BA threads

[Nick Clarke 3Participant @nickclarke3]

If you know the difference between what you measure and the accurate value you are OK. My grandad, a retired bricklayer and builder had a very old and worn folding rule that he told me was 1/4" short, but if accuracy was important he knew to subtract 1/4" from the measured reading.

However…….

Our first GP after moving to Birmingham 35+ years ago was a lovely guy and a very skilled medic, but he had an old blood pressure meter in a wooden case – the type with a column of mercury. He would tap the glass tube, presumably to settle the mercury (couldn't see from my side), and then scribble a sum on the desk pad to get rid of a zero error.

Not totally confidence inspiring!

[old martParticipant @oldmart]

When you look at a magnified BA thread, it is rather a strange shape with almost sine wave profile, somewhere in there is the 47.5 degrees.

The angle of Morse tapers is also rather mysterious, every size having a different angle.

Edited By old mart on 01/02/2021 18:36:46

[Martin ConnellyParticipant @martinconnelly55370]

I know this thread is wandering a bit but there's a YouTube video by machinethnking called, I think, 'The Origins of Precision'. Tells the story of how the inch became fixed to 25.4mm.

Martin C

[Michael GilliganParticipant @michaelgilligan61133]

Posted by old mart on 01/02/2021 18:34:54:

When you look at a magnified BA thread, it is rather a strange shape with almost sine wave profile, somewhere in there is the 47.5 degrees.

…

.

… and [rather like Joe's recent query about Whitworth], that's because it was designed around tangential roundings:

**LINK**

https://www.ring-plug-thread-gages.com/PDChart/BA-thread-data.html

MichaelG.

[Nigel Graham 2Participant @nigelgraham2]

Ah, that explains why my thous are so often not as meant… I must be using too varied a mixture of old tools…

'

Further to BA, its diameters were originally metric and in geometrical progression, but of course some committee of bureaucrats decided the odd numbers surplus to requirements, so broke the series!

I have looked at some common thread and spanner standards. BS and ANSI go by screw diameter in mainly logical steps, and their flats are generally to constants or at least logical.

The SI M-series though… How did they generate it? Perusing my poster-size Tracy Tools chart shows it is not at all as consistent as it likes us to think. It follows regular increments for a few sizes, then suddenly jumps to another set. There are any number of M-fine pitch variants and intermediate diameters.

The flats for the common medium sizes are rounded from 1.6D, 1.5D for larger. So why a 17mm (not 16mm) spanner for M10? Below M5 the sizes seem very arbitrary indeed.

M5 – 8mm A/F (1 : 1.60)

M6 – 10mm (1 : 6r)

M8 – 13mm (1 : 1.625)

M10 – 17 (1 : 1.7)

M12 – 19 (1 : 1.583r)

M14 – 22 (1.57)

M16 – 24 (1.5)

M20 – 30 (1 : 1.5)

M24 – 36 (1 : 1.5)

Commercial flange-nuts' stamped, tapered profiles need anything but the nominal M-series spanners.

One of my Round Tuit moments is to equip some of my Myford 7's modern-made accessories with coherent fastenings:

Moving the rear tool-post needs a 5/16 –inch (BSF) and 17mm spanners; the fixed steady's clamp-nut seems from the ISO-Fitnowt range.

The milling-machine's clamp-set (the common commercial type) is all 3/8-inch X 20, either UNC or BSW, but needs a 17mm spanner. I have ordered an M6 clamp-set and it will be interesting to see if a 10mm A/F spanner fits it. Holtzappfel in Swiss inches?

One of my tilting vices has a large nut and 4 screws, apparently all metric but fitting only an adjustable-spanner.

The slitting-saw arbour nut is of Not-Known-Here-Guv A/F but co-incidentally fits (ish) a particular, ancient and very rusty spanner of uncertain parentage I found lurking in the Harrison lathe's chip-tray…

Sir Joseph Whitworth set out to end this palaver…..

