Harrisons gearing and friction.

I have been thinking more closely about Harrison and his gearing and i want to make a Harrison clock of my own design but i lack some detail. Harrison used small dia rollers to run the clock arbors on to reduce friction but does anyone know what the end of the arbors should look like, what i am trying to say is that the end of the arbor (in therory) should not touch the roller but run like a train wheel which is self centering, as trains were well before his birth does anyone have the answer, David

Harrison used four anti-friction devices for diffferent applications. These were:-

1 Use of loaded wheel material--oily woods in segmented wheels with radial grain

2 Anti friction wheels- pairs of plain edged wheels seperated horizontally with the shaft sitting on top of both wheels

3 Anti friction arms- two long arms ( about 12") with radiused ends, shaft sits in the notch between two arms, counterbalenced and sprung to allow arms to oscillate over a small angle- used on H1 on the axles of the inertial arms.

4 "Rollers".

The idea of the wheels (2 + 3) is to reduce friction by using mechanical advantage. Imagine that you have a 1/16" shaft resting on a 3" plain wheel with plain bearings with some friction u. When you turn the shaft 1 turn the bearing shaft rotates by 1/48th turn so the friction is u/48, a large amount. However there are two plain wheels at each end of a horizontal shaft so the overall reduction is u/12 but still a great advantage and totally in keeping with the concept of no lubrication. So the principle emerges that you get a friction reduction of vr/4 with four wheels where vr is the velocity ratio i.e. the ratio of the shaft to wheel.

You stop axial movement on the main shaft by turning a short taper to reduce the diameter to the wheel contact section so the shaft is centred between the cones, the turned down diameter in contact with the wheels being the smallest that you can get away with!

The only currently made anti-friction wheel device that I know of is the prop balencer made by SLEC. The same idea can be used for loco wheels or motor rotors. There is a single wheel bearing on a tumbling barrel in the Stott Park Mill.

The roller bearing is very different to the anti-friction wheels, in the roller bearing the details of the roller constraint set the frictional losses however lateral shaft constraint can follow the concept of cones as above.

Regards,

Alan

It's possible that a very scholarly book I've recently encountered may be of assistance to you:

John Harrison's Contrivance by Stuart Harrison published by Fionchra Press, Cedar House, Orford, Suffolk, IP12 2NQ. ISBN 978-184426-773-6. I don't believe there is any familial connection between the author and the subject of the book.

It would not be too innapropriate to use ball bearings in a clock. There is a sample copy of the BHI Journal on the BHI website which discusses this application. About 20 years ago I was given a Devon Clock Co "Sea Clock" kit. which is very loosly based on H1. It has two compound pendula on two shafts connected by toothed quadrants and sittting in a total of four ball races. After assembly it would only run for around 30 swings before stopping. That was because the grease had not been flushed from the bearings. After trying IPA and meths with the bearings in place, I dismantled the clock then ran the races in white spirit which finally removed the grease.

Like the original, this deviant clock has the escape wheel mounted between the ends of the shafts. A pallet arm on each shaft alternatly locks on the escape wheel. This arrangement is very tricky to set up and somewhat counter-intuative. If both arms loose lock then the scape wheel spins up to very high speed then chops off two grasshopper's legs!

The clock is a very poor timekeeper however although it does use pendulum oscillators you can pick up the whole clock and rock it around without missing a beat. The poor timekeeping is due to a lot of different factors. The Q of the resonator is very low, there is loads of friction in the oscillator which causes each beat to roam over a considerable range. The saving grace -like some other clocks- is that the variations average to a mean rate that is not too far out over a day.

Regards,

Alan

Many thanks to those who have replied to my question and i dont want to appear rude but i have studied Harrison for over 30 years and made quite a few grasshopper clocks but my main question was about friction and his use of anti friction rollers/wheels and what profile the arbors have when abutting up against those rollers, its an area which i have never thought of and more to the point never seen in print, regards David

David,

If your question is "what form do the pivots have in Harrisons clocks?" then the answer for H2 which uses anti friction wheels for most arbours is that they are straight cylindrical pivots like any clock. I had the opportunity to see H2 disassembled at the Greenwich museum last year and was a bit surprised at this. I guess that the shoulder prevents the arbour to move out of position and the pivots look a bit longer than a "normal" clock. I am not a Harrison expert but hope that this will help a little. I recommend a visit to the observatory in Greenwich to study the displayed clocks where this should be visible

Regards Anders.

Further to Alan Gray's description of friction wheels it is important to differentiate between rolling, sliding and static friction. Henry Sully (1680-1728), working in Paris sent a marine timekeeper to George Graham in 1724. This clock has friction wheels supporting the balance wheel.

Because there is no drop on Harrison's asymmetrical grasshopper escapement he had to reduce friction throughout clock.'s fitted with it The drop of anchor and dead beat escapements in conventional clocks produce the audible tick and shakes the frame thereby releasing static friction.

Harrison's ultimate pendulum clock, now known as the Royal Astronomical Society Regulator is the clearest example of his methods of reducing friction.

The geometry of the friction wheels in this clock only allow motion in one direction. because the wheel arbor and friction wheel arbors are a right angle the bisector of which is in the line of force being applied to the wheel. This means that there is efficient rolling friction in the contact between the wheel arbor and the friction wheels. His friction wheels have lignum vitae collets to minimise sliding friction. PTFE has a similar coefficient of friction as lignum vitae and is the material of choice for modern constructors. Friction wheels only allow motion in one direction; the wheel arbors are held against the friction wheels by the driving force.

