Boxford metric lead screw fitted to imperial lathe?

.

An interesting counter-argument is, I think, presented here: **LINK**

http://www.galleyrack.com/images/artifice/machine-shop/ornamental-turning/literature/english-19c-other/amateur-mechanics-v01n01-1883-01-google-nosource-pp-017-018-img-028-029-ornamental-lathe-screw-threads.pdf

MichaelG.

Ha ha, presented 136 years ago, so perhaps not quite so pertinent to the 21st century (and this thread)?

I did not bother to study it. If I came across an ‘orphan’ thread and needed to replicate it, I would either do that or alter it to a different thread form - whichever was most practical.

Interesting as all this has become, it is perhaps easy to overlook the fact that it only came about because of the rogue fitting of a metric leadscrew to the imperial screwcutting gearbox on an imperial lathe, giving rise to the work-arounds that have been proposed to save the owner the expense and hassle of re-equipping the lathe

What might be instructive now would be to see if similar solutions can be found for a case where a metric gearbox has been mated to an imperial leadscrew.

I feel a full bottle of gin might become an important aid in bringing those answers to an eager public!!

Brian

.

What a hoot! Holtzapffel threads appear to be as irrationally chosen as MT taper angles. But Holtzapffel was a clever bloke. Can anyone explain why he might have chosen such strange TPI values? For example:

1" - 6.58tpi
3/4" - 9.45tpi
1/2" - 13.09tpi
1/4" - 25.71tpi

Seems back in 1883 young thrusters were having bother selling revolutionary new ideas like aliquot pitches to the Old Farts of the day: 'Now that these screws are thus exhibited, they are seen to be a mere relic of rule-and-thumb mechanics which have been kept alive for trade purposes, as if the screw-cutting lathe had never been invented.'

Perhaps ageing Engineers have always preferred to die in a ditch rather than learn new tricks! Good to know Whitworth was once considered a dangerous innovation...

I also enjoyed finding Victorian Engineers attended 'conversazione'. Far posher than typing LOL on a web page! A fascinating glimpse of times past, thanks Michael.

Dave

I'm glad this thread had drifted in to such interesting realms!

Although if anyone wanted to flog me a metric gearbox...

.

Brief summary, Dave ... Holtzapffel contrived a set of thread pitches which were conveniently available on Holtzapffel lathes [not, in principle, unlike what has been discussed here] ... AND, being highly respected manufacturers, these were at risk of becoming de facto standards !!

The linked document presents a reasoned argument against this ... by engineers.

MichaelG.

.

Just in time for an edit: http://ftp.sizes.com/tools/thread_holzapffel.htm

.

P.S. ... For reasons which I may be able to explain anon ...

My current favourite Holtzapffel thread is ‘U’ 

1/10” x 55.11tpi

Edited By Michael Gilligan on 30/10/2019 18:13:06

With a gasp of despair about what ‘Museums’ have become [*]

Here, with precious little information, is the Science Museum’s attempt at an online ‘exhibit’ : **LINK**

https://collection.sciencemuseumgroup.org.uk/objects/co475561/set-of-18-plug-gauges-for-holtzapffels-screw-threads-gauges.

MichaelG.

.

[*] https://www.sciencemuseumgroup.org.uk/about-us/policies-and-reports/digital-strategy/

Edited By Michael Gilligan on 30/10/2019 22:50:18

There must have been some reason Holtzapffel settled on the TPIs that he did. Could it be as simple as that was the resulting thread when using a gear train of a limited number of standard sized change gears? So the resulting TPI was complex but the setting up of the lathe to cut it was simple?

Similar to the way Mr Whitworth arrived at the oddball measurements across the flats of his hexagonal nuts and bolts. It was the largest hexagon that could be made from the limited range of readily available round bar at the time. So you ended up with oddball measurements across the flats but the machining to create them was minimized. It's all about ease of manufacture, not the resulting measurement.

