Ongoing discussion of Ratio Deco

[KMD]

A few months ago I found a big nasty error in DPlanner. I emailed the support and never heard back. There's still an error in the program, and it seems no one cares. Scary.

Can you post this error and what version of dplanner you are using?

 

[FishTaco]

Can you post this error and what version of dplanner you are using?

Ditto.

Thanks,
Chris

 

[rainman_02]

Version 3.1.4

Here is what I emailed to decoplan@gue.com when I came across the "missing time." I understand it is not likely a realistic profile, but when writing a program one has to check out the "weird" cases on the boundary conditions.

I am writing my own deco software as an exercise in better understanding the theories behind neo-Haldanean and Bubble based decompression. As such, I have elected to compare the output of my code, based on Erik Baker's FORTRAN code, to commercially available decompression software packages such as DecoPlanner, V-planner, and GAP. This gives me two neo-Haldanean models and three Bubble models to compare with the output of my code. During testing of some rather unique profiles to check the different parts of my code, I came across an anomaly in DecoPlanner.

The profile consists of 20 minutes at 300', 20 minutes at 150' (intentionally violating the calculated ceiling to force the program to insert deco stops), and finally 20 minutes at 200' all using 12/60, before beginning the final ascent and deco. Deco gasses are 35/25 at 120', 50% at 70', and 100% at 20'. Descent rate is 75 fpm and ascent rate is 30 fpm for the whole dive, last stop at 10'.

DecoPlanner gives a total run time of 199 minutes. GAP's run time is 248 minutes. My program's run time is 263 minutes. All three are utilizing Buhlmann's ZHL-16 with GF 30/85.

Comparing the Bubble models, DecoPlanner's run time is 215 minutes, V-planner's is 213 minutes, GAP's is 214 minutes. I don't have my VPM-B implementation fully working yet. VPM-B (or RGBM in the case of GAP) with standard values and Conservatism +2 (12% increase in critical radii).

My concern is the significant difference between DecoPlanner and GAP (~20%). I know why I have a ~6% difference between GAP and my program -- GAP does not re-compute Gradient Factors when a descent is in between two ascents; my program does (as does DecoPlanner). The close precision (~0.5%) of all three Bubble models makes me think there might be something in DecoPlanner that needs to be addressed.

Also, please don't take this as an attack on DIR or GUE. I happen to be a fan of both. It is also not a pitch to sell my program as the exercise was mainly to help my understanding of the way the two major decompression algorithms worked.

 

[FishTaco]

Anyone care to apply ratio deco to rainman's profile, I'd be curious to see what kind of time it would generate.:blinking:

Take care,
Chris

 

[rjack321]

It will give the wrong answer.

 

[KMD]

Anyone care to apply ratio deco to rainman's profile, I'd be curious to see what kind of time it would generate.:blinking:

Take care,
Chris

Heck no. I am just going to take what Decoplanner or Vplanner or any thing else says to do on the 300' dive and follow it. I have never done a deco dive before, but I should be good. I mean the programs are based on scientific theory right?

Seriously though, with any profile, how much does experience play in the planning of the deco?

 

[ScubaMilo]

Anyone care to apply ratio deco to rainman's profile, I'd be curious to see what kind of time it would generate.:blinking:

Take care,
Chris

Let's see, avg depth of about 217fsw for 60min let's round it to 230 just to be safe.
Ratio deco says:125min of deco
Hmmm.......

Milo

 

[rjack321]

Heck no. I am just going to take what Decoplanner or Vplanner or any thing else says to do on the 300' dive and follow it. I have never done a deco dive before, but I should be good. I mean the programs are based on scientific theory right?

Go for it. Report back if you survive. :mooner:

Seriously though, with any profile, how much does experience play in the planning of the deco?

As of today, I see you have 134 posts which should be adequate amount of internet diving

 

[rjack321]

Let's see, avg depth of about 217fsw for 60min let's round it to 230 just to be safe.
Ratio deco says:125min of deco
Hmmm.......

Milo

So get yer' double 130s and a couple of deco bottles and let's dive!

Plan sh'man, we'll work it out on the way up! :cwmddd:

 

[ScubaMilo]

So get yer' double 130s and a couple of deco bottles and let's dive!

Plan sh'man, we'll work it out on the way up! :cwmddd:

Just you me and Darwin, We'll make a great team.

Milo