Your objection does not apply to the situation we're discussing -- decompression diving.. The diver is obviously exposed to far more than the minimal supersaturation pressures you reference.
And your objection tends to work against you (i.e. against VPM). If we modified the simple integral supersaturation calculation to reflect this threshold supersaturation idea, then we'd simply disregard compartments that never exceeded a certain threshold supersaturation. So, for example, if the supersaturation pressure never exceeded, let's say, 400mb in a compartment we'd not allow that compartment to contribute to the integral supersaturation since, as you point out, there is some level that is harmless.
I've gone through that exercise and it invariably hurts VPM-B and tends to put more distance between VPM and GF which is why I never decided to actually show it. It's just one more thing you could object to and, as you note, VPM-B is already shown to have higher supersaturation exposure due to the deep stops.
The reason it hurts VPM-B is because the deeper stops allow more on-gassing in the intermediate and slow compartments. That additional on-gassing causes more compartments to exceed the threshold sooner. So you might find yourself using, for example 12 compartments in the integral supersaturation calculation for VPM-B, but only 10 for GF. Not for every profile obviously, since at some point all the compartments exceed the threshold just due to the dive being more substantial.
10 vs 12.... not a valid comparison then. The ZHL model gas tracking structure being used throughout here, is strictly a parallel cell design, with each cell as an independent stress. These cells cannot be summed together to create one large over all stress.
Models that are designed for serial or interconnected cell use (adding cells), have typical 3 or 4 cells max. These are widely spaced cells to prevent the overlap or duplication of information. These typically have further controls or multipliers to equalize or normalize values to a common base.
Your method does none of those. Your summing of 10 vs 12 is clearly not a valid or equalized method. What you do now, is without control, or validation. It is one big junk number.
My 4 ATA diver death situation, vs the weekend skier situation, highlights exactly the problems that your simplistic method creates. The same value for both conditions, but obviously the ski condition is harmless. However, your simple method has no way to differentiate the two, it does not recognize the good from the bad. It does not place any weight on the deathly situation, or eliminate the safe condition.
Stress in ascent occurs under elevated ambient pressure, but it has no weighted corrections to normalize values in the method.
The majority of your big number on the surface, comes from the harmless lower stress end, and therefore obliterates any important differences that occur at the early high stress end.
With these kinds of inbuilt errors and obfuscations, this simple method is not useful.
*************
The universally accepted measure of stress in a parallel design, is the peak value of each discrete cell at any given time. It does not permit the averaging or summing of these.
.