[Nick WheelerParticipant @nickwheeler]

Posted by Nigel Graham 2 on 03/02/2021 23:19:39:

I have looked at some common thread and spanner standards. BS and ANSI go by screw diameter in mainly logical steps, and their flats are generally to constants or at least logical.

The SI M-series though… How did they generate it? Perusing my poster-size Tracy Tools chart shows it is not at all as consistent as it likes us to think. It follows regular increments for a few sizes, then suddenly jumps to another set. There are any number of M-fine pitch variants and intermediate diameters.

The flats for the common medium sizes are rounded from 1.6D, 1.5D for larger. So why a 17mm (not 16mm) spanner for M10? Below M5 the sizes seem very arbitrary indeed.

M5 – 8mm A/F (1 : 1.60)

M6 – 10mm (1 : 6r)

M8 – 13mm (1 : 1.625)

M10 – 17 (1 : 1.7)

M12 – 19 (1 : 1.583r)

M14 – 22 (1.57)

M16 – 24 (1.5)

M20 – 30 (1 : 1.5)

M24 – 36 (1 : 1.5)

One of my tilting vices has a large nut and 4 screws, apparently all metric but fitting only an adjustable-spanner.

The slitting-saw arbour nut is of Not-Known-Here-Guv A/F but co-incidentally fits (ish) a particular, ancient and very rusty spanner of uncertain parentage I found lurking in the Harrison lathe's chip-tray…

Sir Joseph Whitworth set out to end this palaver…..

I think you should stop looking at standards, and consider some actual parts.

Metric units are consistent, and relate to each other in relevant ways.

While it would be good if you could generate bolt sizes and their heads in a logical sequence, one look at the BA mess shows why that is a bad idea in practice: it leads to really peculiar sizes. Metric threads are biased towards ending up with sensible numbers, which wouldn't happen if you went with a logical 1:10 progression. Much the same applies to the bolt heads; compare a 5/16 bolt with its M8 equivalent, and you'll see that the heads are similarly comparable at 3/8 and 10mm. And it's all irrelevant, because any experienced mechanic picks his size marked spanner by judging the fastener size by eye.

The designers of metric strike me as being pragmatic enough to know when to accommodate real world values in a logical and we thought out way, like fuel economy – we could do it as km/l, but l/100km gives a much more sensible number.

[Howard LewisParticipant @howardlewis46836]

It may be a mistake to become too fixed on hexagon size vs fastener size, although logical sequence does have advantages.

Sometimes hexagons differ from what we take to be the "norm" . On some Renaults, the M8 nuts on the studs securing the carburetor were 12 mm A/F because of space considerations (Presumably forced by Solex )

Don't forget that during WW2 to conserve material, BSW and BSF head sizes were reduced , (So 5/16 BSW / 3/8 BSF became 1/4 BSW /5/16 BSF size ) ditto BA hexagons were available "next size down", presumably for the same reason, and to aid a scale appearance.

Sometimes differences are driven by practical considerations, where the manufacturer realised that not everyone might have two spanners of the same size. This resulted in the locknut not being the same size as the adjuster, so the adjuster might be M8 with a 13 mm hexagon, but a 10 or 11 mm A/Ff locknut enabling adjustment and locking with two different sized spanners that would be readily available.

Sometimes a manufacturer changed thread standards when introducing a new model. The Leyland Tiger Cub was to manufactured BSW / BSF standards. The brake slack adjusters were identical, apart from the hexagon of the adjuster, to those fitted to the successor Leyland Leopard, which was to Unified thread standards.. Since the slack adjusters were physically interchangeable, it was not uncommon for a fitter to need a 3/4 A/F spanner on one side and a Whitworth spanner on the other. This was far better than having an expensive vehicle off the road for 24 hours awaiting a "genuine" replacement..

Howard

[Tim StevensParticipant @timstevens64731]

Japanese motorcycles seem to avoid the 13mm spanner size – maybe on superstitious grounds? – using 12mm in preference except for really HD 8mm stuff which could be 14mm.

