"re-assessing deco profiles and deep stops", plus other bits..

[rossh]

Absolutely (although I believe you meant in vivo; in vitro is in a petri dish).

Consider the Big 2 (Buhlmann+GFs and VPM). Buhlmann arbitrarily models 16 (12, 8) compartments, assumes that an exponential law governs ongasing and offgasing in each of these (the on/offgassing model) and further assumes a linear law specifies what minimum ambient pressure can be tolerated (the deco model). Testing was used to adjust all the constants to make the model "safe" (for a certain value of safe )(although ZHL16A used formulas to derive the a and b deco model constants from the on/offgasing model constants, if anyone knows how those formulas were arrived to, I'd be interested). GFs are just a dodgy hack to fix up the deco linear law to do deep stops (put down the flamethrower, that's what I'm diving too ). VPM uses bubble modeling for the deco model; the on/offgasing model is the same.

Yeah, blah blah blah, quite right, sorry. The point here is that none of that even pretends to describes what's happening to the body. Never mind a body in diving conditions, working, or not, getting cold, getting dehydrated and so on. Cheers,

Matthieu

This is a good topic, as I feel many with an interest in modelling, don't appreciate the relevance of this.

The Deco models in use today do not try to reproduce the physiology. Too many interested people fail to understand this. A deco model tries to model the danger zone and creates limits here.

It is currently impossible to reproduce the human circulation system in a math model. If such a thing was invented, it would be put to important use in medicine with the diagnosis of a 1000 human diseases and ailments.


Deco models are built around parameters we can measure in the water. That is: pressure and time. Many other variables affect the required deco time, but we can't measure or use them in a dive. As a diver in the water, we can only measure pressure and time which leaves very little for a deco model to work with.

If we could measure more variables, like body temp, circulation rates, and actual on / off gassing, then we could make our deco models more capable. For that I think we would need sensors and probes on the body, communicating to a computer. Now - here is the interesting part of such a capable model; For most of the dives we do, it would suggest LESS time than we do now. That's simply because our models and calibration we use today already have a safety buffer that fits the average diver and average situation. A more accurate model calculation would give less time. Are today's divers comfortable doing less time when its appropriate? Probably not.


As pointed out, many older models use only averages of time and pressure to track exposure. There are many versions of this approach - with a varying amount of cells in series or parallel. Then they typically apply a set of constant or factors to determine ascent limits. These simple model limits are fitted values to match observations. The problem here is the simple time + pressure averaging coupled with factors / constants, will quickly gets out of step with the physiology. Hence the GF is applied to patch up ZHL. RD uses several rule sets to compensate.

(Mark, this is why ZHL will never make you happy - the further you go to the extremes of the profiles, the further you venture from the fitted calibration zone).

Bubble models (VPM/RGBM) go one step further, by applying the science of microbubble behavior to the time / pressure averaging. All dives involve microbubbles, so it seems this approach is a lot closer to the physiology. Bubble models seem to span the diving spectrum without need for adjustment.


Another model concept that I thought showed promise was the DCIEM. Here they made ultrasound observations of microbubbles in test dives. From that they created tables that gave consistent levels of microbubbles. Interestingly, this also aligns nicely with the bubble model approach too.


Regards rossh

 

[anarchista]

if anyone knows how those formulas were arrived to, I'd be interested). GFs are just a dodgy hack to fix up the deco linear law to do deep stops (put down the flamethrower, that's what I'm diving too ). VPM uses bubble modeling for the deco model; the on/offgasing model is the same.

Such easy removal are carried out on one year of physics at Warsaw University.

Forum AKP Krab AGH • WyÅ***8250;wietl temat - Meandry dekompresji

Yeah, blah blah blah, quite right, sorry. The point here is that none of that even pretends to describes what's happening to the body. Never mind a body in diving conditions, working, or not, getting cold, getting dehydrated and so on.

Jones decompression model maps well to push a perfusion model.
Documentation will be completely indigestible.

Rebreathers Poland :: Zobacz temat - "Proste" pytanie

IMHO, largely, yes.

