NEDU Study

[Kevrumbo]

. . .Also important, I hold in VERY high regard the person that can show a change of opinion when presented with new evidence. Dr Mitchell has changed his preferences from deep stop oriented diving to shallow stop oriented diving in response to being presented with new evidence.

Hell . . .I found out the hard way as a patient of Dr. Mitchell in 2013, and was first made aware of this "retro practice" as early as 2009, with the Truk Stop Dive Guides (and Dive Ops Manager @Kelvin Davidson at the time) taking "peculiarly" extra long O2 shallow stop profiles each dive day for four consecutive days before taking their day-off break.

The issue at play here seems to be that if you are using a bubble model and you know that your slow tissues are being subjected to this effect that the best you can do in the absence of proper calibration of the model is guess about how to fix it..... or use a different model.

Intuitively you must be right but padding your shallow stops based upon gut feeling shouldn't be the way we decide if we've done enough deco. The models should be calibrated correctly to begin with, if you ask me.

R..

Well that's where are right now . . .an individual trial & error "guesstimate" on how much to deemphasize deep stops along with the longer O2 shallow stops.

So far for a gas content model like Buhlmann w/ GF's Lo/Hi, I'm reading OC & CCR Divers posts of using 40/70, 50/80 and a trend just in this thread of suggestions of 60/xx and 80/80 (post 76 above for Air) toward "99" GFLo and a surfacing range selection of from 70 to 80 GFHi.

 

[rossh]

Simply put Ross, indeed the cheapest insurance is an extended shallow O2 profile stop, especially on top of bubble model schedules for those still choosing to use them, which the NEDU Study has shown -put unnecessary inert gas supersaturation decompression stress on slow & intermediate tissues upon surfacing.

Kevin suggests that there is a meaningful difference in the amount of on gassing in slow tissue, caused by initial slow ascents / deeper stops. Lets check that idea in the context of tech model profiles using real science measures.......

Has anyone really done the homework on this Kevin - used proper deco math to study this? lets do that now.

As we can see.... All ascents on gas the slow tissues, in all models.

In the 170 ft 30 min example, the slowest limiting cell (#8) was on gassing up to 50ft, and then swaps over to off gassing at the 40 ft stop - in ALL profiles. Cell #12 was on gassing for all the dive, until the last stop - all profiles. Cell #16 on gassed for the entire dive. It's the same conditions for all profiles.


Let me say it again.. Using real science measures, the differences caused by deeper stops on shallow time, are so tiny, they are barely visible. The large effect from deeper stops idea that is being promoted, is a fallacy.


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

The issue at play here seems to be that if you are using a bubble model and you know that your slow tissues are being subjected to this effect that the best you can do in the absence of proper calibration of the model is guess about how to fix it..... or use a different model.

Intuitively you must be right but padding your shallow stops based upon gut feeling shouldn't be the way we decide if we've done enough deco. The models should be calibrated correctly to begin with, if you ask me.
R..

Well that's where are right now . . .an individual trial & error "guesstimate" on how much to deemphasize deep stops along with the longer O2 shallow stops.

I sense you both might be trying to imply the thought, that the existing gas kinetics formula, used universally throughout deco modelling and dive computers, is some how not up to the job?? ie. is the suggestion that formula work OK on the shallow stop parts, but not at the deeper stop parts ??? Is this why you feel the need to manually add an abundance of extra time that is beyond the amount derived by underlying gas kinetic formula?

Let me just kill that thought right here....

The deepest stop of all, is the bottom segment - 30 mins for example. We all on gas at the bottom. We fully trust the gas kinetic formula for the bottom segment and expect it holds the correct information for the ascent. But why would anyone think the gas kinetics would go wrong for a few brief minutes in the deeper stop part of ascent of one particular profile type??? Makes no sense. Obviously if the formula work on the deepest part (the bottom), and work in shallow stops, they will work equally well in the stops in-between



This practice of ad-hoc adding lots of shallow time, is grossly out of proportion with the real needs of the model and on/off gassing formula. People are welcome to add all the time they wish, but it's not justified by gas kinetic conditions.

.

 

[Patoux01]

I sense you both might be trying to imply the thought, that the existing gas kinetics formula, used universally throughout deco modelling and dive computers, is some how not up to the job?? ie. is the suggestion that formula work OK on the shallow stop parts, but not at the deeper stop parts ??? Is this why you feel the need to manually add an abundance of extra time that is beyond the amount derived by underlying gas kinetic formula?

Let me just kill that thought right here....

No... I'd say that Diver0001 was pointing to the fact that VPM has no real parameters in there, they're all "oh those values looked nice when I typed them". I'm not gonna look up the newsgroup where it was posted, you know damn well which post I'm talking about.

I might be wrong though, but that's how I read his "absence of calibration".

 

[Diver0001]

I sense you both might be trying to imply the thought, that the existing gas kinetics formula, used universally throughout deco modelling and dive computers, is some how not up to the job?? ie. is the suggestion that formula work OK on the shallow stop parts, but not at the deeper stop parts ??? Is this why you feel the need to manually add an abundance of extra time that is beyond the amount derived by underlying gas kinetic formula?

...snip....

.

This is a straw man and such a transparent one that it isn't worth responding to so I'm not going to.

For the benefit of people wondering what I meant by "calibration" what I mean is that it has become abundantly clear that BUBBLE models like VPM, RGBM and Thalmann are not suitable for use in technical diving because they do not calculate stops correctly.

