I believe, I believe!!!

[wedivebc]

wedivebc, how about conducting tests shaken and unshaken, and dip tube purged and dip tube not purged, a 2 X 2 matrix of combinations to see which is the predominant effect, and posting the results?

Actually we sort of did that. Once we decided to see what was going on we left the tanks over night. Analyzed in the morning and all readings were low. We shook, rattled and Rolled the tanks Readings came out as expected. Second time round we blended then opened the valve for a second before analyzing which also gave us normal readings.

 

[WarmWaterDiver]

So either agitation or purge seems to be equally effective then?

The core issue is the dead leg in the dip tube for whatever the last gas was used to top off the tank?

 

[IceBergSlim]

I have found that if you put some really small ballbearings inside the tank you can shake it like a can of spray paint.

/not intended as sarcasim

 

[airmonkey]

Funnily enough I did my TDI nitrox course literally yesterday and this got mentioned and the instructor said it was true! (The shaking the tank..)

 

[wedivebc]

So either agitation or purge seems to be equally effective then?

The core issue is the dead leg in the dip tube for whatever the last gas was used to top off the tank?

I don't know about equally. I would say a small purge or quite a bit of shaking seems to give the same results. When we just gave it a couple of shakes the readings came up but not to the target we were expecting.

 

[miketsp]

I know a little about physics and gas and entrophy and the like. I'm absolutely amazed. And frightened.

This just goes to show the difference between theory and practice.

The last time we had a thread on this subject I chipped in saying this was ridiculous as according to Dalton's Law each gas has to expand to fill the available space. But then I went away and did some calculations on the diffusion rate of two gases at 200bar assuming that they start separated and how long would they take to mix.
The answer was scary - some hours.
At 200 bar the gases act like viscous liquids.
They will mix totally if left alone for long enough.
The mix should also be sooner if the fill rate is low as you will get a good mix (turbulence) before the contents get too viscous.