There's a thing that's sometimes forgotten: the ppO2, in and by itself, is irrelevant to deco (so long as it's above 0.15 bar or so, that is ). It's important however because, oxygen being metabolised, you don't need to decompress it. So whatever pressure is taken by the oxygen is taken off the the rest, the inerts, which we do need to decompress. The higher the ppO2, the lower the ppN2 and ppHe, and therefore the less ongasing or the more offgasing.

Considering ongasing on the bottom, say 60m, 7 bar. With a ppO2 of 1.3, ppInerts will be 5.7. With 1.2, it will be 5.8. That's less than 2% extra. Not going to change the deco obligations all that much (of course "not all that much" is a relative concept).

Considering then, what you were saying, the ascent, say from 1.3 at 60m straight to 21m. If the ppO2 is kept stable at 1.3, ppInerts will be 1.8 bar. If, on the other hand, oxygen is lazily added and the diver gets there at 1 bar ppO2, ppInerts will be 2.1 bar. Okay, bit of a difference. But, compared to what it was on the bottom (5.7), you lost 3.9 bar in the first case and 3.6 in the second. Most of the gains with regards to ppInerts during the ascent are due to the ascent itself, not the ppO2. And it's not all that long, anyway, so again it won't change your deco obligations all that much.

When it's rather more important to have a high (and stable) ppO2, is when doing the stops.

IMHO, I'm not qualified to do any of this on CCR, it's all theory, and all that.

This, unfortunately, is pure nonsense.
Few people know but it's possible oxygen DCS.
(Review Book:
"Possibility of Selection Decompression for Diving Apparatus type AMPHORA" 2012 y.
"Possibility of Selection Decompression for Diving Apparatus type CRABE" 2011 y. know better modelers from our modern perspective.)
Level ppO2 in decompression has an effect on the rate of off-gassing of inert gases. Unfortunately, the action is too complicated physiology too high ppO2 limited perfusion (model Jones)

In such situations, showing the specific tissue saturation equations we talk about specifics. A dozen free intervals have low resilience to be controlled decompression below the surface.
Janos in highly simplified version is presented. #21 (permalink)

P1=12,4 + (25,91 - 12,4)(0,5^(2/1,88) = 18,86
P2=12,4 + (32,08 - 12,4)0,5^(2/3,02) = 24,83
P3=12,4 + (37,00 - 12,4)0,5^(2/4,72) = 30,73
P4=12,4 + (39,68 - 12,4)0,5^(2/6,99) = 34,77
P5=12,4 + (39,8 - 12,4)0,5^(2/10,21) = 36,32
P6=12,4 + (37,44 - 12,4)0,5^(2/14,48) = 35,15
P7=12,4 + (33,13 - 12,4)0,5^(2/20,53) = 31,77
P8=12,4 + (27,75 - 12,4)0,5^(2/29,11) = 27,03
P9=12,4 + ( 22,28 - 12,4)0,5^(2/41,2) = 21,95
P10=12,4 + (18,03 - 12,4)0,5^(2/55,18) = 17,89
P11=12,4 + (14,83 - 12,4)0,5^(2/70,69) = 14,78
There is an additional saturation of these compartments.
P12=12,4 + (12,10 - 12,4)0,5^(2/90,34) = 12,104
P13=12,4 + (9,80 - 12,4)0,5^(2/115,28) = 9,83
P14=12,4 + (7,86 - 12,4)0,5^(2/147,42) = 7,90
P15=12,4 + (6,29 - 12,4)0,5^(2/188,24) = 6,33
P16=12,4 + (4,99 - 12,40,5^(2/240,03) = 5,03

You emmbee present it in a version less than one year physics student at any European university.

greet rc

 

[Philippe GERIN]

When I run GF,s 15/85 I get out of the water feeling fine, If I run buleman without gf's I get out the water feeling like shit!!
very unscientific!

But real life experience. How about this one: when I run SP=1.1, I feel fine but if I run SP=1.3 I feel like shit. This is for standard dives, not deeper or longer dives. I run GF 30/85 by the way.
If anyone could shed some light, I'd be happy to learn.
Best
Philippe

 

[rossh]

Now, if you're looking for studies that suggests that deep stop may not be good, which wasn't what I said, but anyway, here you go:

• Empirical evaluation of the efficacy of deep stops in air decompression dives. (USN)

Maybe now its time to expose some of the details behinds these.

