Quote:
Originally Posted by saxo3
Interesting questions, hard to answer. It is very difficult to do any quantitative analysis and try to calculate thermal effects inside a highly compact watch movement comprising many individual components with different material properties.
Used materials, masses, dimensions, thermal contacts and anchoring, friction, viscosity of used oils etc. make a quantification (nearly) impossible.
Take the Nb-Zr-O Parachrome spiral. In order to get a rough approximation about the impact of temperature changes, you would need to determine the so-called expansion coefficient for the Niobium-Zirkonium-Oxygen alloy for a dynamic (no static) situation, other physical properties as well, all without knowing the alloy composition, how the spiral was tempered and so on. Hopeless.
For any systematic study one has to be sure that the entire watch has reached a thermal equilibrium, then your only free parameter is this stable temperature, let it be in water, at air or anywhere else.
Just in case you wish to investigate further, which I don't suggest, you would need to measure the watch temperature in thermal equilibrium and obtain a more precise measurement of the amplitudes (305-315 vs. 290-295).
Anyhow, with such high amplitudes I don't expext any changes w.r.t. timekeeping, it questions more how precisely and reproducibly your timegrapher can measure.
Maybe you can join us with your 32xx data?
|
That's a LOT of variables. I don't think I have the equipment or the expertise to dive into it as deeply as I would like to.
From the timekeeping perspective, it doesn't seem to make a lot of difference. It consistently stays within the -2 to +2 range, and it's
extremely consistent when factoring in my daily activity. If I am sedentary during the day and wear the watch to bed, it's consistently right about -2 SPD (which I surmise has to do with the watch resting on an angle where there is more friction as it loses the bulk of those two seconds at night). If I am sedentary-to-mildly-active during the day and leave the watch DU or DD when sleeping, it is close to 0 SPD. If I am extremely active during the day, it's about .5 to 1. It's amazing how something with so many moving parts is so accurate and also how many different factors impact this accuracy. But it absolutely blows my mind that whoever regulated this watch did so in a way that could accurately account for my own (often vastly different) behavioral variances to still operate within that -2 to +2 SPD range. Rolex is awesome!
The timegrapher I have is one of those imported 1000 models, so it's not precise enough for me to dive into this deeper, and I don't have a sufficiently accurate means of measuring temperature either. That said, it does make me curious how much the temperature of watches impacts amplitude variance when cross-comparing. If my house is 22 C and someone else's 27 C, how much does that variable alone impact measurements.
By the way, thanks for organizing all of this information! Next time I log some data I'll be glad to add to the mix!