my guess would be the dripping of the wax causes it
since they'll never burn at precisely the same rate, you'll get a situation where one side (A) drops, and so the other (B) is heavier.
B will then cause the system to rotate (I'm guessing very slowly at first) but this will bring the flame in to closer proximity with the wax, causing B to drip and lose weight; A will then be heavier and the process repeats. This will build up over time to create the fairly large motion you see in the video.
Any other ideas?
Edit: So yeah, what everyone else has already said.
(
JeremyBeadlesHands, Sat 12 Jan 2013, 12:41,
Share,
Reply)
I reckon that there *could* also be a second effect that would cause oscillation, to do with the change in air pressure above and below the candle created by the difference of cold and warm air. Hot air expands and so the bit of hot air above the candle would provide a *lower* pressure and therefore push less down on a given side of the candle than the cold air below it would push up.
If you imagine that somehow side A has dipped down a little into the cold air below it, it will (initially) now have colder air on both sides, whereas side B that has moved up a little will have moved into the hot air above it, and so will have warmer air below it. So overall I'd think that maybe this would cause a greater push UP on the displaced side A, than the push UP that you get on the displaced side B, and so the candle would correct itself, most likely overshooting the level-position and continuting this cycle.
I haven't really thought about it hard enough to decide if that's really an effect though (for example, whether you would expect it to cause the candle to continue to oscillate or to quickly settle down). Needs testing. It's a shame you can't get a candle to burn in a vacuum, as sucking the air out of the vicinity would be a good test!
If it is indeed a factor then I'd reckon it wouldn't be nearly as important as the loss of mass that you rightly suggest.
(, Sat 12 Jan 2013, 14:06, Share, Reply)
nice idea
hadn't considered it, I assume it would have some effect, but I'm guessing the flame would provide some pretty fast heating action to the air in the vicinity so I don't know how transient this effect would be each time it oscillates.
(
JeremyBeadlesHands, Sat 12 Jan 2013, 14:23,
Share,
Reply)
But if you get out at the A end...
...too early and the system hasn't reached equilibrium and is still looping round the B end, that magic moss will have grown all over your eggs!
Or something similarly prime like.
(, Fri 18 Jan 2013, 22:36, Share, Reply)
Boyle's Law
If there were a change of pressure then, according to Boyle's Law, the operator would be overcome by an uncontrollable urge to yawn in order to equalise the pressure in his ears.
As this didn't happen I can't see how it ca be an atmospheric pressure thing.
(, Sat 19 Jan 2013, 5:32, Share, Reply)
PS: regarding the loss of mass due to the wax melting, I think that an important factor is also that when side A tips down, this allows the wax to drip off the recess that the flame has hollowed out, whereas side B is held up and so wax collects in its recess. Thus B is heavier than A and so we get a rotation. And then the situation is reversed.
A bit like if you had two buckets on the edges being filled with water from a non-spillable source - they empty each time they get tipped down.
(, Sat 12 Jan 2013, 14:09, Share, Reply)