Hi Giovanni:

This is long, few discuss it, and Giovanni opens a host of issues that few
have clarified.

Formost, I agree with you sentiments completely.  I'm happy you were
motivated to dissect RGBM, its not easy but at least Wienke makes his
thinking known, albeit reading the various versions of RGBM reminds me of
cereal standing in milk for some time: whether its corn or bran, it can be
tasty if one is hungry enough but it is fundamentally ?? :)

The opposite, the days of proprietary algorithms is, in IMHO, much
worse.  Few work in decompression with the passion of the few represented
in ftp.decompression.org.  Few articles on most decompression theories
exists for widespread dissection as Wienke has made available, which you
have analyzed and hopefully others will, in a kind of 'peer' review.

Consider a review of his recent book, which also is a glimpse into Wienke
himself in general:

"
Comparing books again, subtle variations in some equation hint that other
complex equations could be simplified. For example, Wienke chose to drop
fC02 and fH20 in equation 10.7, p.202 compared to TDD p.76; in p.72
equation 4.15 Wienke drops water's specific density compared to TDD p.122.
The resulting changes are, to knowledgeable readers, insignificant, but in
reviewing other equations, one wonders what further simplification may be
possible. A transcendent notion one gets for models is that precisions of
formulas are inconsequential if they do not translate in physical effects
in real dives. Such a notion is key to remaining skeptical about models,
and their subsequent complex equations. Its ideal for writers of
decompression models to speak from the point of diving such decompression
models themselves, such as Wienke does, to 'fill the gaps' between what is
surmised and what occurs."

You can read the whole review here:

http://www.amazon.com/exec/obidos/tg/detail/-/0941332977/qid=1031923154/sr=8-1/ref=sr_8_1/102-4995323-8596101?v=glance&s=books&n=507846


At 10:22 AM 9/13/2002 +0200, Giovanni Marola wrote:
> > ...  Suuntos corrective RGBM algorithms are
> > documented by Wienke in widely available publications and aggressive
>
>Yes there are many papers by Wienke, most of them are available
>on the net but IMO, there is also a lot of confusion in these papers.

The 'free' internet papers are very informative, but Wienke makes subtle
changes in several versions and considers a printed version more
'final.'  He cuts and pastes what he needs from past articles so the new
articles look identical, except in key areas.  It can be a pain to read, as
they are 'works in progress' and he is liberal about posting drafts.  Most
of it is good, but there are errors, particularly when something is
evolving in his mind such as the approach of using an equation of state to
modify the bubble's critical radii.

>Let me do an example.
>
>In the paper titled  "Abyss-Reduced Gradient Bubble Model"
>by Bruce Wienke
>(http://www.abysmal.com/pages/articles.html)
>I have found at page 9:
>
>Reduced gradient bubble model.
>Crucial to all bubble models are the concepts of critical radii....
>....
>1/r = 1/r0 + DP/k                                                        (47)
>....
>Table 3: Micronuclei excitation radii.
>13       0.89        153      0.49
>.......


The above critical radius formula is VPM, which he 'adapted' in most
versions of RGBM.
In the last year or so, he has decided to look at equation of states, which
I personally believe is problematic, but its resulting REAL WORLD tables
for diving remains to be seen.

>whereas in the paper titled  "Reduced Gradient Bubble Model"
>by Bruce Wienke and T. O'Leary
>(http://www.scuba-doc.com/rgbm.pdf )
>I have found at page 12:
>
>Reduced gradient bubble model.
>Crucial to all bubble models are the concepts of critical radii....
>....
>r = alfa + beta(T/P)**1/3 + k(T/P)**2/3                     (61)
>....
>Table 5: Micronuclei excitation radii.
>13       2.10        153      1.18
>.......
>
>As you can see the two papers are basically equal, but
>the critical radius has two completely different formulas
>and a set of completely different numerical values in the table.
>How is this possible????

Its worse, Giovanni, the table has been revised in the rgbm71602.pdf draft
of Wienke's
http://groups.yahoo.com/group/divingaccidents/

pressure excitation radius pressure excitation radius
P (fsw) r (um) P (fsw) r (um)
13 0.174 153 0.033


>In addition there is an other very, very strange thing. The formulas
>which follow in both papers are equal and are exactly those
>one can find in the VPM pionered by D. Yount.