The Navy test - did NOT use any deep stops! If you take away the paper's Title and conclusions, and take a good look at the study for yourself, you might come to a different conclusion.

* No deep stops used in the test -at all - none! They instead tried to simulate them by working its tissue model backwards. They took their BVM model (navy use only - not deep by our standards) and reversed into a shallow profile shape that would simulate the same tissue effects of the slightly deeper profile. All this artificial fiddling was done to get the profile to finish at the same time as the base control test model (VVAL18). They instead should have let the BVM make its own profile naturally and accept the different finish times, but they didn't. So this is why the test profile ended up with an unnatural multilevel profile (and why the test went of the rails). They had a multilevel shallow profile - not a deep profile, and no deep stops to be seen anywhere.

* Next - and this is the kicker; the test was stopped because the base line control model (VVAL18) - was about to drop out the bottom of the test limits and invalidate the entire project. That's why it stopped - not because of the BVM model. The artificially simulated BVM profile was right down the middle of the test parameters (3 to 7%) exactly where it was calibrated to be. The test was stopped because the base line calibration data from its VVAL18 - the basis of many other navy testing projects - was about to invalidate this project by dropping below 3%. I'll say it again - this test stopped because the VVAL 18 control model datum, was 5 dives short of mandatory invalidation of the entire project!

A note on Navy testing risk rate: it is done with exaggeration (cold / exercise) to force DCS, so they can better measure it. That's why you get unrealistic risk numbers like 5% out of navy data - the risk numbers need to be normalized back to real world values.

So why did the navy label it deep, when there was none? <removed by author>

Regards rossh

 

[Mark Chase]

But real life experience. How about this one: when I run SP=1.1, I feel fine but if I run SP=1.3 I feel like shit. This is for standard dives, not deeper or longer dives. I run GF 30/85 by the way.
If anyone could shed some light, I'd be happy to learn.
Best
Philippe

You have a low tolerance to high PP02s

I have run two hour dives on 1.4 finishing on pure 02 and i feel fine but even I find over 3.5 hours on high PP02 I start to feel crap flue like symptoms and suffer chesty symptoms post dive.

This is build up of defensive mucus on the lungs which in turn restricts lung efficiency and reduces off gassing so deco becomes less effective.

Its a tough balancing act.

On OC i regularly dived 1.4 on the bottom and would gas switch at 66m 40m and 9m all on 1.6PP02. I often did hot dives where the depth was a little more than planned (change of dive site) and id end up doing the bottom phase on 1.5-1.7PPo2 and yet i never remember getting these symptoms. It seems something unique to CCR and i have always wondered why.

Safe to say my OC dives were shorter with my longest only 3hours max run and the vast majority in the 2 hour range

Bearing in mind how long it takes me to get the symptoms on CCR this may well be the deciding factor although I always wondered if it was the variable PP02 of OC that was preventing it.


ATB

 

[anarchista]

You have a low tolerance to high PP02s

I have run two hour dives on 1.4 finishing on pure 02 and i feel fine but even I find over 3.5 hours on high PP02 I start to feel crap flue like symptoms and suffer chesty symptoms post dive.

This is build up of defensive mucus on the lungs which in turn restricts lung efficiency and reduces off gassing so deco becomes less effective.

This year's conference (Polish Hyperbaric Medicine and Technology Society) was given information about the evaluation of the toxicity of oxygen to the lungs. More information will be in the next issue of PHR.
There is another mechanism that impairs gas exchange at high ppO2 a spasm of the arteries. It is the defense of the body and reduces blood circulation, blood flows a small stream and can not transport at the level of the inert gas assumed by the model.
"Basics Hyperbaric Therapy" Institute of Maritime Medicine Military Medical Academy

rc greet

 

[emmbee]

Thanks for that Ross, very interesting.

* No deep stops used in the test -at all - none! They instead tried to simulate them by working its tissue model backwards. They took their BVM model (navy use only - not deep by our standards) and reversed into a shallow profile shape that would simulate the same tissue effects of the slightly deeper profile. All this artificial fiddling was done to get the profile to finish at the same time as the base control test model (VVAL18). They instead should have let the BVM make its own profile naturally and accept the different finish times, but they didn't. So this is why the test profile ended up with an unnatural multilevel profile (and why the test went of the rails). They had a multilevel shallow profile - not a deep profile, and no deep stops to be seen anywhere.

