Gradient Factors - What is Everyone Using?

[LFMarm]

If you want to avoid testing DCS on yourself, you can dive what world class expert on DCS dive themselves. Here is what they dive: Simon Mitchell 50/70-80, Neal Pollock 40/60-70, David Doolette 70/85.

I personally like to be more conservative and use a ratio of 0.83 between GF High and GF Low. Typically 50/60 and switch to 70/85 for bailout.

 

[EFX]

The maximum tolerable supersaturation already varies by compartment. Perhaps not the the ideal way (i.e., iso-risk), but finding better numbers requires testing.

Making them user-configurable also has the same problem as today's single GF: what do you set them to?

For any change we make, especially a change made to the models, we have to ask: What benefit do we get out of it. Setting GF's for 16 tissue compartments makes things more complicated and may have unattended consequences for our safety. We already don't know exactly how conservative a dive will be going from a GFHi of 90 to a GFHi of 70. Sure, no one will argue that the dive will be more conservative for DCS safety but how much more safe is it? A reduction of 90 to 70 suggests a 22% increase in safety but I doubt seriously that such a change will be linear. And now we're going to multiply that uncertainty by 16!

 

[100days-a-year]

Decompression " theory "
Correlation doesn't imply causation

 

[grantmac]

Having a slope from GFSlow to GFFast doesn't seem any more complex than GFLow and GFHigh, but again I'm not an expert so I tend to dive what the experts dive. Just because it hasn't been investigated doesn't mean it's not a valid thought however.
Research seems to indicate that slower compartment supersaturation has a greater effect on DCS compared to faster. Maybe ending the dive with the faster compartments at a higher M-value than the slower ones rather than simply setting a GF for the leading compartment.
But again I'm not an expert, it seems the theory keeps changing which is a good thing. Who knows where it'll go?

 

[inquis]

Maybe ending the dive with the faster compartments at a higher M-value than the slower ones

Yep, and this is already a characteristic of Buhlmann ZHL-16C. However since probably everyone uses GFs on top of that, the more the two factors differ, the less this is the case. (Hopefully you're getting an inkling of perhaps why the low GFlow values have fallen out of favor.)

 

[EFX]

Research seems to indicate that slower compartment supersaturation has a greater effect on DCS compared to faster. Maybe ending the dive with the faster compartments at a higher M-value than the slower ones rather than simply setting a GF for the leading compartment.

The faster TC's (tissue compartments) are already set for higher m-values throughout the entire dive. Take a look at the m-value tables for Buhlmann ZH-L16 in the paper "Understanding m-values" by Erik C. Baker, P.E.. Included are values for half-time (HT) and slope. Unless you're doing a very fast bounce dive the faster TC's rarely control the dive. This is because of their fast HT's. Those TC's on gas fast but they also off gas faster than slower TC's. Often TC's 1 and 2 off gas almost completely on the ascent.

The Buhlmann algorithm I'm familiar with applies the GF's to all TC's. The TC with the highest supersaturation becomes the leading TC. Based on the settings of this compartment and other dive data an NDL time remaining or stop depth and time will be calculated. So, there is no need to set a particular TC with the GF's.

 

[dmaziuk]

For anyone reading it: pay attention to the "Consistency of M-Values" bit.

 

[grantmac]

I'll have a read through that on duty tomorrow. Looks like a good resource for folks who want more than a passing understanding.

 

[exodusmaximus]

You are echoing the thinking from 15-20 years ago. This article is about 4 years old now. It summarizes more recent thinking.

Well, most my tech diving was done then, lol

 

[exodusmaximus]

Yes, which makes 30/70 less conservative than 50/70, according to all the research for over a decade. Deep stops increase the risk of DCS because slow tissues continue to on-gas during the deeper stops.

Ok, that makes a bit more sense. I guess 50/70, 50/75 is the way to go