We have a very old pendulum grandfather clock - the pendulum is probably about 2 feet long with a slow tick. I've noticed that in the hot weather, when it's probably been warmer inside the house than normal, the clock is running *fast*. I would expect it to run slow, since the pendulum will lengthen slightly in the warmer temperature and
t = 2 pi sqrt (l/g)
as l increases, so does the time period t between ticks
The pendulum is a single rod, not a multi-bar rod with two different metals so as to compensate for thermal expansion.
My thought too. I wonder if it is a "spring" one rather than having traditional weights. Whilst the spring unwinds very slowly, there's a lot of area in contact with a very thin film of often rather thick, sticky oil. If I had a bit more time I'd have a look at what temperature does to the viscosity of air, which is also slowing the pendulum.
Sorry, I should have said. It uses weights rather than a spring. Bloody nuisance having to remember to wind it every morning :-)
Would viscosity of lubrication oil and air around the pendulum affect the timing of the pendulum? I can imagine them causing more energy to be lost as heat.
If the pendulum is hung from a flexure pivot, then lubricant doesn't affect the "free" period of the pendulum, but the escapement does give it a little "push" (after all, it has to supply some energy to overcome air resistance). So less friction in the drive mechanism may give you slightly more "push".
Engineering Toolbox shows both dynamic and kinetic viscosity of air increasing with temperature, which is not what I was initially expecting. On reflection, though, that is right. It's been a long time since I thought about kinetic theory.
We have a similar pendulum clock and see that it also runs a bit fast in this weather. My guess was also that it was the lower viscosity of the oil, or perhaps lower air drag on the pendulum.
How is the pendulum bob fixed to the rod? I have a 1900-ish pendulum clock whose flat circular bob is supported at its bottom. So as the temperature rises the bob's center of gravity moves upwards and tends to compensate for expansion of the rod downwards. The clock keeps excellent time, within 5 seconds per day when compared with a radio-controlled clock. This is despite its pendulum rod being made of wood which I would expect to be affected by atmospheric humidity, but it doesn't seem to be.
That's a *very* interesting point. Wood seems to have an expansion coefficient of around 3 x 10^-6, the first reference I checked gave brass as 10 to 19 * 10^-6 (both per degree C).
The pendulum on this clock supported at the bottom like that. Goodness knows how the length is adjusted, because the weight doesn't move freely up and down the rod, resting on a nut that you can wind up and down a threaded rod. It looks as if you have to rotate the whole weight to move it up and down the screw, but if that's how you adjust it, it's very stiff and I'm not about to try it in case I break something. I just correct it each morning (and maybe in between if I notice it's wrong) so it remains no more than a couple of minutes fast.
+/- 5 seconds a day is pretty bloody amazing for a clockwork clock. It's good even for a quartz clock: the real time clock on my PC loses or gains more than that, but then gets corrected by scheduled process syncing with time.windows.com (or whatever) every 24 hours; I think the worst I've seen it (compared with time.is) is about 90 seconds out.
How long did it take to get it that accurate, given that you have to make a small adjustment to the pendulum length and then wait ages to see how much time it has lost over maybe 24 hours to work out how much further you need to adjust the screw, iteratively homing in on the correct setting. I suppose if you know the gearing of the clock (pendulum ticks per minute of hand movement) you can time a few ticks and extrapolate from there - but the more ticks you count, the less measurement error there will be.
Actually, it's uncanny. When I went to check the pendulum fixing just now, I found that the clock had stopped literally a minute or so ago, after I'd forgotten to wind it this morning. What is the chance of it stopping exactly at the time that I check the clock?
This is a very old full-height grandfather clock that has been in the family for years. It's named Francis after a distant ancestor who was a clock-maker. It would have been incredible to have found one of his own clocks in an antique shop, but my parents did the next best thing and bought one from a clockmaker who had lived in the same part of the country at about the same time - which I think was early 19th century. I presume the mechanism is original - in design, even if some parts have been replaced like-for-like over the years.
You'd tend to expect a big clock to have a deep sonorous chime, but this one has a little bell about the pitch of a bell used on a shop counter for attracting the assistant's attention - a bit weedy, as if its voice hasn't broken. The chiming mechanism is a bit uneven: I presume it has a wind vane that crudely regulates the interval between the chimes, and sometimes for no apparent reason you get a few chimes with a bit of time between them and then a couple of hurried ones at the end, as if it's saying "Sod it. Twelve o'clock is a lot of hard work. I'm nearly there, try and get the rest over quickly." But that just gives it character ;-)
Harrison's H4 was pretty incredible, but in fact the French were close behind
formatting link
Can't immediately spot the accuracy of the H4, but this link has an unreferenced statement suggesting that mechanical marine chronometers got to 0.1 seconds per day, giving an accuracy of 1 - 2 miles after a sea voyage of a month.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.