Hi Bazyle ,

There was no consistently accurate method of setting out large (more than a few inches) linear distances in an ordinary production shop before about 1920 . Around this time the high accuracy jig borer started to get into general use . One of the first companies to manufacture these was Societe Genevoise .

Jig borers initially used two methods for high accuracy work :

(1) End on end stacks of slip gauges set between precision stops . Very accurate but tedious since a new stack had to be assembled and set for every move .

Some more inspired machinists no doubt made themselves a selection of length bars equivalent to slip gauge stacks needed for more commonly used distances .

(2) A precision screw thread and basically lead screw dials . Very accurate screw threads were available long before the 1920's but don't seem to have been applied to anything except ordinary lathes much earlier .

As time moved on other methods appeared including long Vernier scales and precision glass scales both viewed through a built in light and magnifier system .

Nothing much happened then until Heidenhein introduced an early version of electronic DRO in the early 1950's .

The above applies to ordinary industrial use . For scientific work optical setting out methods were in use very early on and these were developed progressively into modern interferometric and laser systems .

The clock making industries had mini versions of jig borers a good bit earlier than general industry . These were mostly screw based and their primary use was 'spotting' the centres for clock plates .

Notes :

(1) When designing accurate machines there is no way of 'beating the system' . Somewhere in the machine there has to be a fundamental length reference . This can be a screw , scale , assemblage of slips , a laser system or one of many other things . However once a reference system has been decided on there is no way of making it 'more' accurate by adding complexity .

(2) Very few machines have ever used a rack as the fundamental reference . Probably because it is much more difficult to make an accurate long rack than an accurate long screw .

There's a lot more to this .

Regards ,

Michael Williams .

Edited By MICHAEL WILLIAMS on 31/03/2012 15:35:06

Hi S,M ,

The Merton nut was used in the first attempts to make graticules for scientific work and measuring scales .

Making the earliest accurate screws was quite a task . People like Maudslay cut their initial master screws literally by hand with endless testing and fine corrections needed . Once the best screw had been made by hand there was a painstaking method of making a better copy basically by screwcutting but with an axial micrometer like adjustment on the cutting tool so that by cutting very slowly and making minute adjustments all the time better pitch accuracy was achieved .

Regards ,

Michael Williams .

I think it's some where in vol 1 ME (1898), an artical about cutting threads free hand with chasers.

Ian S C

But how did he know he had pitch errors unless he had something accurate as a reference?

Ian

Bazyle,

http://www.practicalmachinist.com/vb/general/expiriments-leadscrew-lapping-218177/

is worth reading … especially Posts #19 and #20

MichaelG.

Bazyle ,

Straight answer is probably tape measure .

Apart from some woodworking machines few industrial tools have ever used a rack as a means of setting out measured lengths . In the period you mention accurate working in a run of the mill workshop was limited to about 18 inches at most – anything longer was done with ruler and tape .

Interesting bit of confusion exists in descriptions of machines of your stated period and earlier . Many authors in the past have mixed up accuracy of setting out measured distances with accuracy of straightness , flatness or roundness .

Achieving high accuracy in (eg) flatness was more than a hundred years ahead of achieving the same accuracy in measured lengths . Maudslays hand scraped surface plate as a reference for flatness would have been accurate to one micron or better but this level of accuracy in measured lengths could not be achieved as a regular thing until the 1980's . Even the early jig borers were only unconditionally accurate to about ten microns for measured lengths .

Ian ,

But how did he know he had pitch errors unless he had something accurate as a reference?

Just so – all the means mentioned for improving accuracy are really just calibrating the basic reference against an external master reference and providing a built in means of correcting minor errors . Basically you have just copied the master reference onto the machine reference – but whatever you end up with is still the fundamental length reference for the machine .

Michael Williams .

Bazyle,

"hackuracy" is a great term … it made me smile.

I think your local works would have either relied upon a leadscrew, or spaced it out with dividers.

For the benefit of future readers: It's worth mentioning that a Gaertner-style "effective pitch adjuster" could be used [for example] to convert a 26tpi leadscrew to 1mm pitch [or vice versa] … very handy for those occasional jobs.

MichaelG.

Thanks Michael,

I found a facsimile reprint sold by Amazon for a good price and have ordered a copy at quite a good price after reading the restricted pages. Looks to be quite a fascinating book.

Best regards

Terry

Hi Terry ,

You are quite right. Maudslay built a screw generating machine that held a blade at the required helix angle and cut into a soft metal bar, this was followed by a cutting tool which actally cut the reference thread in the bar.

I've read abut this in several different places but have always concluded that there is either a lot more to it than described or that this was just an early experiment . If you think about it the process as described would be generating almost random pitches after a few turns .

I think it was the ancient Greeks that used something similar to generate the helix for their screw pumps but at a very much lower accuracy level .

Most early feed screws were cut by hand and these became progressively more accurate as time went on .

Hand cut feed screws survived in use right up to quite recently . Stothert and Pitt in Bath had a giant size planing machine with hand cut feed screws where the original fine lines of centre pop marks marking the helix were still visible .

Regards ,

Michael Williams .

Why is everyone talking about the early 19th century when the question was about the early 20th century?

Terry,

The full text of Jesse Ramsden's paper is available here: http:// http://www.fer3.com/arc/imgx/Text.pdf

and an excellent commentary, here: http://www.fer3.com/arc/imgx/Commentary-on-Jesse-Ramsdens.pdf

Both thanks to the expert Mr Morris

http://sextantbook.com/category/chasing-tenths-of-an-arcminute/

MichaelG.

John,

Yes, you are right … the problems were substantially solved.

What is interesting, and a little depressing, is how much of that knowledge has since been forgotten or ignored. That's why, every once in a while, we need to dig through the history.

MichaelG.