Unusual Escapement

[David NobleParticipant @davidnoble71990]

While on holiday recently, my wife and I visited a stately home where i saw this clock. Viewed from distance, I couldn't see how the pendulum was getting an impulse until I looked closer. As I'm a new clockmaker, I hadn't seen anything like it. It seems to have solved the inherent problem of inertia when the wheels stop and start.

Edited By David Noble on 25/05/2022 20:33:03

Edited By David Noble on 25/05/2022 20:33:58

[David NobleParticipant @davidnoble71990]

[lfoggyParticipant @lfoggy]

Looks amazing but I don't really understand what's going on here. Was the clock running when you saw it?

[David NobleParticipant @davidnoble71990]

Hi ifoggy,

Yes the clock was running. In the middle photo, you can see a spindle coming out of the clock which was rotating and pressing onto the base of the pendulum. I suppose strictly speaking, the pendulum is describing a circle rather that oscillating.

The title of my post is also misleading as it isn't an escapement in the true sense.

David

Edited By David Noble on 25/05/2022 21:52:08

[Sam StonesParticipant @samstones42903]

Hi David,

I just had to turn your pictures around.

Cheers,

Sam

PS I haven't figured out how the impulses are transferred.

[David NobleParticipant @davidnoble71990]

Thanks for that Sam

There are no impulses! The thin needle is driven from the clock mechanism and without the pendulum, would be free to spin out of control. With the tip of the pendulum pressing on it the spin is restricted as it tries to push the pendulum away.

David

[Michael GilliganParticipant @michaelgilligan61133]

[duncan webster 1Participant @duncanwebster1]

There is a clock with a conical pendulum in the museum at Liverpool. I think it came from an observatory (perhaps Bidston). I couldn't see how it worked either. If you super-impose 2 sinusoidal motions at right angles you get a circle, but not sure how that helps. This shows that the period of a conical pendulum depends on the cosine of the angle of the rod to vertical, so needs to be controlled accurately. perhaps someone can come up with an explanation

[Jon LawesParticipant @jonlawes51698]

Interesting stuff, thanks for sharing.

[Michael GilliganParticipant @michaelgilligan61133]

That one is a magnificent piece of work, Duncan … and very sophisticated

The conical pendulum, in the dome atop, is not the timekeeper as such

… I suppose it best qualifies as something more like a flywheel.

**LINK**

http://resources.schoolscience.co.uk/POL/insight/bondclock.html

MichaelG.

.

P.S. __ The first couple of times I saw it, it was working … but more recently it was just a dead thing, badly lit.

P.P.S. __ This is worth a look:

https://dash.harvard.edu/bitstream/handle/1/37364391/Partners%20in%20time:%20William%20Bond%20&%20son%20of%20Boston%20and%20the%20Harvard%20College%20Observatory.pdf?sequence=1

Edited By Michael Gilligan on 26/05/2022 06:39:51

[Michael GilliganParticipant @michaelgilligan61133]

EUREKA !!

.

An unexpected find ^^^

MichaelG.

[David NobleParticipant @davidnoble71990]

Thank you Michael, I was just thinking that I ought to have taken a video.

David

[lfoggyParticipant @lfoggy]

Is the 'pendulum' working like a governor? The needle rotates with a constant torque, which drives the suspended bob, which moves outwards, increasing the cirumference of its rotation, until the forces are balanced? Hopefully that point would be associated with the correct rate of the clock. Sounds like there would be a lot of room for error in this arrangement.

[John HaineParticipant @johnhaine32865]

Horolophile's videos are very interesting – several escapements built from Lego plus other clocks working.

