Ongoing discussion of Ratio Deco

[Phil K.]

Catching this flaw is no small matter. You're doing us all a service. Your post is one of the most important uses of a public forum like this. I don't think you should let it go with an email to tech support. Why not email JJ directly at jarrod@gue.com?

 

[rjack321]

I'm not sure plugging in an insane profile and getting an insane answer is a "flaw".

 

[rainman_02]

In that case, then all three software packages should agree within a certain tolerance...since all three are supposedly running the same basic algorithm (Buhlmann with Gradient Factors) and all three probably started with Eric Baker's FORTRAN code.

Just as I would expect an unmodified VPM-B and RGBM profile to be pretty close to each other (since they're almost the same algorithm...both based on the original VPM), I would expect various Buhlmann algorithms to be similar.

Redundancy in computing applications requires a minimum of three computers running the same algorithm. The system is fine as long as all three agree within tolerance. When one starts to diverge, it is singled out as a "bad" computer. This is evident in the way airplanes crossed the ocean before GPS (triple Inertial Navigation Units). It is also evident in the Space Shuttle navigation system. Three primary computers running the same algorithm. When one diverges out of tolerance, a fourth running the same algorithm is brought into comparison with the two "good" computers, assuming they are both good since they match. If the new computer doesn't match the other two, a fifth computer is brought in, but it is running a different software package that should return the same information as the first four. The true "bad" computer is singled out by comparison.

As such, regardless of the profile, when two computer algorithms (based on the same equations) compute one solution and a third computes a different solution, the most likely problem is with the one that is different (thank you Occam).

But thank you for confirming that people don't care because "it's not a profile *I'd* ever dive!"

Phil K., I'll send it to JJ and see what he says.

 

[rjack321]

They aren't actually redundant because they may or may not be based on the same source code. Clearly you are outside the specs of the code (in at least one case, possibly all three). And one gives a different answer, not a big surprise.

This is like planning a moon mission with the space shuttle. An "incorrect" answer is not necessarily meaningful.

And, this is a thread about Ratio Deco. How is this "flaw" comparable or relevant?

 

[lamont]

They aren't actually redundant because they may or may not be based on the same source code. Clearly you are outside the specs of the code (in at least one case, possibly all three). And one gives a different answer, not a big surprise.

goes to the correctness of the underlying algorithm. large differences on psychotic profiles may be smaller but still significant differences on sane ones...

 

[rainman_02]

Lamont, that's exactly my point. And it is why people test the boundary conditions of their code. If it works at the boundaries (where most errors occur) then it will most likely work within the useful range of the program.

It is relevant because DecoPlanner is the GUE method of comparison of a profile to ensure "proper" deco. But if the baseline assumption (the algorithm in DecoPlanner, in VPM, or in Ratio Deco) is not accurate, how can the results be accurate?

 

[rjack321]

goes to the correctness of the underlying algorithm. large differences on psychotic profiles may be smaller but still significant differences on sane ones...

That's why friends don't let friends dive decoplanner

 

[Kevrumbo]

Number 1
Here is GI's profile
66m/220' for 25 min

Depth Time Gas
66/220 25 16/50

48/160 :20
45/150 :20
42/140 :20
39/130 :20
36/120 :20
33/110 1
30/100 1
27/90 1
24/80 1
21/70 5 50%
18/60 2
15/50 3
12/40 5
9/30 8
6/20 13 O2
Total 42
******
Here is the Ratio Deco 1:2 profile
That means 1 min BT = 2 min deco on two bottles @ 220'/66m
For every min of bt do 2 min Deco on the two deco gases. The set point forthis ratio is 66m/220
So for 25 bt then do 50min deco on two bottles
25 min on the Nitrox 50
25 min on the O2
Make sure the 36m/120-24m/80' is 1/2 the time of the nitrox 50 = 12min or so

66m/220' for 25 min

Depth Time Gas
66/220 25 16/50

51/170 1 80% ata's
48/160 1
45/150 1
42/140 1
39/130 1
36/120 2 65% ata's
33/110 2
30/100 2
27/90 2
24/80 2
21/70 7 50%
18/60 7
15/50 3
12/40 3
9/30 5
6/20 12 on 6 off and 6 up O2

Total 50 min on the deco bottles (Nitrox 50 and O2) plus deep stops.
++++++
As an aside the Decoplan total deco time is = 54 min starting at 36m for 1
min. So as you can see a little longer than yours and therefore certainly not dangerous or bending people as you claim. In fact Decoplan on a whole is
even shorter with a total of 54 min and mine is 65.
------
Number 2
Here is another GI profile
50m/190' for 20 min [Typo error; should be 51m/170']
Depth Time Gas
51/170 Bt= 20 16/50

