RGBM vs Buhlmann

[College Diver]

I recently spoke to someone who claimed to know a mathematician who derived both of the algorithms just for kicks. The end result being that some "inconsistencies" were found which appeared to show that the Buhlmann was a bit more lenient in situations which could ultimately result in a user of this algorithm getting hit.

The final opinion was the RGBM was a much safer algorithm to use.

I assumed the two were the same thing.

I have a few questions about this:

Has anyone on the board taken the time to derive both algorithms, and if so, were similar inconsistencies found? If it is in fact the case that these inconsistencies exist, does the inferred conclusion follow from the data?

 

[stas]

What you are asking doesnt really make sense. These algorithms are numerical implementations of the RGBM and Buhlmann models of decompression. What might make a little more sense is to derive the equations of the models themselves starting from the basic assumptions of each model (which are different by the way). This while an interesting and not very difficult exercise of working with ODE's, does not answer which model is safer in terms of incidence of clinical DCS of its users.

One can try to compare run times produced by each model for a given dive profile, but that wouldn't be an indicator of their relative safety. These are models with fundamentally different approaches to safe off gassing.

Empirical data is the only thing that can be used to evaluate their relative safety factors. At the end of the day one remembers that real decompression is an incredibly complicated physiological process which we know depends not only on the person but varies from dive to dive for a given person. Then you pick the model which sounds better to you (or to your dive computer manufacturer) and go with it.

Having said this, I will add that it is now becoming clear that bubble nucleation is is a factor in decompression and their growth needs to be controlled. Thus a modern decompression model should try to incorporate controlling these bubbles into itself.

 

[College Diver]

it is now becoming clear that bubble nucleation is is a factor in decompression and their growth needs to be controlled. Thus a modern decompression model should try to incorporate controlling these bubbles into itself.

IS what you are referring to here possibly the same thing as the Micro bubble level associated with the Uwatec computers?

 

[stas]

Yes, Uwatec implements the Buhlmann ZH-L8 ADT model which tracks microbubbles. Thats what the letters ADT mean (streight Buhlmann does not track bubbles)

 

[amascuba]

RGBM uses a bubble model to calculate decompression and incorporates a slower ascent with deeper stops.

Buhlmann uses a dissolved gas model to calculate decompression and incorporates a faster ascent followed by longer stops at shallower depths.

Bubble models assume that the dissolved inert gases in your body will bubble at deeper depths and account for that by slowing down the ascent rate by utilizing short deep stops followed by a shorter shallow stops. This allows the body to more effectively off-gas by allowing your blood circulation carry the dissolved gases through your bodies natural filter, your heart and lungs, before the bubbles get too large. By doing this you are letting your fast tissue compartments to off-gas more efficiently during the deep stops followed by your slow tissue compartments to finish off-gassing in your shallow stops.

The dissolved gas model doesn't take into account the bubble formations at deeper depths. It assumes that you will do the majority of your off-gassing at the shallower stops by forcing bubble growth during the ascent followed by a longer hang at shollower depths. You might hear some people refer to this as "bend and mend".

 

[Dr Deco]

Hello College Diver:

Haldane-type Models

All decompression models, starting with that of John Scot Haldane, have purported to be based on a theoretical model. John Scot Haldane’s was established on an idea of limited, but stable, supersaturation. This notion persisted into the 1990s. While incorrect, it has the clear ability to generate tables for decompression, and it allows extrapolation to deeper depths and longer bottom times. This idea, for example, was the basis for the US Navy dive tables.

Theoretical Objection – Nuclei

The major problem with the stable supersaturation concept was that a free gas phase could [supposedly] form with very limited dissolved nitrogen supersaturations. This is not in agreement with theory or experiment. Typically, several tens – if not hundreds – of atmospheres of supersaturation are needed to form gas bubbles in water. Water is a liquid with a very high cohesive force between its molecules, and voids do not form within it with ease. The dissolved nitrogen oversaturations in diving are slight and not capable of [de novo] bubble formation.

The solution to this dilemma was to introduce into diving the concept of preformed tissue micronuclei. These preformed nuclei were accepted in virtually ever endeavor [even baking] with the exception of diving and, their acknowledgment was a late arrival on the scene.

Theoretical Objection - Surface Tension

Surface tension is the force exerted by molecules of a liquid at the gas-liquid interface, that is, the surface of the bubble. The smaller the bubble, the greater is this tensive force, and hundreds of atmospheres of dissolved nitrogen pressure are need to form bubbles when they are very small [less than a fraction of a micron].

Nuclei

Liquids, however, always have microbubbles or voids in them; this is the result of thermal motion of the molecules. There is a distribution of sizes, many very small ones, and a few large ones. Very large microbubbles are very scarce until one gets to elevated temperatures. Near boiling, e.g., one observes [vapor] bubbles with ease.

Larger microbubbles can be formed – that is , enlarged - from nuclei by a reduction in pressure of the liquid surrounding them. This is referred to as “hydrodynamic cavitation” and, it occurs when liquids move (Raleigh cavitation) or surfaces separate (Stephan adhesion). Musculoskeletal movements [physical activity] also generate these reductions (“stress assisted cavitation”). :sprite10:

I suspect that the number and size or tissue micronuclei differ from individual to individual. This is conjecture and is unproven to date.

Application of the Two Models

The Haldane type models [US Navy, PADI, and Buhlmann] are concerned only with dissolved nitrogen. They emphasize controlled but rapid ascents to near the surface in order to increase the gradient between dissolved nitrogen in the tissues and the lungs. The Microbubble Models, conversely, have slower ascents and pause at deeper depths. The deeper stop causes the microbubbles [always present, remember] to remain small; the surface tension [Laplace pressure] of the small bubbles will be large and continue to cause a shrinkage of the bubbles. Taking advantage of the Laplace pressure is unique to microbubble [two-phase] models.

How are they Made?

All decompression procedures attempt to reproduce the real world. Whatever may be the underlying theory, the deco table will match real diving experience. This is accomplished by collecting dive data and analyzing it to adjust parameters of the model. In particular, dives with DCS are required to define the limits. “Clean” dives are only minimally helpful. Especially for two-phase models, computers are required to obtain the best fit and adjustment of the model’s mathematical parameters. This can be a very extensive process and requires a large [e.g., a Cray] computer.

It seems that the adjustments for the deep stops cause the model parameters to forbid a series of relatively easy, shallow dives. Three dives a day allowed by the USN, DCIEM or PADI tables are sometimes not allowed by the RGBM.

This is an attempt to give a short but concise explanation.

Dr Deco :doctor:

 

[College Diver]

Thank you very much for your time - while I do not completely understand everything you have written here in your explanation, I am in the process of studying the physics involved with diving. thanks again.
Alan Crane

 

[NiclasG]

The 2 major brands(Uwatec & Suunto) track microbubbles. About the other
ones I do not know. Suunto uses RGM folded over a Haldanian model.

Uwatec has it's own stuff on top of Buhlmann. Bühlmann did the work for Uwatec when he was old and ready to retire. In the beginning he wasn't particulary interested but then he
caught fire and got very interested and then he died. Since then some 2-3 other
dive physiologists have worked/validated parts of the Uwatec model.

Buhlmann's book on dive physiology is a bit outdated but still a good place
to start for somebody interested in diving. The major impact of the
book is that it presented things in a way that was easely understandable
and implementable in source code.

No model is static but gets extended and improved as time passes,
at least Uwatec used to be an active company in the research community. This is important
for funding new research and advances in diving safety. Diving is small money for manufacturers (tiny volumes) and as business diving hasn't grown for years.

Niclas