitted through a gear train to reach the leadscrew or feed rod and then through a rack and pinion set all bets are off as to finish.

S_J_H
10-11-2010, 12:57 AM
A belt drive should provide a smoother finish all things being equal. But you bring up a good point Wierd.

Back to the topic, I was just looking for info related to the old split bed precision lathes and the screw cutting systems that were offered for it.
The way I understand it, these lathes were not really a part of the evolution of screw cutting, it was already in use and many of the lathe builders offering this type of lathe also built engine lathes. They had a different purpose and were built very similar to the smaller watch makers lathe.
Some quotes taken from http://www.lathes.co.uk may help to explain things better than I can.

The Stark Company, originally of 222 Moody Street, Waltham, New England, U.S.A., bear an important name in the history of American close-tolerance engineering, for the firm were almost certainly the originators of the Precision Bench Lathe - as eventually made by a variety of firms including: Derbyshire, Levin, Bottum, American Watch Tool Company, B.C.Ames, Hjorth, Potter, Pratt & Whitney, Rivett, Wade, Waltham Machine Works, Wade, Pratt & Whitney, Rivett, Cataract, Hardinge, Elgin,, Remington, Sloan & Chace, and (though now very rare) Ballou & Whitcombe, Frederick Pearce, Sawyer Watch Tool Co., Engineering Appliances and Fenn-Sadler.
Stark's claim as originators of the type was bold and unequivocal (and printed on all their sales catalogs), with the first examples being built by John Stark personally in 1862, well before any of his competitors - who were also mostly from the Waltham area. A special section of the Archive, assembled from literature of the 1870s, has been devoted to these very early and important Stark machines.

Almost as soon as the first Precision bench lathe had first appeared in 1862, there was a demand for some form of screwcutting facility and before long most manufactures of this class of lathe were offering the facility in one (or both) of two ways: either a long-travel top slide through a universally jointed shaft from a train of changewheels driven from the headstock spindle, or by the "Chase" system, using a "Master Thread" and "Follower". The latter system was developed by Joseph Nason of New York who obtained US Patent No. 10,383 on January 3, 1854 for an "arrangement for cutting screws in lathes."
I'm going to build a version of the feed system shown in the patent for my Stark.
Might even use cnc to cut the master threads, sort of a full circle type trip.:D
Steve

BillTodd
10-11-2010, 06:24 AM
I can envision how it all works and the mechanics seem sound, but I'm no engineer. Anybody have any thoughts?


Given that common/practical threads tend to have a limited number of turns rather than length. Then the limits of this lever system seem fine. I suspect that the lever does not give a perfectly linear conversion if analysed geometrically but it will be perfectly OK in practice.

Sound like a good project, please post pictures of your build :)


Does anybody have knowledge of the variable feed system shown in the above patent ever being used on any lathe or other type of machine?

In the last year or so a new machine was launched claiming to be able to cut any pitch thread within its limits.

http://www.youtube.com/watch?v=pJRCN9G-jBk

I suspect that it uses a lever arrangement to divide the hand turned lead screw into the movement of the rotary cutter. I asked the inventor on a forum, but he wisely declined to answer, saying only that the system did not use electronics.

Bill