I have to say I was a bit queasy when Andrew Fock said at Eurotek that the French Navy studies were not as good as the USN one because they let the profile run its course and so the "deep stop" one had a different, longer, runtime.

Next - and this is the kicker; the test was stopped because the base line control model (VVAL18) - was about to drop out the bottom of the test limits and invalidate the entire project. That's why it stopped - not because of the BVM model. The artificially simulated BVM profile was right down the middle of the test parameters (3 to 7%) exactly where it was calibrated to be. The test was stopped because the base line calibration data from its VVAL18 - the basis of many other navy testing projects - was about to invalidate this project by dropping below 3%. I'll say it again - this test stopped because the VVAL 18 control model datum, was 5 dives short of mandatory invalidation of the entire project!

What you're saying here is that the baseline VVAL profile was, err, "safer" than what VVAL18 model predicted, and the deeper profile was in line with it, so if they had completed the study they'd have shown that VVAL was wrong, and that would have been awkward at best. Did I get this right?

If so, that's certainly interesting but I wouldn't think it really changes the conclusion.

So, for me, the general idea remains:

Deep stops are supposed to be good because they're supposed to reduce bubble formation and that, in turn, in taken to mean less bends. That's the original Pyle stops/pearl divers predicate. VPM/RGBM/thermo models/... take the second step for granted, attempt to minimise bubble growth, and it just so happens that you get deep stops, kinda proving step one above. That's my understanding.

But looking at those studies, it turns out that deeper stops do not imply less bends, and do not even imply less bubbling. Not necessarily. That doesn't invalidate "deep stops", here we run into the difficulty of proving a negative. It does however mean that you can't just say "deep stops are good" and it does suggest that the general approach of ever decreasing the lower GF to increase safety may not be correct.

Personally, from a theoretical POV, and staying with Haldanian models, I'd be more interested in increasing the offgasing halftimes with time to try and compensate for cold and dehydration... IMHO.

Cheers,

Matthieu

 

[anarchista]

Personally, from a theoretical POV, and staying with Haldanian models, I'd be more interested in increasing the offgasing halftimes with time to try and compensate for cold and dehydration... IMHO.

Can emmbee document your views.
Jones model deals with such issues. This model in which the half-off-gassing depends on the perfused tissue, the mass and speed of blood circulation. the link already posted.
http://rebreathers.pl/forum/download.php?id=102
"Fundamentals of Pathophysiology Dives" Military Medical University

greet rc

 

[rossh]

What you're saying here is that the baseline VVAL profile was, err, "safer" than what VVAL18 model predicted, and the deeper profile was in line with it, so if they had completed the study they'd have shown that VVAL was wrong, and that would have been awkward at best. Did I get this right?

If so, that's certainly interesting but I wouldn't think it really changes the conclusion.

Matthieu

All good testing has a control baseline / datum. They ran two concurrent dive profiles. Its not one model vs another - its a test procedure control, to keep the results calbrated. The test procedure failure rates were calibrated using the VVAL18 for a 3 to 7% failure. But it didn't happen - they could not reproduce their own base line data, and it forced the cancellation of the testing..

The conclusions add up to nothing useful - there were no deep stops - they only tested a weird multi-level profile. And get this.... every model ever invented had the diver out of the water long before this navy test profile did. This test sequence failed everywhere - both baseline, and a weird multilevel shallow profile that represents nothing. Were is the justification for making any recommendations on deep stops?

**********

Added Nov29th:

This paper was presented at the UHMS 2008 Deco and Deepstop workshop. Of particular interest, is the follow on discussions between the navy's Wayne Gerth and his peers at the conference. You should buy the book (from UHMS) and read it for yourself. Many of the questions / comments from the other Dr.'s are critical of Gerths' methods or conclusions too. He did not receive any compliments on this.


************

Here it is:

Here is the a slide from a presentation. Notice how line A1 suddenly never has an incident? After 140 test more dives, that were supposed to be calibrated with reliable VVAL18 to a failure rate of 5%..... This is why the test was cancelled - the baseline was about to fall out the bottom.

For those interested looking into this more deeply, here is the original presentation at the DAN 2008 tech conference: deco.hhssoftware.com - /dan_2008/
.