Before the war, lots of oddities occurred – Humber used 10mm fine threads with BSW head sizes, and Morris in both early Cowley & Oxfords, and later T series Mgs, for example, use metric fasteners with W heads. Some of their engines were made by Hotchkiss, using continental fittings, but British heads to avoid complications when serviced in British garages.

And Norton Commandos mixed BSF and UNF threads – even with the same thread length and hex socket sizes …

Cheers, Tim

[Peter Jones 20Participant @peterjones20]

Posted by peak4 on 31/01/2021 20:02:59:

Here you go folks, some interesting reading for your next tea break; link to part 2 at the foot of the page
https://www.sizes.com/library/technology/thread_BA1.htm

From the 2nd page section 9
"For, as has recently been pointed out by Mr. Bosanquet,⁵ it is easy to cut a thread, whose pitch differs from one millimetre by an amount which may for all ordinary purposes be neglected (1/155300th), with a guide-screw based on the inch by the addition of a wheel of 127 teeth"

Now of course we can cut an exact 1mm pitch thread with a 127 tooth gear since the inch is defined as 25.4mm
There's an interesting article HERE on the varying definition of the the "Inch" with the passage of time. (I have posted that one before, but it's still worth a read.)
http://metricationmatters.com/docs/WhichInch.pdf

Bill

Edited By peak4 on 31/01/2021 20:04:12

Thanks Bill, it cleared up some questions I had from when I tried researching metric system in early 2000~2002 as I was teaching a 'machine shop' course at Motorcycle Mechanics Institute.

At least it justified my statements to students that they had been using metric system and metric measurements since the 1860's (plus, of course, they only use 'metric' money)

You would not believe the 'arguments' from people saying they 'don't understand metric'

I always asked if they ever went shopping and needed to take someone with them to figure out how to pay

[Peter Jones 20Participant @peterjones20]

Posted by Howard Lewis on 04/02/2021 15:52:06:

It may be a mistake to become too fixed on hexagon size vs fastener size, although logical sequence does have advantages.

Sometimes hexagons differ from what we take to be the "norm" . On some Renaults, the M8 nuts on the studs securing the carburetor were 12 mm A/F because of space considerations (Presumably forced by Solex )

Don't forget that during WW2 to conserve material, BSW and BSF head sizes were reduced , (So 5/16 BSW / 3/8 BSF became 1/4 BSW /5/16 BSF size ) ditto BA hexagons were available "next size down", presumably for the same reason, and to aid a scale appearance.

Sometimes differences are driven by practical considerations, where the manufacturer realised that not everyone might have two spanners of the same size. This resulted in the locknut not being the same size as the adjuster, so the adjuster might be M8 with a 13 mm hexagon, but a 10 or 11 mm A/Ff locknut enabling adjustment and locking with two different sized spanners that would be readily available.

Sometimes a manufacturer changed thread standards when introducing a new model. The Leyland Tiger Cub was to manufactured BSW / BSF standards. The brake slack adjusters were identical, apart from the hexagon of the adjuster, to those fitted to the successor Leyland Leopard, which was to Unified thread standards.. Since the slack adjusters were physically interchangeable, it was not uncommon for a fitter to need a 3/4 A/F spanner on one side and a Whitworth spanner on the other. This was far better than having an expensive vehicle off the road for 24 hours awaiting a "genuine" replacement..

Howard

I always thought they were Continental and Japanese industry standards. Continental (French/German) 8mm bolts have 13mm head and JIS have 12mm head

It's been 'common knowledge' in the motorcycle industry since the ISO standard was changed in 1965 (ish) Now, all metric threads will interchange even if bolt head size is different. Anyone who has restored a pre-1965 'import' (generally Japanese or German) knows there are differences in thread design

[Nigel Graham 2Participant @nigelgraham2]

Peter –

Thank you –

A fascinating article! I've saved the link in the 'Engineering' folder of my "Favourites".

It's intriguing to note that although the French seemed to be first away with telling the fledgling USA how to measure things, they still use Imperial for everyday use, though their scientists and engineers now use SI metric.