For this reason Harrison had to fit roller bearings on the great wheel arbor of the RAS regulator. (Think of horology as one-way sloppy engineering!) Harrison's design of roller bearings copied from da Vinci was copied by Lord Grimthorpe for the outer bearings of the hands of the Westminster Clock.

Many thanks to all who have replied and i now have a germ of an idea where the inner roller and the arbor shoulder have a small radius which keeps the arbor in line, i didnt like the idea of 2 90deg surfaces rubbing together even though they would have been small, cheers David ps to one of the respondents who mentions Harrisons clocks under test, see my short video on youtube under H1 where i give my experimental clock a good workout

David, Just to make sure that my explanation is understood. The shoulders are perfectly square. However, after refreshing the memory I remember (and can see) that the ends of the pivots are conical and the marks where the rollers made contact and minor wear are not close to the shoulder.

Since it irritated me that I could not remember exactly how it is done I looked thru the book "Time restored" by Jonathan Betts. On page 92 there is a picture of H2 and there you can see that there is a small bridge that serves to hold the bearings for the rollers and also will keep the arbour in place. So for H2 at least it is clear. The roller wheels roll on cylindrical pivots and the pointed cone at the end of the pivot will keep the arbour in place lengthwise.

A final comment that is certainly known by David and Stuart but maybe not by all is that the arbours and pivots were made of brass.

Anders

In answer to the question posed in the original posting of this thread, John Harrison used 'endstones' whenever possible to control axial position and thrust of arbors. His methods are clearly seen by careful examination of the Royal Astronomical Society regulator. In this clock the arbors are made of gunmetal, with the exception of the remontoire arbor which is a tube made of brass. The arbors have the smallest diameter possible where they are in contact with the friction wheels and the periphery of the friction wheel is perfectly square. The ends of the arbors are conical and almost certainly originally abutted against lignum vitae endstones. During subsequent restorations the lignum vitae has been replaced with blued steel. (see photograph). Because of the hand it is impossible to fit a front endstone to the centre arbor, so Harrison turned a collar on the arbor between the friction wheels, thereby taking the axial thrust on the side of the friction wheels. The remontoire arbor and the escapement arbor are co-axial making it impossible to have endstones on the remontoire arbor. He therefore turned a collar similar to the centre wheel arbor. (see photo) The rear roller bearing of the great wheel has a lignum vitae endstone. A similar arrangement is impossible for the front bearing because of the winding square. Harrison makes the diameter of the front bearing much larger than the rear so that the axial thrust is towards the rear endstone. (see photo).

Applying friction wheels to a tapered arbor is a very inefficient way to control axial position and thrust which I do not think would have been used by Harrison but anyone can try it in their own design. (or disagree with me)

Many thanks Stuart and Anders, i have kept my replies short as i am suffering cataract problems at the moment and only have the use of 1/2 an eye which makes reading slow, using end stones makes sense and biasing the arbors where no end stones are fitted is an idea which i had not thought of, i have worked on clocks which have a steel ball as an end stone and of course all carriage clocks have an end stone of sorts for adjusting the depth of the contrate wheel, cheers and thanks David

Hello Stuart, Niloch mentions that you have a book out about Harrison but i cant find it listed anywhere, i have tried the usual isbn No but no luck, cheers David ps i looked at your pics, very impressive work but i cant see, rem fly endstone and rem arbor collar for some reason

Try Shenton Books they have it listed fo £47.50

 

David,

My book, 'John Harrison's Contrivance' is available from Shenton Books, www.shentonbooks.com, 0845 8385523, G.K.Hadfield, www.gkhadfield-tilly.co.uk. 01768 870111, Jeffrey Formby Antiques, www.formby-clocks.co.uk, 01608 650558. and Ian T. Cobb, www.clockmaking-brass.co.uk, 0116 2676063. There are the beginnings of fuller details on my website, www.fionchra.com. I hope that this helps. Stuart.

Many thanks Stuart, i did look at your site and the book looks very comprehensive but i dont have the funds to buy it just yet. I do however take a great amount of detail and idea's from your messages, one thing i didnt mention was i would like to make my clock from Oak like he did in his earlierer work, which builds a new layer of complications into the mix. I have made many clocks over the years but i feel this will be my last one, i am enjoying the research more than the build, so much thanks David

Had a glance at Stuart's book on Sunday, ordered it that night , arrived Tuesday. It is the first detailed account of a John Harrison clock that I have seen, Stuart also gives a great deal of background detail and a concise history of early timekeeping. Some of John Harrison's thoughts and ambitions are quoted which helps form a much better picture of his genius.

 

A very good book.

 

Regards,

Alan.

'Sleeping in Oblivion': http://soptera.blogspot.com

 

First published December 2007. Still available in DVD format (all pages print enabled) as the 3rd Edition.

 

Videos of the 2003-2005 'replica' at www.youtube.com : please search for the word glathoppa and, if you have the time, may I point to a reading the comments for each video (left click on the square chevrons button).

 

 

I'm confused as to why Mr Creed is asking his original question: he has a copy of 'Sleeping in Oblivion'.

Hello David, you have a private message, regards David Creed