New deco planning tool

[dmaziuk]

First stop is at 10 ft. Subsurface puts NDL at 100 ft on air at 5 mins. Staying for 6 mins has the ceiling at 2.4 ft when leaving the bottom and 1.9 ft when arriving at the 10 ft stop.

I understand you've setup a scenario where the tissue pressure is below the GFL-tolerated amount at the surface (therefore a negative GFL-ceiling) but above the GFH-tolerated amount. No argument that mandating GFL<GFH avoids this; however, Subsurface can clearly deal with it. The code is freely available if anyone is interested to see how they handled it.

Yes, you can cap it at calculated or 3 msw, whichever is deeper, and it is only an issue near the NDL, but still: mathematically it does not make sense. And considering that the rationale for GFs was to make the first stop deeper, not shallower, logically it didn't make sense at the time either.

 

[David Carron]

Yes, you can cap it at calculated or 3 msw, whichever is deeper, and it is only an issue near the NDL, but still: mathematically it does not make sense. And considering that the rationale for GFs was to make the first stop deeper, not shallower, logically it didn't make sense at the time either.

Again, I'm not seeing specifically what doesn't make sense to you mathematically. In a dive computer the objective function used in the search for the deepest stop simply answers the question "if stop n were the deepest stop, would immediate ascent to stop n - 1 be possible" which makes total mathematical sense for all values of gflo and gfhi.

Although the original intent of GFs was to modify zhl to reproduce deeper stop ascent profiles, that does not imply that "shallower stop" ascent profiles do not make physiological sense; many entries in the military mn90 tables can only be reproduced with gflo > gfhi

 

[inquis]

Again, I'm not seeing specifically what doesn't make sense to you mathematically.

I can certainly see @dmaziuk's point, it's the first stop that is in question. (Thereafter, it's just as you describe.) The simple implementation looks for the depth at which the GFL line crosses the (max) tissue saturation pressure. In this case, though, that would cross at a negative depth, indicating one can surface... except for tissue tension is above the GFH, meaning you can't surface.

That simple approach to set the initial stop depth obviously doesn't always work. I haven't looked at the Subsurface approach in depth, but the numbers seem quite reasonable. There are no discontinuities as bottom time is increased and ceiling increases monotonically. If one lowers GFL, the ceiling increases monotonically and matches the typical value when GFL=GFH. They've clearly identified a better approach to setting the initial stop depth that works regardless of whether GFL is lower or higher than GFH.

 

[David Carron]

I can certainly see @dmaziuk's point, it's the first stop that is in question. (Thereafter, it's just as you describe.) The simple implementation looks for the depth at which the GFL line crosses the (max) tissue saturation pressure. In this case, though, that would cross at a negative depth, indicating one can surface... except for tissue tension is above the GFH, meaning you can't surface.

That simple approach to set the initial stop depth obviously doesn't always work. I haven't looked at the Subsurface approach in depth, but the numbers seems quite reasonable. There are no discontinuities as bottom time is increased and ceiling increases monotonically. If one lowers GFL, the ceiling increases monotonically and matches the typical value when GFL=GFH. They've clearly identified a better approach to setting the initial stop depth that works regardless of whether GFL is lower or higher than GFH.

This is certainly a simple implementation, but it is not applicable in the context of a dive computer which, at the very least, requires you to take into account the loading during the ascent before comparing to any calculated limits and, as you have noted, may not be applicable in the context of a simulator at all.

It turns out that the approach you describe, anchoring Plo at the immediate ceiling, may actually result in stops becoming shorter or possibly the first stop becoming shallower as diver load increases:

https://fr.deeplysafelabs.com/assets/videos/perdix2_720_fr.webm

 

[dmaziuk]

They've clearly identified a better approach to setting the initial stop depth that works regardless of whether GFL is lower or higher than GFH.

Then it's not ZH-L + GFs. It may be Buhlmann+Baker+Helling or something, but it's not quite what Baker devised.