 

[ebt]

But does the model reflect *exactly* what is happening to the divers body, or is it just a close enough approximation?

Measure with a vernier, chop with an axe. Thats diving.

The Deco models in use today do not try to reproduce the physiology. Too many interested people fail to understand this. A deco model tries to model the danger zone and creates limits here.

Thanks Ross, thats exactly my point. Well, that and that just because the model is capable of outputting stops to cm's and milliseconds, doesn't mean you should be following it *that* rigidly.

Its the same with PO2's, you dont need to be uber anal about their decay during ascents.

 

[emmbee]

Such easy removal are carried out on one year of physics at Warsaw University.

Forum AKP Krab AGH &bull; WyÅ***8250;wietl temat - Meandry dekompresji

You don't have a clue what formulas I'm talking about, do you? You just saw somebody saying he doesn't know something and went straight for the mocking without stopping for one second to consider that you might have misunderstood, that you might not know, or that your assumption that everybody but yourself is an clueless idiot might not be entirely correct.

No, the equations you linked to are not what I'm talking about.

This, unfortunately, is pure nonsense.Few people know but it's possible oxygen DCS.

The reason few people know it is because those who do know that:

1. The conditions for this to happen will not occur in diving
2. Even if they do, the oxygen bend will be negligible compared to the nitrogen/helium bend
3. And even if not, an oxygen bend will tend to resolve itself as the body consumes the oxygen.

The rest of your comments are about as relevant, so I won't bother to reply.

Cheers,

Matthieu

 

[anarchista]

You don't have a clue what formulas I'm talking about, do you? You just saw somebody saying he doesn't know something and went straight for the mocking without stopping for one second to consider that you might have misunderstood, that you might not know, or that your assumption that everybody but yourself is an clueless idiot might not be entirely correct.

No, the equations you linked to are not what I'm talking about.



The reason few people know it is because those who do know that:

1. The conditions for this to happen will not occur in diving
2. Even if they do, the oxygen bend will be negligible compared to the nitrogen/helium bend
3. And even if not, an oxygen bend will tend to resolve itself as the body consumes the oxygen.

The rest of your comments are about as relevant, so I won't bother to reply.

I gave tissue off-gassing of the equation, if you were in this perception, it could tell how the gas is off-gassing. To what extent depth.
The argument about a person once again shows, that they do not understand the presented equations.

Yes, oxygen is consumed, but the existence of such a form of DCS, also known in the literature, as convicted publications (wrote two by Commander Prof. Ryszard Klos). Therefore, it is introduced to reduce the speed of the ascent 1m/min.
Conditions for the occurrence of DCS is a full saturation of hemoglobin, plasma and rapid saturation of the tissues, also accumulated myoglobin oxygen.
Restriction is known and used in decompression chambers and diving in oxygen or oxygen decompression stop at 6m without the 3m.

greet rc

 

[tecdiverdude]

A good passionate discussion is what we are after.. But please remember we are adults and refrain from insults and narcissism...
This thread has caused a lot of dramas.
Please keep it civil or we will have to start using the red button- we don't like doing this .

 

[Janos]

But real life experience. How about this one: when I run SP=1.1, I feel fine but if I run SP=1.3 I feel like shit. This is for standard dives, not deeper or longer dives. I run GF 30/85 by the way.
If anyone could shed some light, I'd be happy to learn.
Best
Philippe

With 1.1 you will be decoing for longer and doing a slower ascent. You don't offgas just through gradient, but also by tiny bubbles, so the slower ascent could help with this, while still assuming the same amount of gradient diffusion.

Or it could be that the fast Helium is getting out on the 1.1 but not the 1.3

Having said that, I think this is "measuring with a vernier and chopping with an axe" (I do like that phrase)

Personally I like 30/75 with the last 15-20mins on pure O2.

Janos

 

[Janos]

PS - Good thread by the way.

 

[Hanssing]

There is a current trend in diving, and particularly CCR diving, to keep inflating ande adding more deco time onto the plan. There is often little or no justification from a decompression risk point of view. Its done purely for safety reasons, to add more safety onto more safety, and its an endless pursuit. I agree that a safe return is the most important part of diving, but this endless increasing of time has no more benefit to reducing risk.

I joked here once about 20/50 profiles, but I expect that profile will arrive and be used one day.