This conical pendulum is deeper than it looks. Back in the late noughties there was a lot of speculation about coupled pendulums. Observing that anti-phase coupled pendulums were less sensitive to horizontal ground vibration, the idea was that quadrature ones might be insensitive to vertical vibrations. It turns out that while two pendulums will easily synchronise in anti-phase it's impossible to make them do so in quadrature. A clockmaker called Colin Fergusson suggested that an alternative would be to have a conical pendulum, which in effect oscillates in two orthogonal planes in quadrature. He did an experiment with a bob on a string suspended from the shaft of a DC motor and observed a very pronounced "resonance" at a certain speed. This seemed to contradict the accepted theory of conical pendulums (as "explained" in one of the references cited above) which has nothing to say about resonance. When I was at school we covered conical pendulums on A-level physics and were told that they were not resonant like an ordinary pendulum.

It turns out this is quite wrong! As Colin observed, experimentally they are resonant! If you do simulations of them, or construct the equations, this is confirmed. They have a rather similar response to an ordinary pendulum, except that their "circular deviation" of rate with amplitude is the opposite sign. Also since the rate depends more strongly on amplitude the shape of their response curve is distorted, falling off more rapidly on the high-frequency side.

So the explanation of the clock in the first photos is that the radial arm applied a constant torque to the pendulum which accelerates it towards its resonant frequency. As it reaches that its amplitude, and therefore velocity, increase so it absorbs much more energy. Once the air resistance torque equals the driving torque the amplitude stabilises – just like a conventional pendulum.

Sadly it turns out that the conical pendulum is not insensitive to vertical motions. I wrote an article on this for Horological Science News, happy to share with anyone interested.

Of course the pendulum doesn't have to have the same period in the two planes, if the suspension is gimballed. Solar/sidereal clocks have been made with a single pendulum vibrating at different rates in the two planes with separate escapements driving separate trains and faces.

Bond's clock uses a conical pendulum governor that runs slightly faster than the main pendulum to drive a very clever escapement which is essentially completely "detached". There's a very nice description of it in Philip Woodward's My Own Right Time.

[duncan webster 1Participant @duncanwebster1]

Is the connection from the string to the motor shaft offset from the rotation axis? If so we would have a different model to the one in my original link

I googled the book in John's post, it is listed in Children's Literature by Amazon! It is however £83 to buy, so I'll give that a miss.

Edited By duncan webster on 26/05/2022 12:51:52

[David NobleParticipant @davidnoble71990]

Posted by lfoggy on 26/05/2022 11:09:12:

Is the 'pendulum' working like a governor? The needle rotates with a constant torque, which drives the suspended bob, which moves outwards, increasing the cirumference of its rotation, until the forces are balanced? Hopefully that point would be associated with the correct rate of the clock. Sounds like there would be a lot of room for error in this arrangement.

Yes, I 'think' that is what is happening but as I said, 'I'm very new at this game'

David

[John HaineParticipant @johnhaine32865]

That's essentially true, but it is much easier to drive the pendulum at its "resonant" frequency. Yes, the pendulum is a "governor" but so is a conventional pendulum. In Bond's clock, and a related one made by Lord Kelvin, the but contacted the inside of a friction ring to brake it at the right radius for the required speed.

To Duncan's point, yes the string was offset from the motor shaft slightly, but once the system settles down the bob is rotating such that it is at the "opposite side" to the offset, so the effective suspension point is just below the drive.

[duncan webster 1Participant @duncanwebster1]

If we have a conventional pendulum swinging with an amplitude of say 1 inch either side of vertical, and then move the suspension point 1 inch either side at the resonant frequency but in anti-phase, the bob will stand still. If the suspension point is moved at a slightly different frequency I think you get a chaotic movement, which absorbs a lot of energy. If you move the suspension point at resonant frequency but with lower amplitude I think it provides the impulse to keep the pendulum swinging. Superimpose a movement in the third dimension at 90 degrees and you get conical. I still think this is different to a fixed suspension point

Edited By duncan webster on 26/05/2022 13:56:20

[John HaineParticipant @johnhaine32865]

Posted by duncan webster on 26/05/2022 13:42:04:

If we have a conventional pendulum swinging with an amplitude of say 1 inch either side of vertical, and then move the suspension point 1 inch either side at the resonant frequency but in anti-phase, the bob will stand still. ….

