Riding deco ceiling on ascent

[Dr Simon Mitchell]

Hello @inquisit and others,

DHM published a paper on this subject in March. It was early released by the authors and so is freely available on PubMed Central. I have also attached it here.

Those of you interested in the physiology and medicine of diving should consider becoming associate members of SPUMS or EUBS so you get access to all papers on release (and the SPUMS ASMs are terrific events with diving and science rolled together). Journal articles are embargoed for a year unless the authors pay an immediate release open access fee as has happened with the one I have attached.

The DHM site with access to previous journal articles is a bit of a treasure trove also.

Simon M

 

[BlueTrin]

It's what Suunto calls the code running on low-end computers. Maybe also Cressi and Mares though I'm not sure about them. Nobody knows what it is but given the official explanation: these devices don't have the CPU oomph to run the full "iterative" RGBM, one could make an educated guess: lookups in pre-computed tables and simplified functions that work "close enough" within the device's intended range.

Thanks for explaining.

 

[Mod63]

It's what Suunto calls the code running on low-end computers. Maybe also Cressi and Mares though I'm not sure about them. Nobody knows what it is but given the official explanation: these devices don't have the CPU oomph to run the full "iterative" RGBM, one could make an educated guess: lookups in pre-computed tables and simplified functions that work "close enough" within the device's intended range.

It's less "low-end" and more "low-power consumption" that prohibits the use of a full RGBM calculation (side note: even most ZHL based computers did/do not fully calculate but break the calculation loop once the result is good enough). And no lookups, it's more a calculation of a fudge factor based on the RBGM model that is then applied to a traditional Haldanian model with compartments, think GF but calculated differently. There is a paper from Suunto on the current version which the call Fused RGBM with some basic explanation

 

[rongoodman]

The only dive computer I remember ever offering the full iterative RGBM model was the Liquivision Lynx with some after-market software.

 

[inquis]

I have also attached it here.

Thank you very much for that paper. I thought it very well written and thorough with regard to their goal. So the big question in my mind remains:

if tissues can support the gradient present when first arriving at a discrete stop, can they not continue to support that level of offgasing?

It seems that answering that would require an empirical study. In light of the limited gains and the associated difficulties in performing a continuous ascent, I doubt there would be much urgency to investigate.

 

[inquis]

I do find it ironic that Deep Stop proponents desire to "protect the fast tissues", while the opponents (based on actual testing) say, "no, because that overloads the slow tissues". Meanwhile, in this ceiling v staged issue, the status quo of staged protects the fast tissues and leaves the slower tissues with relatively greater loading.

As mentioned early on, I think the execution advantages to the staged approach greatly outshine any potential advantages to ceiling ascents, but I thought the consequence to loading interesting nonetheless.

 

[dmaziuk]

It seems that answering that would require an empirical study. In light of the limited gains and the associated difficulties in performing a continuous ascent, I doubt there would be much urgency to investigate.

I think when your leading TC switches, you may have to stop anyway. Also since half-times differ, you'd need to match you "continuous ascent" rate to the leading tissue. You could end up with "interesting" ascent profiles. Checking this doesn't need an empirical study: one could run a few simulations and see what ascent profiles would look like if you tried.

Edit: I wrote that before reading the paper -- thank you @Dr Simon Mitchell -- that's exactly what the authors did. In Fig. 2 you can see the change in ascent rate (black line) when the green line hits the ceiling, and stop at TC switch at 9msw.

 

[dmaziuk]

(side note: even most ZHL based computers did/do not fully calculate but break the calculation loop once the result is good enough).

"Fully calculate" what? There is no "good enough" ceiling: there's either a ceiling or there isn't one.

The only thing you "break" on is NDL: there is no practical point in calculating past, say, 99 if you have a 2-digit display. But NDL is not what ZH-L is really for to begin with.

 

[dmaziuk]

I do find it ironic that Deep Stop proponents desire to "protect the fast tissues", while the opponents (based on actual testing) say, "no, because that overloads the slow tissues". Meanwhile, in this ceiling v staged issue, the status quo of staged protects the fast tissues and leaves the slower tissues with relatively greater loading.

Riding the ceiling keeps fast tissues "closer to M-value longer" but it's the GF85 M-value. If you want to protect them more, lower your GF. This paper also speculates that "... later these fast tissues become
undersaturated and any bubble that may have formed will shrink" which, if you believe bigger bubbles lead to more inflammation, would end up being more "protective".

 

[Mod63]

"Fully calculate" what? There is no "good enough" ceiling: there's either a ceiling or there isn't one.

The only thing you "break" on is NDL: there is no practical point in calculating past, say, 99 if you have a 2-digit display. But NDL is not what ZH-L is really for to begin with.

Fully calculate the ceiling. You do not care if it is 17.95m or 18.05m -> 18m for you deco stop is good enough.