Have I understood the basics of decompression theory, GF99 and SurfGF?

[tursiops]

DSAT has been tested for repetitive NDL and is approximated by a GFHigh of 95. There's no way I'm doing deco at that setting.

Additionally, since there is always the chance of unintended deco, I avoid GFHi=95 just in case.

 

[David Carron]

As I recall back when @huwporter ran the numbers, you have to dive 10/90 to trigger a deco stop -- after "just" exceeding your NDL -- at 9 msw. For anything less silly you'll get at most 6 msw that is practically indistinguishalbe from the 5 msw where many of us are doing our safety stop anyway.

This is incorrect. Many computers do not calculate NDL correctly but, for those which do, there will always be a deco stop as soon as NDL drops to zero.

 

[dmaziuk]

This is incorrect. Many computers do not calculate NDL correctly but, for those which do, there will always be a deco stop as soon as NDL drops to zero.

I suspect you misunderstood: the point is there is a discontinuity in the model when NDL counts down to 0. Gradient Factors can exacerbate it and generate a first stop that is too deep and/or too long for the newbie user at the end of the dive, with little gas left.

However back when we looked at the numbers, the conclusion was that it'd take something like 10/90 to produce that effect; common GF settings would generate a stop at 6 or 3 msw, i.e. the same depth as the "safety" stop we normally do. So that concern is pretty much theoretical only.

Oh, and "many" computers don't calculate NDL correctly? SRSLY?

 

[inquis]

the point is there is a discontinuity in the model when NDL counts down to 0.

Are you talking about a jump in the ceiling? I don't see this discontinuity in the Shearwater ceiling. It increases from zero after NDL goes to 0, resulting in an initial stop depth of 3m after roundup.

 

[inquis]

Many computers do not calculate NDL correctly

Any concrete examples?

 

[dmaziuk]

Are you talking about a jump in the ceiling? I don't see this discontinuity in the Shearwater ceiling. It increases from zero after NDL goes to 0, resulting in an initial stop depth of 3m after roundup.

You can set your GFs so that it won't be the case.

 

[inquis]

You can set your GFs so that it won't be the case.

That may be the case for simplistic implementations (e.g., Baker's original concept), but have you seen it happen with anything from the last decade or two, e.g., Shearwater's implementation? Does Subsurface show this discontinuity?

(Baker's implementation ignored off-gassing during ascent, and I agree that could result in a discontinuity in the ceiling when NDL hit 0.)

 

[dmaziuk]

That may be the case for simplistic implementations (e.g., Baker's original concept), but have you seen it happen with anything from the last decade or two, e.g., Shearwater's implementation? Does Subsurface show this discontinuity?

(Baker's implementation ignored off-gassing during ascent, and I agree that could result in a discontinuity in the ceiling when NDL hit 0.)

Open subsurface planner and plan a dive to 40 msw on 10/80 on air. When the stop shows up at ~6:40, see how deep it is.

 

[inquis]

Open subsurface planner and plan a dive to 40 msw on 10/80 on air. When the stop shows up at ~6:40, see how deep it is.

Thanks for that, and I'd say that's tied to the ceiling calculation using current tissue tension (see core/deco.cpp/#220 in the source). Shearwater doesn't do that.

In the bigger picture, though, I think you're right that it's a non-issue for realistic GF choices.

 

[dmaziuk]

Thanks for that, and I'd say that's tied to the ceiling calculation using current tissue tension (see core/deco.cpp/#220 in the source). Shearwater doesn't do that.

In the bigger picture, though, I think you're right that it's a non-issue for realistic GF choices.

Well... if Shearwater's math doesn't work the same way Baker's does, they're either not running ZHL-16+GF, or they live in the Evil Parallel Universe where 2 + 2 = sqrt( -1 ).