Why 2 gradient factors ?

[rjack321]

The problem is what is "aggressive" when different tissues are considered? "Protecting" the fast tissues (i.e., at depth) is less aggressive -- for them -- but is actually more aggressive for slower tissues. I believe researchers feel there is be a "sweet spot" that is overall best for all tissues, but an evidence-based assessment of precisely where that is lacking thus far. All we have are the broader indications: Doolette uses GFLow = 0.83*GFHigh (yielding the same degree of supersaturation for all tissues). [1] Mitchell has condoned a GFLow range of 55-75. [2] Baker uses a GFLow between 50 and 60 [sorry, no specific reference].

[1] Gradient Factors in a Post-Deep Stops World, 5/29/2019 (retrieved 4/21/2022)
[2] What is optimal decompression, www.youtube.com/watch?v=nIO9qI5XODw, 2020

This is what I was getting at when I posted the "be careful making assumptions about what is/isnt a 'conservative' turn of the GF dial." Agree on the sweet spot issue, just not sure that is even the same spot across different people - best guess is it probably isn't! yay for biological variation!

 

[Nelson_]

In the sense that lowering the "allowed overpressure" value adds conservatism. Remember that overpressure drives gas exchange so reducing it increases your off-gassing time. And with multiple compartments, some of them may still be on-gassing and the whole "safety" thing becomes... not quite that simple.

Agree!

 

[dmaziuk]

Agree!

To answer your question literally: because it takes two points to draw a line.

Since Workmann, the M-value is a line defined as M0 + P * dM where M0 is the surfacing M-value, P is the ambient pressure, and dM is the slope. Two independently adjustable gradient factors allow you to change the slope -- which way you want to change it, and why, is entirely up to you. Baker's original motivation was to approximate the then-popular "deep stop" profiles.

 

[EFX]

To answer your question literally: because it takes two points to draw a line.

Since Workmann, the M-value is a line defined as M0 + P * dM where M0 is the surfacing M-value, P is the ambient pressure, and dM is the slope. Two independently adjustable gradient factors allow you to change the slope -- which way you want to change it, and why, is entirely up to you. Baker's original motivation was to approximate the then-popular "deep stop" profiles.

I was going to post something similar but you beat me to it. Baker didn't want to just make the dive more conservative at surfacing which could be done with just GFHi. He wanted to provide an additional setting (GFLo) in order to adjust the slope of the TC m-value limit throughout the entire dive. This gets implemented in DC's by letting GFLo determine the depth of the first stop. The first stop and the surface provide the terminal points in the profile which gives the slope of the line defined by tissue pressure and depth (or the pressure at depth). You can see the relationship between GFLo, GFHi, and current GF in the graph below.

The number of additional stops needed to arrive at the surface is divided into the difference in GF's. This value, the GF increment, is added to GFLo and then to each shallower stop until the surface is reached so that current GF equals GFHi.

For example, if GFLo/GFHi is equal to 60/80 and the first stop is calculated to be 40 ft. Using 10 ft between stops gives a GF increment value of (80 - 60) / 4 = 5. The ascent profile then looks like this:

Stop depth: 40, 30, 20, 10, 0.
Current GF: 60, 65, 70, 75, 80.

 

[Nelson_]

I was going to post something similar but you beat me to it. Baker didn't want to just make the dive more conservative at surfacing which could be done with just GFHi. He wanted to provide an additional setting (GFLo) in order to adjust the slope of the TC m-value limit throughout the entire dive. This gets implemented in DC's by letting GFLo determine the depth of the first stop. The first stop and the surface provide the terminal points in the profile which gives the slope of the line defined by tissue pressure and depth (or the pressure at depth). You can see the relationship between GFLo, GFHi, and current GF in the graph below.

The number of additional stops needed to arrive at the surface is divided into the difference in GF's. This value, the GF increment, is added to GFLo and then to each shallower stop until the surface is reached so that current GF equals GFHi.

For example, if GFLo/GFHi is equal to 60/80 and the first stop is calculated to be 40 ft. Using 10 ft between stops gives a GF increment value of (80 - 60) / 4 = 5. The ascent profile then looks like this:

Stop depth: 40, 30, 20, 10, 0.
Current GF: 60, 65, 70, 75, 80.

View attachment 765063

Very clear! But, still no explanation why we use 2 different GF values. Forget Pyle and bubbles, why don’t we use the same value for both?

 

[scubadada]

Very clear! But, still no explanation why we use 2 different GF values. Forget Pyle and bubbles, why don’t we use the same value for both?

Gradient factors were used to adjust Buhlmann to mimic the profile of the, then popular, bubble models. I believe we are still recovering from that, with GF low moving up toward GF high. As per the Doolette article, there may be reasons to keep the GF low less than the GF high.

If the goal of a gradient factors was simply to make the algorithm safer/more conservative, this could have been achieved with a single factor that shifted the M-value line to the right a uniform amount like GF 85/85. No stop dives are made more conservative with only the GF high changing the allowed exposure compared to the native algorithm. GF low changes the slope of the line in addition to the right shift and decrease in intercept provided by the GF high

Keep in mind the derivation of the Buhlmann algorithms. ZH-L16B had the a coefficients decreased from ZH-L16A, and thus the Mo decreased for compartments 6, 7, 8, and 13. This shifts the M-value line to the right as the slope of the line remains the same with the lower intercept. ZH-L16B is suggested for use in tables. ZH-L16C underwent further changes with additional decreases in the a coefficients and Mo for compartments 5-12 and 14. ZH-L16C is suggested for use in dive computers and is the version used by most of us. So, ZH-L16C is more conservative in the medium and many of the slow compartments compared to the original algorithm, ZH-L16A.

The decompression algorithms are all models that have successfully decreased the risk of DCS with diving. Dissolved gas models, bubble models, and several other models that are also out there, may all have some features that mimic actual human physiology.

 

[dmaziuk]

https://www.shearwater.com/wp-content/uploads/2012/08/Deep-Stops.pdf

Understanding M-Values | Erik C. Baker

Understanding M-Values By Erik C. Baker In conjunction with an array of hypothetical "tissue" compartments, gas loading calculations and M-values compose the major elements of the dissolved gas or "Haldanian" decompression model. Through the use of widely-available desktop computer programs...

scubatechphilippines.com

 

[inquis]

still no explanation why we use 2 different GF values.

Because the goal was to shift AND tilt the M-Value line. That requires 2 numbers. The tilt gave more time at deeper stops, which was in vogue at the time.

You're welcome to set them equal, if you like, achieving only a shift. However, when the preeminent researchers run different values, there may be good reasons to follow suit.

 

[Tracy]

Very clear! But, still no explanation why we use 2 different GF values. Forget Pyle and bubbles, why don’t we use the same value for both?

Some people do.

 

[tursiops]

Very clear! But, still no explanation why we use 2 different GF values. Forget Pyle and bubbles, why don’t we use the same value for both?

You can. Some do. Personal choice and need.
Baker could have had just one GF that applied to the entire depth range. He provided two so you could adjust the "conservatism" as you call it, to be different at depth than near the surface.