My Journey into UTD Ratio Deco

[Kevrumbo]

And to answer your question about what I'm comfortable with, I dive a 70/70 GF and would try 100/70 if Shearwater would ever implement my change request.

If you wanted to actually do this ascent [GF 100/70] you could do it by setting the shearwater to 99/99, do the ascent through to 20ft(?) then reset the GF high to 70%. You'd then discover that you missed stop time at 30 and/or 40ft though since the GF99 had you ascend past the newly entered GF70. So you'd be in a ceiling violation. GF100/70 basically creates a situation where you are continually exceeding the 70% safety buffer on the m-value on at least one tissue (maybe more) throughout the whole ascent. That's why I said it wasn't mathematically possible (well maybe it is, but shearwater is not going to write code that creates a continuous deco ceiling violation).

The other option is to wait until the "GF99" indicator reads 70 at an arbitrarily chosen shallow deco stop depth profile i.g. 6msw on Oxygen; or perhaps deeper at 9 or 12msw while still breathing on Nitrox50. . .

What does the “GF99” display value mean? - Shearwater Research

 

[dmaziuk]

Dive rite inaccurately described that. It is how close you are willing to venture at any point in the ascent.

By that definition the only mathematically possible setting is GFHi = GFLo, i.e. one gradient factor. Which is perfectly fine as long you don't want to change the slope of your M-value line, only shift it closer to ambient pressure line. (But it's not what the pictures, incl. ones in Eric Baker's articles, actually depict.)

 

[rjack321]

By that definition the only mathematically possible setting is GFHi = GFLo, i.e. one gradient factor. Which is perfectly fine as long you don't want to change the slope of your M-value line, only shift it closer to ambient pressure line. (But it's not what the pictures, incl. ones in Eric Baker's articles, actually depict.)

GFlow is the bottom of your allowable offgassing window
GFhigh is the top
(Erik Baker and Doppler both illustrate this, Diverite does too but the text is not accurate)

GF 100/70 is saying the bottom is above the top, which is why Shearwater (and Ross via multideco) won't ever be able to write anything like this. Another way to say it is you have cleared your 80ft stop and should move up to 70ft, yet once you arrive at 70ft you have violated your ceiling. It's an illogical mathematical argument.

 

[rjack321]

The other option is to wait until the "GF99" indicator reads 70 at an arbitrarily chosen shallow deco stop depth profile i.g. 6msw on Oxygen; or perhaps deeper at 9 or 12msw while still breathing on Nitrox50. . .

What does the “GF99” display value mean? - Shearwater Research

Which ends up being a 70/70 profile

 

[huwporter]

Dive rite inaccurately described that. It is how close you are willing to venture at any point in the ascent.
See Erik Baker's original work on this topic
http://www.ddplan.com/reference/mvalues.pdf

But in the interest of discussion, if implemented, it would look like this figure.

This figure might make sense visually, but it requires your computer to predict the future and decide (at some point in the ascent) that 100% GF is no longer ok and it should start dialing the maximum overpressure gradient back so that no pressure group exceeds 70% by the time you surface.

If you wanted to actually do this ascent you could do it by setting the shearwater to 99/99, do the ascent through to 20ft(?) then reset the GF high to 70%. You'd then discover that you missed stop time at 30 and/or 40ft though since the GF99 had you ascend past the newly entered GF70. So you'd be in a ceiling violation. GF100/70 basically creates a situation where you are continually exceeding the 70% safety buffer on the m-value on at least one tissue (maybe more) throughout the whole ascent. That's why I said it wasn't mathematically possible (well maybe it is, but shearwater is not going to write code that creates a continuous deco ceiling violation).

I'm not sure I agree - the way that gradient factors are implemented in most deco software (i.e, "start the first stop as soon as any tissue is about to exceed the GF Lo value, then move in equal gradient factor steps so that after the last stop no tissue exceeds the GF Hi value on surfacing") then it's perfectly possible to calculate a GF 100/70 ascent - it's just the GF steps are negative. Example:

I'm not saying that I'd like to dive this profile, but I don't think there is anything in the maths that prevents GFHi (= "gradient factor at the surface") from being a lower % value than GFLo (= "gradient factor at first stop").

 

[Michael Guerrero]

What @huwporter said. What this means in practical terms is that you drive toward the m-values of your leading tissues at the beginning of deco (which will be your fast tissues) and then slowly step back throughout the deco as your mid and slow tissues take over as the leading tissues, surfacing at the GF hi value.

 

[dmaziuk]

I'm not certain what the "bottom of allowable off-gassing window" would mean in that context: what's wrong with ambient pressure line being the bottom?

 

[elmo]

Post-fail

 

[elmo]

Sorry for the post quote-fail above. To avoid putting words into @rjack321 's mouth. I've repeated it correctly below.

GFlow is the bottom of your allowable offgassing window
GFhigh is the top
(Erik Baker and Doppler both illustrate this, Diverite does too but the text is not accurate)

Actually Erik Baker says "GF lo generates first stop" and "GF hi (surfacing value) maintains safety margin". See fig 4 (last page) of
https://www.shearwater.com/wp-content/uploads/2012/08/Deep-Stops.pdf
It is clear Baker's objective was to use gradient factors to modify a Buhlmann profile to generate deep stops, which is why you almost always see gf-lo as less than gf-hi, but there's no reason why the same method can't be used to extend shallow stops without inducing deep stops. I am not advocating for that but mathematically it works just fine. Subsurface and GUE DecoPlanner both let you do it.

 

[rjack321]

That is what 30/85 was originally trying to achieve deepstops yes.

Re: the Decoplanner output. This is just linearizing the ascent and scaling the 100 down to 70 across the deco range (in theory correct but...) Implementing this means you would be forced to keep ascending once your GFlow is not longer 100 (or any of the successively smaller values because GFlow is implemented as the bottom of the offgassing window). So you would forced to ascend to the next stop in order to get offgassing credit on at least one compartment and you would get no offgassing credit at all for any compartment less than GF100 - all of which makes no sense. Assuming you could even make a 10ft step at 10ft/min without violating a ceiling - although you probably can for modest BTs.

@Kevrumbo 's idea of diving 99/99 and letting the GFhigh degrade to 70 later in the profile is going to be the closest you are going to get to a workable approach. Although all of this is all pretty far off from a RD profile - those are roughly 5/95.