Regards

As you know, i have been bent a little three times on vpm-live on the x1. I call it bent, others call it skindbends, and niggles. But pain in shoulder and skin rashes do tend to make me back off a bit.

This was on maximum setting of +5 conservatism, doing two dives or more.

When i dive now, i use gf 70/35.

But my thinking after tinkering with buhlman post dive analysis, is that the on/off gassing model with no safety is to blame. The thing everyone that looks at the exponential functions realises, is that you get a "tracking" error if you offgas just a bit slower than you on gas, in regards to the arbitrary timeconstants that are used for the model. This trackingerror becomes more pronoumced over successive dives with short SIs, and gets reset when you stay out for a long enough time.

In laymans terms, if you have an empty tissue and expose it to 1 bar inerts, it will ongass at a certain described rate. Now if you have the same tissue loaded with 1 bar inerts it will ofgas at exatly the same rate. If that is not the case you will get into trouble during multiple dives......

Funnily, thats recommended by many agencies.

The ongassigband the offgassing for the compartmentloading is used in gf and in vpm....

Not to mention that the trimix model was never really tested, but just hacked by relative solubility for the halftimes.

So in my oppinion buhlman and therefor part of the vpm-foundation is very shaky, and for some divers even 35/80 is too ageessive, i know because it is for me.

Nicolai

 

[rossh]

When i dive now, i use gf 70/35.

But my thinking after tinkering with buhlman post dive analysis, is that the on/off gassing model with no safety is to blame. The thing everyone that looks at the exponential functions realises, is that you get a "tracking" error if you offgas just a bit slower than you on gas, in regards to the arbitrary timeconstants that are used for the model. This trackingerror becomes more pronoumced over successive dives with short SIs, and gets reset when you stay out for a long enough time.

In laymans terms, if you have an empty tissue and expose it to 1 bar inerts, it will ongass at a certain described rate. Now if you have the same tissue loaded with 1 bar inerts it will ofgas at exatly the same rate. If that is not the case you will get into trouble during multiple dives......

Funnily, thats recommended by many agencies.

The ongassigband the offgassing for the compartmentloading is used in gf and in vpm....

Not to mention that the trimix model was never really tested, but just hacked by relative solubility for the halftimes.

So in my oppinion buhlman and therefor part of the vpm-foundation is very shaky, and for some divers even 35/80 is too ageessive, i know because it is for me.

Nicolai

Nicolai,

Did you mean to say 35/70 ?? That is a huge amount of deco, and way beyond what's normally required for successful deco. Remember that ZHL was man tested at 100/100 !

For 99.99% of the diving world, the models of all types in use today, seem to work just fine. The Haldane / Schriener gas tracking mechanism is component of most models, and been in use for a long time.

People with skin bends are a unique breed. The real causes are still unknown. I know of one person who got them, they had PFO fixed, they add extra deco, and still gets the skin rashes. Skin bends do not seem to be well correlated to deco quantity. It seems if your susceptible to skin bends, then you get skin bends on any model and any amount of realistic deco. Hence this unique group appears to need an unrealistic amount of excess deco to resolve the skin rash problems every time. It's not the models fault.

From a modelling perspective, skin bend divers are outside the normal diver range, and models need not account for this condition. Why would they? To do so would mean doubling up on existing deco amounts - for everyone ?? Models are made for the normal healthy diver. It's up to every diver to self assess and make additions accordingly for the conditions.

The alternative is that we use your unique condition as the model datum. Then make every other diver add excess time to double up on their deco times. That's not right.


Regards rossh

 

[GLOC]

There must be enough technical dives being conducted now to gather some of this data. If we put together a survey that people could enter some demographics, a dive log with profile, results, etc we could start that process. Occurences of bends or dying are hard to model as such a rare event (thankfully), but you could model after-dive effects, whether people feel tired, sluggish, sprightly. We just need people willing to invest a little time after each dive to fill in a survey. Maybe set up a survey-monkey survey, approach people here, spread to rebreather world, YD and other locations, and see if we get enough to give meaningful answers.

The problem with this is two-fold.