I think from MG's earlier link that Holzapffel had not originally used change gears but used a pattern thread and follower to copy threads which was and still is used most recently on the little unimat lathes. Those original threads would have been struck by eye. It was common practice for brass turners, of candlesticks etc to just make threads by hand and eye as a very loose fit was adequate for the work in question. Likewise for threads in wood.
You can see that the half inch thread for example was trying to be 13tpi (still used today) and so 1in roughly half that. The guy who made the first pattern did incredibly well to get it that close.

It was only years later that some awkward SOB insisted on it being dead accurately repeatable so they had to measure the reference masters and work out how to make them on a precision lathe.

.

As Bazyle has correctly noted; and as per my previous ‘just in time’ edit [*]. The original threads, contrived by John Jacob Holtzapffel, pre-date the leadscrew version of the lathe. ... But yes, your underlying logic is reasonable.

MichaelG.

.

[*] repeated for convenient reference:

http://ftp.sizes.com/tools/thread_holzapffel.htm

Edited By Michael Gilligan on 31/10/2019 05:19:13

For anyone interested in the Holtzapffel digression:

Vol. II of ‘Turning and Mechanical Manipulation’ is available for download here: **LINK**

https://ia802900.us.archive.org/22/items/pli.kerala.rare.6416/pli.kerala.rare.6416.pdf

... and, quoting Charles Holtzapffel:

Between the years 1794-1800, the author's father made a few varieties of taps, dies, hobs, and screw tools, after the modes explained at pages 635 and 636

MichaelG.

Just to close this because the problem is now completely solved. Skip to the last paragraph if you don't want to read the process.

I made a mistake in my initial assessment. The lathe was cutting incorrectly but not for the reasons I thought. I purchased a thread pitch gauge and made some more careful measurements and the leadscrew is genuinely imperial which is, I guess, good.

Using some information from this thread I made a giant spreadsheet with all the various gear ratios, and what pitches each setting would create. I then cut a full range of A, B, C, D, E and 1, 2, 3, .... 8. From that I determined that the ratios between each was correct - so A-B halved the pitch, C-D halved, etc. 1-2 made the equivalent ratio of 8-9 but actually did more like 8.5 to 9.5. Based with that information, I was fairly confident there wasn't anything horrifically wrong with the gearbox.

I checked all the gears in the train under the left hand cover and they all matched what they said they should be, despite the lathe behaving as if the 56 tooth gear was actually a 60 tooth gear (if you put 60 teeth in the spreadsheet, the numbers matched the measurements).

I then took the gearbox off, checked all the gears that transferred drive across the gearbox matched, etc. which they did.

Whilst doing that, I noticed that there was a non-original looking spring washer and spacer washer holding the leadscrew drive gear on. For some reason, that gear had been put on backwards and with a very small tweak from the spacing washer it will mesh with the incorrect gear. It, as it turns out, is supposed to mesh with the final gear (gear 8 under the right hand lever) and it had been put together meshing with the penultimate one, gear 7. That threw the overall ratio out slightly, leading to my issues. I've now flipped the gear round (there's a spacer on one side of it), removed the extraneous hardware and it meshes perfectly with the rest of the gearbox and give the correct ratios.

Once again thank you for all the helpful advice, numbers and support. It gave me enough of an understanding to work out the problem and fix it. Now I have a properly working imperial lathe, I just have to metrify it which is much less daunting with everything as standard!

Hello again Cupboard,

How satisfactory to hear at last the full story. I don't think it would have been possible for anyone here to guess at the way the drive was coupled to the leadscrew; no wonder the results were odd and confusing. Thank you for completing the saga

I am reminded of one query we in the forum were able to solve 'remotely'. for a posting some time ago. It all boiled down to him using a change wheel of 54 teeth that had been incorrectly stamped as 45 teeth and only by suggesting he actually counted everything was the truth of that situation found

Regards

Brian