39m/130' :30
36m/120' :30
33m/110' :30
30m/100' :30
27m/90' :30
24m/80' :30
21m//70' 3 50%
18m/60' 3
15m/50' 3
12m/40' 3
9m/30' 5
6m/20' 8
Total 25min
**
Here is a Ratio Deco 1:1 profile @ 45m/150 and 1 deco gas.
Again this is a BT = DT using a single Deco of Nitrox 50% @ 45m/150 and because it is 51m/170' then you add 10min to the deco (5min fo reach 10' deeper)
So 51m/170' for 20 min is 30min deco
21m/70' - 9m/30' is 15min
6m/20 - 0m/0;' is 15 min -- Most at 6m/20' and 6 up

51m/170' for 20 min

Depth Time Gas
51/170 Bt= 20 16/50

39m/130' 1
36m/120' 1
33m/110' 1
30m/100' 1
27m/90' 1
24m/80' 1
21m//70' 4 50%
18m/60' 4
15m/50' 2
12m/40' 2
9m/30' 3
6m/20' 9
6 up
Total 30min
++++++
Again as an aside, Here Deco plan total deco is 20min deco with first stop at 21m/70'. So again much more conservative.
----------
Number 3
Here is an another GI profile
38m/130' for 45 min

Depth Time Gas
38/130 45 21/35

27/90' 1
24/80' 1
21/70' 3 50%
18/60' 2
15/50' 2
12/40' 3
9/ 30' 5
6/20' 8

Total 25
*****
Here is the Ratio Deco equivalent profile
BT = DT 1:1 @ 45m/150 but because we are shallower by 6m/20' then we can subtract 10 min. So 45min BT = 45 min deco less you 10 min is a total of 35 min on the Nitrox 50 bottle.

38m/130' for 45 min

Depth Time Gas
38/130 45 21/35

27/90' 2 - 80% ATA's
24/80' 2
21/70' 5 50%
18/60' 5
15/50' 2
12/40' 2
9/ 30' 3
6/20' 12
6 up
Total 35 on the nitrox 50 bottle plus deep stops.
+++
As an aside the decoplan is total deco 37min with a first stop at 21m/70'
again I am more conservative.
___
Number 4
Here is another GI profile
60m/200ft 120min

48m/160 4
45m/150 4
42m/140 4
39m/130 4
36m/120 10
33m/110 10
30m/100 8
27m/90 7
24m/80 6
21m/70 25
18m/60 20
15m/50 20
12m/40 15
9m/30 15
6m/20 120
to 0 20

******
Here is my Ratio Deco profile
Bt=2 X DT @ 66m/220 or 1:2 Ratio
So 60m for 120 min is 240min on two bottles if 66m./220' but we are doing 60m/200' dive so 6m/20' is 10 min less. So 120 bt = 230 Deco on two bottles
1/2 for the nitrox 50 = 115 min
1/2 for the O2 = 115 min
Make sure the 36m/120 bottle is 50% of the Nitrox 50 time = 57min

60m/200ft 120min

48m/160 4 - 80% of ATA's
45m/150 4
42m/140 4
39m/130 4
36m/120 13 - 35/30
33m/110 13
30m/100 9
27m/90 9
24m/80 11
21m/70 30 -- Nitrox 50
18m/60 30
15m/50 15
12m/40 15
9m/30 25
6m/20 115 -- O2
to 0 20 20% of o2 time

Total Deco from 36m is 287min
++++
As an aside the deco plan is Total deco = 289min with 39m/130 as the first stop
____
Number 5
Here is another GI profile:
90m 30min 10/70

72 1
69 1
66 1
63 1
60 1
57 1
54 2
51 2
48 2
45 2
42 2
39 2
36 5
33 4
30 4
27 3
24 2
21 5 50%
18 5
15 5
12 8
9 12
6 34
10 up

Total 100 min

Here is another Ratio Deco 3:1 profile:
90m 30min 10/70
Is 1:3 so for each min of bottom do 3 min of deco on three bottles @ 81m/270
But because we are at 90m/300' then we add 15min
So BT 30min = 90min deco + 15min
Total = 105 on three bottles
3 bottles
O2 is 40% of total
Nitrox 50% is 40% of total
36m/120' is 20% of total
Still do 80% ATA is 1 min
65% ata is 3 until you reach the gas switch

72m/240' 1
69m 1
66m 1
63m 1
60m 1
57m 1
54m 3
51m 3
48m 3
45m 3
42m 3
39m 3
36m 5
33m 5
30m 3
27m 3
24m 4
21m 12 50%
18m 12
15m 6
12m 6
9m 8
6m 36
8 up
Total 105 min for the three bottles plus your deep stops
As aside Decoplan is total deco of 151 with a first stop at 57m/190

Full context can be viewed here (and a good discussion of the incident which precipitated it all). . .

 

[Kevrumbo]

Ratio Deco/Gas Planning Excel '97 version attached below (Enable Macros when prompted). . .

 

[larry nichols]

im very greenhorn is there a quick way to summorize what u guys are discussing?