1. Firstly, by collecting profile data will not be enough to prove a point. You need to know state of hydration, how well the profiles were followed, body mass/make up (lean/fat), body temp, PFO test, pre-existing medical issues, medication etc. These are all factors which may effect the decompression efficiency/effectiveness. You need to nail some of these down so that you are working from a smaller variable base. I believe that RossH with his V-planner live project provides some statistical feedback on the profiles loaded there but I don't know if he captures whether the divers got bent or not.

2. Secondly, if people aren't even will to fill in reports when something goes wrong (incident reports) why would they fill in reports when something goes right (they didn't get bent)? I know that DAN have teamed up with the majority of dive computer software manufacturers to collect profile information at the click of a button but not sure where this goes or what happens to it.

Good idea though!

Regards

 

[anarchista]

This is a good topic, as I feel many with an interest in modelling, don't appreciate the relevance of this.

The Deco models in use today do not try to reproduce the physiology. Too many interested people fail to understand this. A deco model tries to model the danger zone and creates limits here.

It is currently impossible to reproduce the human circulation system in a math model. If such a thing was invented, it would be put to important use in medicine with the diagnosis of a 1000 human diseases and ailments.


Deco models are built around parameters we can measure in the water. That is: pressure and time. Many other variables affect the required deco time, but we can't measure or use them in a dive. As a diver in the water, we can only measure pressure and time which leaves very little for a deco model to work with.

If we could measure more variables, like body temp, circulation rates, and actual on / off gassing, then we could make our deco models more capable. For that I think we would need sensors and probes on the body, communicating to a computer. Now - here is the interesting part of such a capable model; For most of the dives we do, it would suggest LESS time than we do now. That's simply because our models and calibration we use today already have a safety buffer that fits the average diver and average situation. A more accurate model calculation would give less time. Are today's divers comfortable doing less time when its appropriate? Probably not.


As pointed out, many older models use only averages of time and pressure to track exposure. There are many versions of this approach - with a varying amount of cells in series or parallel. Then they typically apply a set of constant or factors to determine ascent limits. These simple model limits are fitted values to match observations. The problem here is the simple time + pressure averaging coupled with factors / constants, will quickly gets out of step with the physiology. Hence the GF is applied to patch up ZHL. RD uses several rule sets to compensate.

You have a model of decompression Jones, perfusion model. To determine the perfusion information you need oxygen consumption. This easily obtained in CCR, pressure drop in the oxygen cylinder is the main source, indirect information is pulse.
The effect of exercise on saturation and desaturation, which gives a better assessment of a specific optimal decompression diving.
The effect of the application of a variable half-off-gassing time, but it is not a computational problem.

rc greet

 

[rossh]

Now, if you're looking for studies that suggests that deep stop may not be good, which wasn't what I said, but anyway, here you go:

• Bubble incidence after staged decompression from 50 or 60 msw: effect of adding deep stops.

http://www.ncbi.nlm.nih.gov/pubmed/15892549

.
Lets now take a look at the French navy tests and paper. This test sequence used 6 different test profiles (EAP 1-6), some of which are realistic and typical of tech diving (except for the deep air part).

There were 12 navy divers used in the test sequence, but 3 of these had previous DCS incidents in their careers.

For tests 1 to 4, the first stop was at ½ max depth ATA. For test 5 and 6 (trimix), first stop was at 1/3 max depth ATA. The first stops used were not as deep as tech diving standards, and the trimix stops were a long way short. The model used for the EAP profiles is described as being based on an extended Haldane style, but is not specified exactly.

Only two test profiles raised concerns in these tests:

*****

1/ EAP 2: profile 60m, 20 mins, AIR, with 100% O2 deco.

m min gas
60 20 air
27 1 air
24 1 air
21 1 air
18 2 air
15 2 air
12 4 air
9 6 air
6 5 100% O2
3 10 100% O2

One diver reported joint pain DCS. This diver was given an MRI and found to have a pre-existing condition: a bone infarction of the humeral diaphysis (dead bone tissue in the center of the upper arm).

Comment:

This diver has a previous dive injury that should have excluded him from testing, and possibly diving in general. This DCS event is not related to the test sequence, but instead this diver was predisposed to a DCS occurrence.

The profile above is rather ordinary, and has been carried out thousands of times by divers of all types. Today's deco planning tools will add typically about 5 to 10 minutes to that ascent time.