Serious workers see the similarities of VPM in the RGBM algorithms such
that when VPM states a null, RGBM restates as:
[[123^4567+890]+kappa/.006*nRT]x 0, and kappa is no where to be found :)

However, as of this writing, no VPM computers exist, and while the morality
of profit over a public domain algorithm can be debated, there are unique
ideas of RGBM implemented in the new Suunto computers albeit NOT the
critical diffusion and phase algorithms that make bubble models unique and
thus a critical reason NOT TO USE ANY DIVE COMPUTER for routine
decompression unless the decompression algorithm is well described [ such
as Uwatecs Buhlmanns, albeit transforming slowly into something proprietary.]

When one is faced with rigorous math Wienke uses to describe physical
events then find fault in RGBM somewhere, if not its apriori, one may
equally be skeptical about all equations that comprise the full RGBM suite.
However, RGBM has ideas imbedded in its key equations to address 'real
world' research on diving such as increase DCI risk with multi day diving
... and its seems to be useful to use the concept, rather than the actually
RGBM formulae, to achieve this.  Whether its a simple counter than reduces
NSL by Y given Z atop classic Haldanian NSL, it appears Suunto was guided
by the net effect of the radii adjustments to VPM and its dive resultant
tables, and then extrapolated to Haldane by Suunto. Note I say VPM, because
until roughly a year ago, the bubble mechanics in RGBM was based on
VPM.  Regardless, adjusting radii is an RGBM idea.

With that said, Suunto's RGBM is not the same as the full blown RGBM, nor
is the Abyss implementation.  I do not know if ANYONE has the full RGBM
working as envisioned by Wienke.  HOWEVER, we do know that VPlanner
embodies much of VPM, BUT with some modifications by Erik Baker, who now
champions VPM.  Vplanner is easier to use, compared to the Mathematica
scripts and Basic code released in the late 1990s.  So look to Vplanner to
evaluate bubble model generated tables.

>However this contradicts a recent statement by Wienke
>(http://scuba-doc.com/rgbm.html)
>
>Here is Bruce Wienke's response to our answers:
>      Just read your Newsletter Mailbox answer to question about RGBM vs VPM.
>There is quite a bit more all should know (major differences) -- actually
>the RGBM
>abandons the gel physics of the VPM as NOT applicable in toto to blood and
>tissue.
>With all due respects to my friend and decreased colleague, David Yount, I
>must go
>on record as NOT accepting that VPM gel dynamics apply routinely to the body,
>nor the properties he studied.  Such VPM type bubble seeds have NEVER been
>  found in the body -- nor outside of "gel-like" media. RGBM (EOS) bubbles do
>recover VPM bubbles in limiting circumstance of material strength and
>pressure,
>but that is not important to the RGBM. Naturally occuring bubble "seeds"
>in the
>atmosphere and oceans are NOT akin to VPM gel bubbles -- NOR should they be.
>The body, oceans, and atmosphere are NOT gel.
>.....

This is a change of heart about 2 years ago.  People speculate that as RGBM
portions may increasingly appear under license in dive computers, the
difference between RGBM and VPM must be made more apparent as VPM is public
domain.

>_________________
>
>So I have a question: If the body is not a gel why RGBM use the results
>which have been obtained considering the body as a gel?

I don't know, and that's fundamentally why I am very interested in seeing
the new RGBM tables based on equations of state, and if they differ
substantially from VPM tables.

A brief discussion on decompression.org stated simply, that bubble seeds
inside human bodies hover at 96-99F which hardly effect PV=nrT, make EOS
RGBM look more like VPM!

1/r = 1/r0 + DP/k

[VPM]

r = alfa + beta(T/P)**1/3 + k(T/P)**2/3

{RGBM}



>As you can see there are some very strange and contradicting
>things about RGBM.
>
>Giovanni

Without a doubt!  Got milk?
Warm regards,
Marv Gozum
Philadelphia PA