What's new
What's new

Increasing leadscrew accuracy

wfrancis

Aluminum
Joined
Apr 18, 2012
Location
San Francisco, USA
I'm interested in the idea that a lathe with leadscrew can machine a new leadscrew which is more precise (yes, I know that's a problematic word) than the one in the machine when change gears involved.

I know this is a potentially vast subject so I'm going to narrow the scope considerably - I'm interested in the period when the first screw cutting lathes emerged and had hand filed (or via some other primitive method) leadscrews. I'm not (in this query) interested in the 100 methods the Moore company had in making their leadscrews more precise.

I'm including a link below which is one place I have read such a claim. Thiout's lathe is another one which is claimed could make more precise screws than the one in the machine (though via an interesting reducing linkage, not gears).

However, the only way I've been able to noodle that this is possible is if the entire initial leadscrew was uniform, but the pitch was slightly wrong, say 8.1 TPI instead of the desired 8 TPI so, via gearing, you could machine a new leadscrew that was the desired 8 TPI (satisfying "more precise").

While I can't find a source for it now, I believe I've also read the reduction also would average out errors in a non-uniform leadscrew, but to my way of thinking it would reduce the error per turn of thread, but cumulatively the error would still be there.

If anyone has knowledge of this second case and/or relevant literature I'd be very happy to review it.


The Most Powerful Idea in the World: A Story of Steam, Industry, and Invention - William Rosen - Google Books
 
This is a great topic for the "History" side of this forum!

A similar concept is used to create accurate dividing plates.

A forum member, Dr. George Langford, knows how to create "original" dividing plates with any number of holes using this concept. Perhaps he will chime in on lead screws.

John Ruth
 
Are you familiar with the technique of leadscrew pitch compensation? (Google has elected to not include your linked page in what it will show me, so I can't read that text.) In a nutshell, the leadscrew nut is captured on the carriage axially as normal, but is also allowed to rotate under the control of a linear cam running along the ways. The cam is customized for that specific machine, to compensate for minor pitch irregularities in the leadscrew, using precision measurement techniques to determine distance actually traveled for a given leadscrew rotation.

This was a fairly common precision "master maker" technique once leadscrews could be made accurately enough that the remaining pitch irregularities were equivalent to a small fraction (1/30?) of a revolution. It is not effective at compensating for a systematic (global) pitch error over long distances.
 
Are you familiar with the technique of leadscrew pitch compensation? (Google has elected to not include your linked page in what it will show me, so I can't read that text.) In a nutshell, the leadscrew nut is captured on the carriage axially as normal, but is also allowed to rotate under the control of a linear cam running along the ways. The cam is customized for that specific machine, to compensate for minor pitch irregularities in the leadscrew, using precision measurement techniques to determine distance actually traveled for a given leadscrew rotation.

This was a fairly common precision "master maker" technique once leadscrews could be made accurately enough that the remaining pitch irregularities were equivalent to a small fraction (1/30?) of a revolution. It is not effective at compensating for a systematic (global) pitch error over long distances.

My Hauser Jig bore has this.
 
Moore's book has a whole chapter on the master lead screw:

7. MOORE MASTER LEAD SCREW ................ 187
Material of the Lead Screw . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Thread-Grinding of the Lead Screw. . . . . . . . . . . . . . . . . . 189
Lapping of the Thread . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Lead Screw Nut ..................................... 191
Lineup of the Lead Screw . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Step Gage Calibration of Lead Screw . . . . . . . . . . . . . . . . 195
Periodic Error ....................................... 198
Moore Lapping Specialists . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

dee
;-D
 
Somewhere in one of my books on early machining there is a description of a method. The customer would supply a screw that was to be improved. The machinist would make a follower that looked like a saw to fit the screw. The idea was to average the thread pitch. The sample screw would be mounted on the lathe, the follower on the carriage, and used to make a new screw. Then the new screw would be used as the lead screw to cut another one. Reportedly there would be no improvement after the third iteration.