*****

2/ EAP3 is two air dive profiles with just 10 min surface interval.
50m, 15min, air, 10mins SI, 50m, 15min, air.

m min gas
50 15 air
15 1 air
12 1 air
9 2 air
6 3 air
3 7 air
Surface 10 mins
50 15 air
18 1 air
15 1 air
12 1 air
9 2 air
6 5 air
3 13 air

The concern raised in the tests was that they observed an increased Spencer count when compared to their base MN78 profile.

Comment:

The significance of Bubble counts via ultrasound / Spenser/ KISS seems to be poorly understood, and some research reports will make or imply a basic miss-understanding in their comparisons. This is explained below:

a/ Two identical dive profiles, with different ascent / deco shape will have almost EXACTLY the same volume of gas to be exhaled from the body. All too often these research papers include or imply invalid assumptions that "profile x contained more gas or bubbling, than profile y". That is not possible.

The deco shape has little effect on the total excess gas volume to be removed from the body. That is because the excess gas the diver absorbs occurs during the bottom section of the dive, and the bottom segments are identical in these test profiles. It is the bottom section that has the highest on-gas rate (greatest sub-saturation and tissue pressure imbalance exists). The bottom segment is where most of the inert gas is absorbed from. As the diver ascends the on gassing will cease, and the off gassing commences. This point is located before any deco stop takes place in any deco profile. Therefore both profiles started with the same volume of gas to discharge from the tissues. Any suggestion that profile x made more gas than y, is misleading.

b/ All supersaturation (decompression) makes microbubbles. You can see an everyday example of this when we open a bottle of coke or beer. Normal off gassing removes inert gas from the tissues in two states: as a saturated dissolved state, and as a gaseous microbubble state. This gas is transported in the blood and will pass through the heart and onto the lung, where it is trapped and exhaled. If we place an ultrasound listening device over the heart / lung, you will hear the microbubbles passing. This is NORMAL decompression in progress.

More microbubbles does NOT mean higher risk! Research in this area has failed to find correlation between microbubble volumes and DCS risk! Despite the obvious desire to link DCS risk to Spencer count, that simple relationship is not there. <changed Dec 3>

More microbubbles merely indicate more off gassing in progress. Off gassing is the objective of decompression. The more off gassing that is achieved - the less gas remains in the tissues to do harm.

Consider what a DCS event might look like? It would likely retain gas at the injury site (the tissue), leaving less gas to be passed onto the heart / lung. That would mean LESS gas to be heard in the heart / lung area, and a lower Spencer scale! Higher Spencer grades mean good normal deco in progress.


c/ A longer deco time means the SAME amount of off gassing will be spread out over a longer time frame. It does NOT imply lower bubbling or less gas from the profile. If we compare a Spencer scale on two different dives, then it will be lower for a longer deco - simply because the SAME amount of microbubbles and dissolved gas are spread out more thinly, over a longer time period.


d/ Microbubbles - size matters! The Spencer scale has no method of assessing bubble size, and therefore total gas volume. The Spencer / KISS grades are a measure of noise levels, which implies a bubble count. The problem here is the geometry of a sphere (bubble). A bubble's volume will increase 8x for a 2x increase in its diameter. For any one sample, if the bubbles were doubled in size, there would be a 75% reduction in the bubble count. That is most significant when our only measure is noise created from bubble counting.

The Spencer scale has no ability to measure actual or relative volume of gas passing the sensor. In order to truly compare off gas rates, we need to include the volume of gas passing into the lung, and we cannot do this yet.

This discrepancy of size may go a long way to explaining why no two divers give the same Spencer result. There are problems in getting a consistent value from Spencer scale readings across the population. If the average bubble size is influenced by each individuals physiology, then trying to measure the gas volume by bubble count, will not give a consistent result.

--

Finally, I'd like to point out that there are table sets in use today, that have been tested for consistent Spencer grade and microbubble count. The are the DCIEM tables. These did not try to eliminate microbubbles, but instead they keep bubble scores consistent. Ron Nishi wrote once, that helium based dives have about double the bubble score over nitrogen based dives (probably the 2.65 ratio N2:He).
The DCIEM tables are used in the Canadian military, and are mandatory for commercial work in Canada. These tables align well with bubble models and other modest deep stops systems.

Rossh