Calibration means fixing it so it calculates the necessary shallow stops properly.

Buhlmann has been calibrated multiple times over the years by using different sets of M-values and more recently by adding gradient factors. Something similar must surely be possible for the common bubble models as well because as it is, the shallow segment of the dive gets calculated to be too short and too deep.

... and ... until such time that the algorithms are calibrated we should simply not use them. This is the take home message one gets from the NEDU study and it's the elephant in the room that has Ross panicking.

R..

 

[rossh]

This is a straw man and such a transparent one that it isn't worth responding to so I'm not going to.

For the benefit of people wondering what I meant by "calibration" what I mean is that it has become abundantly clear that BUBBLE models like VPM, RGBM and Thalmann are not suitable for use in technical diving because they do not calculate stops correctly.

Calibration means fixing it so it calculates the necessary shallow stops properly.

Buhlmann has been calibrated multiple times over the years by using different sets of M-values and more recently by adding gradient factors. Something similar must surely be possible for the common bubble models as well because as it is, the shallow segment of the dive gets calculated to be too short and too deep.

... and ... until such time that the algorithms are calibrated we should simply not use them. This is the take home message one gets from the NEDU study and it's the elephant in the room that has Ross panicking.

R..

Oh... that is just so wrong.

Real ZHL (Buhlmann) is calibrated to man testing.... VPM-B has none, but follows the existing gas kinetic formula like other models, and was aligned with various points in other model and table data at the time.

GF is not a "calibration".... GF is an ad-hoc patch, applied to the end. GF is the opposite of calibration... GF is not used to make equally proportional deco any more. The way GF is abused now, is to deliberately bias and set it to create an over sized shallow stop portions on the end.....

GF is being used as a "dial a plan" system - no one cares what the underlying model design concept is anymore. Today they add so much false extra shallow time onto the plan, that its lost all connection the underlying model attributes. These days, one could use any model as the basis and apply large amounts of GF. They are interested in a desired fitted result - not a deco formula.


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

The nedu test measured two kinds of shallow stops. ZHL and VPM-B both follow the attributes of the successful A1 profile. Nothing in tech matches the failed A2 profile.

The underlying message from the nedu test, is that the generic gas kinetic formula work, and ad-hoc methods do not.

ZHL,VPM-B follow the gas kenetic formula... the nedu test confirms the proper and correct or implied orientation of deco limits of existing ZHL and VPM-B.

Add-hoc methods, like RD, like GF like DIR classroom theory are the problems.... quit trying to blame VPM-B for other peoples problems.






.

 

[Diver0001]

Oh... that is just so wrong.

You see, this is where you're off the rails, Ross.

The NEDU study tells us not to trust bubble models.

And we don't.

When you drill all the way down to the core of this (and similar) discussions with you. This is is. Nothing you can say is going to make people start trusting VPM again. It needs fixing, not a bunch of "baffle them with bullsht" discussions on the internet.

So what is necessary now is for the people who understand those algorithms to phone Simon or David, work with them and fix it. Until that happens we should not use those algorithms for technical diving.

There is it. That's the core issue laid out in black and white.

R..

 

[rossh]

You see, this is where you're off the rails, Ross.

The NEDU study tells us not to trust bubble models.

R..

No.... the nedu showed us to ignore the BVM(3) TDGM model.... fortunately nothing like it exists in the real tech world.


Too many people try to play the "guilt by word association" game..... playing on the word "bubble" or "deep".


VPM-B is not the BVM(3). They are very different, so please don't get it confused. Mathematically, VPM-B is much closer to successful A1 VVAL profile, than the failed BVM A2 profile.


That implied connection by words alone, has been the basis of the anti-VPM attack all along.... Real science math, real science formula show there is no connection. I have been pointing this out with facts and diagrams, here and elsewhere.

There is no real science to connect the nedu test to VPM-B, or tech diving profiles. The nedu test paper makes no claims to represent anything beyond what it tested.

The attempted explanations and positions to make any tech connection so far, by all involved, are opinion based theory only. But when put to analysis, they have failed.


.

 

[Diver0001]

You are now trying to play the "guilt by word association" game..... playing the word bubble or deep.
.

This is another straw man but I will respond to this for the benefit of people who haven't read previous threads.

Simon demonstrated quite conclusively that VPM suffers from EXACTLY the same problem that BVM(3) has.

R..

 

[Diver0001]

Well,

If we are going to play the credibility card again then who are we supposed to believe?

1) A highly respected and thoroughly professional decompression scientist who is both an MD and a PhD, a hyperbaric specialist and a technical diver.
or
2) A computer programmer who sees his baby being threatened and is on the internet spinning up one straw man after another about the work of #1.

I won't presume to dictate who people should believe but I have made my choice.

R..

 

[rossh]

Well,

If we are going to play the credibility card again then who are we supposed to believe?

1) A highly respected and thoroughly professional decompression scientist who is both an MD and a PhD, a hyperbaric specialist and a technical diver.
or
2) A computer programmer who sees his baby being threatened and is on the internet spinning up one straw man after another about the work of #1.

I won't presume to dictate who people should believe but I have made my choice.

R..

Same choice as always.....

a/ the interpretations and opinion positions of a trained Dr., who also publicly promotes his personal agenda for change,

or

b/ the math and trusted deco science formula that shows the reasoning of a/ above is not correct. Its also carried out by an amateur scientist, who coincidentally builds some of the best deco planning tools.

....... reputation vs facts.

.