Back-calculating M0-values of the RDP

[dmaziuk]

One actually might then determine M0 by taking a plus b times the ambient pressure (absolute) at sealevel, but this then is unrelated to any dive profile or no-stop relation...

Right, that is in one of Erik Baker's papers too, and yes: Workman's M0 and dM are not related to specific profiles. I played with it a little sometime ago, one of the notebooks in GitHub - dmaziuk/diy-zhl: DIY ZHL: Buhlmann diving decompression model in python has the M-values converted from ZH-L16. There's also numbers from various sources (in the python file), take a look if interested.

There's nothing there about deriving M-values though.

 

[Jacob BV]

If you know the outcome you are looking for and you know the input model which has parameters, instead of trying to do this "back calculation" just curve fit the input to the output. I don't know what coding languages you are comfortable with but for example in Matlab the function fmincon can solve things like this. Now with the input parameters you can just back calculate the back calculation

 

[Arie0510]

The question more or less is: "what is the input model", so which function to feed to fmincon.
Though the authors in [1] mentioned some of the ingredients, they did not mention all ingredients...

 

[leadduck]

I think he meant only that he had translated the title.

The English title is "Decompression - Decompression sickness"

 

[loosenit2]

The English title is "Decompression - Decompression sickness"

Albert Bühlmann - Decompression — Decompression Sickness - [PDF Document]

Albert Bühlmann - Decompression — Decompression Sickness

documents.pub

 

[Jacob BV]

The question more or less is: "what is the input model", so which function to feed to fmincon.
Though the authors in [1] mentioned some of the ingredients, they did not mention all ingredients...

The underlying model is just a diffusion model of independent half times. So for each depth intergrate these through time, you can even feed into fmincon that an input is the ascent rate. Any assumption you do not know they have made make it a variable to be optimised by fmincon.

 

[Arie0510]

I might have an explanation for the deep asymptote of 262 fsw, though it is a bit artificial.
For depths until 100 fsw the no-stop times have been floored to a multiple of 5 minutes.

Solving the no-stop relation

D - 20.15 = 803 * t^(-0.7476)

yields t=5.01 minutes for D=261 fsw, and t=4.98 minutes for D=262 fsw. Flooring that to a multiple of 5 minutes would give that the first no-stop time of 0 is for D=262 fsw...

Still, this is a bit artificial, since for depths deeper than 100 fsw the no-stop times have not been floored to a multiple of 5 minutes...

 

[Arie0510]

I tried to reconstruct the Square and Blue book M0-values in Tabel III of [1], which are based on a square profile, but there I also had difficulties retrieving the ones for smaller half times.

Unfortunately the wayback machine internet archive has not succesfully stored "the Blue Book" (Powell, Spencer, Rogers: Doppler ultrasound monitoring of the gas phase formation following decompression in repetitive dives, 1988) from the Rubicon foundation archive (pdf is corrupt). I hoped to find more details about the way they back-calculated the M0-values in that reference.

Does anyone happen to have downloaded that pdf from the Rubicon archive (4429; DSAT_1988.pdf) before the site went down?

 

[dmaziuk]

@tbone1004 have you heard from Rubicon folks recently?

 

[Arie0510]

I just stumbled into this post on this forum, een extensive excel spreadsheet by @EFX .
I will start playing around with that excel file to see if that can provide me with some more insights...