I used to be a regular user of Wikipedia's "Answers" Q&A site. Its Maths category included a sizeable section on converting between the systems, and most of the questions were pretty obviously from children wanting respondents to answer their homework questions, such as "How many miles in 80km?".

I would answer by showing you look up the required, widely-published constant and multiply or divide accordingly, but the site was plagued by two or three who delighted in turning it into a baffling nightmare of Algebra (needless, really) and Dimensional Analysis (which it isn't). They would show great long chains of intermediate conversions – sometimes down to an inch and back up- for a sum like that; then often make mistakes in their own arithmetic!

I think I'd point out that for precision you use the appropriate converter, but for a real journey from one town to another, 1km = 5/8mile is often close enough because you are not normally travelling between the towns' mapping-points. I certainly would not though "take pi = 3", as someone once told me his daughter's teacher set in a homework exercise!

Some years ago I happened to see someone's souvenir copy of the programme for a recital celebrating Oslo Cathedral's organ back into use after a major overhaul. I don't know when Scandinavia went metric, but I noticed that the instrument's voice specifications were all in fus (feet).

[Howard LewisParticipant @howardlewis46836]

Organ pipes seem always to have been measured in feet. $ ft, 8ft, or 16 ft. Presumably because under the normal ambient conditions in churches and cathedrals, pipes of those lengths could be fine tuned to give exactly the note wanted, depending on whether it was open or stopped.

Without good air conditioning, a heatwave or deep frost would upset anyone with perfect pitch!

Howard

[Nigel Graham 2Participant @nigelgraham2]

Some pipes are in whole-and-thirds feet, too, such as 2-2/3 ft.

They all have some form of tuning sleeve, but I don't know how much the longer metal flue pipes are affected by temperature changes. I'd guess not noticeably except to perhaps those blessed with particularly good ears for pitch, even for the 16ft and (on some) 32ft, pipes.

Humidity changes might the more serious problem for these complicated instruments.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Howard Lewis on 05/02/2021 16:50:18:

.

Organ pipes seem always to have been measured in feet.

.

The Divine unit of measure !

MichaelG.

.

https://commons.wikimedia.org/wiki/File:Europe_a_Prophecy_copy_K_plate_01.jpg

Edited By Michael Gilligan on 05/02/2021 17:24:22

[Nigel Graham 2Participant @nigelgraham2]

LOL!

He is probably using the Standard Astronomical Unit in that painting.

There is at least one other version of the image, showing God apparently measuring the diameter of the planet Earth.

It struck me that the artist might not have realised the Creator would of course know to measure a sphere with calipers not dividers, but either way, the instrument in that second image is more closed, intriguingly resembling part of the badge of a certain social society originally a guild of church-builders!

[Nigel (egi)Participant @nigelegi]

The notes from the committee meeting where the BA system was introduced are available:

Report of the British Association for the Advancement of Science (1882) – Report of the Committee consisting of Whitworth, Siemens, Bramwell, “appointed for the purpose of determining a Gauge for the manufacture of the various small screws used in Telegraphic and Electrical Apparatus, in Clockwork, and for other analogous purposes.” 52nd Meeting (1882): biodiversitylibrary.org/item/95237 Page(s): Page 311 to 314

[duncan webster 1Participant @duncanwebster1]

Posted by Nigel Graham 2 on 05/02/2021 17:19:33:

Some pipes are in whole-and-thirds feet, too, such as 2-2/3 ft.

They all have some form of tuning sleeve, but I don't know how much the longer metal flue pipes are affected by temperature changes. I'd guess not noticeably except to perhaps those blessed with particularly good ears for pitch, even for the 16ft and (on some) 32ft, pipes.

Humidity changes might the more serious problem for these complicated instruments.

Note from an organ pipe depends on length and speed of sound in air, both of which are temperature dependant. I think speed of sound increases with sqrt T whereas length of pipe will depend on T so less effect than you might imagine. Either way I think you'd need to listen very carefully.

Mine has wooden pipes, and coeff of expansion of wood is very low, so perhaps in a church organ the wooden ones go up due to air temp change, but the metal ones come down due to combined effect?

[Author]

Posts