Mailing List Archive

Mailing List: techdiver

Message Display

> >Just a nit-pick; MiG Plan generates tables based on the Huggins and
> >Buhlmann algorithms.  It is not accurate to say this person was bent on
> >MiG Plan; he was bent on Buhlmann tables.  Some of the other programs have
> >made changes to the models, and in these cases it may be somewhat accurate
> >to say that a person was bent on them.  MiG Plan develops tables 
> >faithful to the original design of the models, and therefore deserves 
> >neither credit nor blame for the reliability of its output.  
> 
> I have to take exception with this statement.  If you run MigPlan with
> "Normal" bias, then you are running the Buhlmann algorithm.  However,
> once the diver/user chooses any of the "more conservative" bias settings,
> then they are running something other than pure Buhlmann, they are running
> a MigPlan modified Buhlmann.  Changing Dr. Buhlmann's "a" or "b" parameters
> does constitute modifying the algorithm.  And, once the algorithm is
> modified, relying on the testing done on the "pure" Buhlmann tables is no
> longer valid.
> 
> Maybe we are just picking on semantics, but I do believe that by allowing
> the user/diver to choose different conservative factors, that you have
> departed from the "pure" Buhlmann algorithm.   Probably the only approach
> to adding conservativism to the Buhlmann algorithm that essentially does
> not modify the Buhlmann algorithm is using the next greater depth or
> the next greater bottom time when planning the dive.
> 
> This applies to the other dive/decompression software out there that
> makes the statement that the testing done by Dr. Buhlmann is all that
> is needed to validate their modifications to the algorithm, but especially
> to those that modify Dr. Buhlmann's parameters.

I personally think you are meddling much more with the model by adding 
bottom time or depth than by changing the a's.  

The safety of the model is based on two things; how accurately the 
compartments reflect what is going on in the body, and how close to the 
"safe" supersaturation limit you are willing to go in each compartment.

Nearly everyone agrees that it is not prudent to modify the compartments
without testing.  Therefore, nearly all the methods that try to make the
tables more conservative try to back off the supersaturation limit to
increase safety.  This is a good method because it both directly increases
safety in a statistical sense and helps compensate for small errors in the
design of the model itself. 

Since this is the goal, the best way to achieve it is through a direct
method; pulling back the factor a's.  This perfectly acheives the goal. 
If you try to add bottom time or depth or %gas to the equation before
putting it into the model, the results are inconsistant.  You may get an
increase in conservatism in the slow compartments but not the fast ones,
for example.  This approach can also be hazerdous on multi-level dives. 

Say, for example, you are adding bottom time.  If you add bottom time to 
the shallow end of your dive, the model might think you outgassed 
from some compartments more than you actually did, resulting in a 
violation of the model.

   Jody Svendsen
   MiG Technologies

Navigate by Author: [Previous] [Next] [Author Search Index]
Navigate by Subject: [Previous] [Next] [Subject Search Index]

[Send Reply] [Send Message with New Topic]

[Search Selection] [Mailing List Home] [Home]