Mark F wrote:If you managed to fully fill a canister with it tilted then you must have created a vacuum in the one being refilled. Had it held butane and you froze it before refilling? That could explain it.
I'm not sure what you mean by "vacuum". The vapor in these canisters will generally be at saturation. At the temperature of my freezer that would be around 50 kPa for n-butane and 200 kPa for isobutane. If by vacuum you mean less than 1 atmosphere then regular butane would fit that definition. But the atmospheric pressure is irrelevant when transferring between two canisters as it's a closed system. What matters is the pressure differential.
Anyway, I was using Jetboil Jetpower canisters so it should have been primarily isobutane, not n-butane, along with some percentage of propane. For one test I mixed the contents of two canisters into one of those larger 450g canisters, then transferred it back so that I'd have pretty close to identical mixtures in both 230g canisters.
I think your notion of an "air" space preventing transfer is incorrect. It's not air, it's fuel vapor, and it will increase or decrease in volume dynamically. I think the head space in the upper canister is more significant for a tilted setup as it dictates when the transfer switches between liquid feed and vapor feed. Vapor feed is much slower as it requires a lot more heat transfer. And it fractionates the mixture as well. Even with that, I am close to certain that I could transfer an entire full canister UP into an empty canister, simply by providing enough of a temperature difference.
Twice your method appeared to work for me with the transfer slowing to a stop close to 100% full. Was that a coincidence? One of those times I decided to re-chill the receiving canister and that's when it filled completely, to 125% full. So I'm still not sure about this. I presume you have successfully used this method yourself. What specifically did you do? It may be that the details are as important as the angle.