How can that be true? What happens to the stored energy in the bob?

The analysis of the conical pendulum assumes that forces are applied to the bob in the x and y directions, not to the suspension, but the results are essentially the same as the simulation which assumed that the suspension point moves.

[duncan webster 1Participant @duncanwebster1]

Posted by John Haine on 26/05/2022 21:23:36:

Posted by duncan webster on 26/05/2022 13:42:04:

If we have a conventional pendulum swinging with an amplitude of say 1 inch either side of vertical, and then move the suspension point 1 inch either side at the resonant frequency but in anti-phase, the bob will stand still. ….

How can that be true? What happens to the stored energy in the bob?

…….

On reflection, the bob goes up and down vertically, but doesn't move side to side. Superposition means you can consider the suspension fixed and the bob swinging (ie moving side to side and up and down from the chord to the arc), then superimpose the whole frame of reference moving with the same amplitude but out of phase, so the bob goes up and down and the pivot goes side to side. This only in one plane, I have to think a bit more about 2 planes.

I don't understand your second point, which simulation?

Edited By duncan webster on 26/05/2022 21:47:36

[duncan webster 1Participant @duncanwebster1]

I should have repeated, the bob doesn't move side to side only in the unique case that its amplitude relative to the suspension point is exactly equal to the amplitude of the suspension movement. This would be quite difficult to achieve in practice, and I suspect would be unstable, any perturbation being likely to set the bob moving sideways. If I had a suitable geared motor I'd set up an experiment, but too many other irons in the fire

Jim? Radford described a clock in ME many years ago where the pendulum was impulsed by moving the suspension from side to side. Unfortunately I didn't scan the article, and I've got rid of paper copies, the loft was creaking.

[John HaineParticipant @johnhaine32865]

**LINK**

https://en.wikipedia.org/wiki/Riefler_escapement

I still don't see this Duncan – if the bob is stationary relative to the suspension point but the latter is moving, the bob has momentum at least as it passes through the mid position, and therefore kinetic energy – where does that go for it to be stationary when it reaches an extreme?

Re simulations – the article I wrote for HSN described these as well as the analysis.

[duncan webster 1Participant @duncanwebster1]

The bob isn't stationary relative to the suspension point, it is swinging at its resonant frequency. The suspension point is moving relative to the rest of the world sinudoidally with the same amplitude but 180 degrees out of phase. Add the two together and the bob has no sideways motion relative to the rest of the world.

[Sam StonesParticipant @samstones42903]

Back again, David.

Duncan and John's input, have added more food-for-thought!

Contemplating the effects of shifting the pendulum suspension horizontally in sync with its time constant can, to say the least, be tricky. The conical pendulum system demands even more concentration.

With reference to Michael’s Farcot/Carrier-Belleuse Conical Mystery Clock link …

**LINK**

… although difficult to follow; between 0:56 – 1:00 of the video, and also the wheel and its speed variation briefly visible between 1:10 – 1:14, it appears that the bob of the conical pendulum is tracing an ellipse. Since the contact point of the pendulum tip visibly slides back and forth a short distance along the ‘torque arm’, the effective torque transfer varies.

The suspension of the pendulum appears to be a ‘twin’ gimbal, similar in principle to Hook’s universal joint.

Q1. Does this form of suspension have little x/y influence?

Q2. Would it be safe to assume that the elliptical trace results from not only the varying torque, but also the earth’s rotation?

Sam

Edited By Sam Stones on 27/05/2022 23:35:48

[David NobleParticipant @davidnoble71990]

Hello Sam,

Yes I did notice the wheel speed variation that you mention. I thought, that as the speed of the wheel is varying, the radius that the pendulum is describing would also vary. Therefor tracing an ellipse.

Or, there is something causing the pendulum to trace an ellipse which in turn slows the clock.

Or, 'radical thought' is there a normal escapement hidden inside and the conical pendulum is for show? 

David

Edited By David Noble on 28/05/2022 08:10:23

[Author]

Posts