Bill
 
Template.jpg

To expand a little on what sfriedberg said, this photo hints at the correction method. A follower rides on the template to somehow move the tool fore and aft to correct irregular pitch errors. It seems to be on a special pitch correcting tool holder. The source doesn’t help here (Lathes, Screw Machines and Turning Mills by Thomas R Shaw, 1903).

It obviously begs the question of how did they know how to shape the template. Shaw does help here, referring to someone in London in the 1870s who had to make a very accurate screw for dividing glass diffraction gratings. He obtained a very accurate leadscrew, and used it to move a travelling microscope for viewing an accurate scale, and recording the errors. These were plotted to produce a diagram from which to make the correction template. Yes, this does beg other questions about the accuracy of the scale, etc.

In A History of Machine Tools 1700-1910 W. Steeds writes that in 1826 Bryan Donkin was using a wavy correction bar to correct irregular pitch errors. Another source* says that Donkin was cutting screws with errors never greater than 1/30,000"

* Donkin was a truly remarkable engineer, who deserves to be better known. The shortcoming has recently been addressed in Bryan Donkin – The very Civil Engineer 1768-1855 by M Greenland & R Day, 2016.
 
The method I have read about involved making a lead screw, then a long lap, then using the lap to average out the errors in the lead screw you just made.

Next step, measure and then compensate for any measured long pitch errors with gears etc and use the lead screw from above to make another more accurate one.

Then repeat with lapping and measuring the newly produced lead screw.

Then use that lead screw with compensation to make an even better one.

Each process cycle produces a better lead screw.

No one seems to address how you measure with increasing accuracy.

Lee
 
No one seems to address how you measure with increasing accuracy.

Early length standards were usually line standards, so I imagine you could get good accuracy with a microscope mounted to the carriage looking down on a length standard.
 
No one seems to address how you measure with increasing accuracy.
There's a huge literature on this subject, going back 150 years or more. Just because they don't teach it in school doesn't mean people weren't talking about it, reporting on techniques, criticizing others' techniques, and continually advancing the state of the art.
 
For those unable to read it, here's a screenshot of the passage in question


I'm uploading a large-ish file but I think it's showing small here. Is there a way to fix that? Or do I have to link from off-site?
 

Attachments

  • Screen Shot 2018-02-25 at 2.17.52 PM.jpg
    Screen Shot 2018-02-25 at 2.17.52 PM.jpg
    19.4 KB · Views: 128
  • Screen Shot 2018-02-25 at 2.24.42 PM.jpg
    Screen Shot 2018-02-25 at 2.24.42 PM.jpg
    14 KB · Views: 105
Compensated Lead Screw

This is probably not the information you are looking for but here is a slide from a comparator that has a compensated leadscrew. This is the slide from a Gaertner Comparator. I don't know how old it is.

The third photo shows the compensation bar and the last photo shows the arm attached to the nut. The compensation bar pivots in the center and is adjusted by the screws on the ends. The nut is not fixed to the carriage but is trapped by screws on each end. The nut is free to float and the follower on the arm allows the nut to rotate. I really don't know how they were calibrated. They probably started with a pretty accurate screw. Some of the literature shows checking with microscopes. This bar is damaged. The bend in it looks like it was hit or dropped.

Here are some references that are on Google Books:

Thread-cutting Methods: A Treatise on the Operation and Use of Various Tools ... - Franklin Day Jones - Google Books

American Machinist - Google Books

For more information about precision screws, this has probably more than anyone could want. Pretty interesting:

The Precision Screw in Scientific Instruments of the 17th-19th Centuries by Randall C. Brooks. Here is the download link:

Leicester Research Archive: The Precision Screw in Scientific Instruments of the 17th - 19th Centuries: With Particular Reference to Astronomical, Nautical and Surveying Instruments

At one time Lindsey Publications had a book about Leadscrews, Gears, and Pantographs. You might be able to find one of those used.

The Moore book has some background material on early screw making and dividing. It available on Archive.org.

Foundations Of Mechanical Accuracy : Free Download & Streaming : Internet Archive

This is an interesting subject that most people take for granted these days.

Terry

DSC_5970.jpgDSC_5966.jpgDSC_5961.jpgDSC_5964.jpg
 








 
Back
Top