Regulator Geeks 2: Scubapro's Balanced Regs

Please register or login

Welcome to ScubaBoard, the world's largest scuba diving community. Registration is not required to read the forums, but we encourage you to join. Joining has its benefits and enables you to participate in the discussions.

Benefits of registering include

  • Ability to post and comment on topics and discussions.
  • A Free photo gallery to share your dive photos with the world.
  • You can make this box go away

Joining is quick and easy. Log in or Register now!

Hi Regulator Geeks!

I'm following these discussion with great interest being crazy about scuba gear and an engineering student.

The amount of details and careful considerations is really nice!

It would be nice to have a chat about Atomic aquatics too, since they would in theory solve the annoying 'problem' of deposits - and their relative damages - inside an unsealed piston reg (read MK25)

Any chances for a Zoom meeting about them?
 
It's up!


Thanks for the likes :)
Oops completely missed it. What an epic ressource to have that discussiion available!

When is the next one? :)
 
Thanks for sharing these very educational insights and tests. I also loved the Mk19 video. I'm just learning about all this in my search to buy my first own equipment and I am amazed about the depth of knowledge you have developed and and also the intellectual integrity you show (challenging your own tests, and by taking in the comments you receive afterwards). Please keep sharing, it is a fantastic source of information for us (the ones that still need to be educated).

Greetings from Belgium,

Dirk
 
Whoa! Look what I just found!
There's a company in the UK that markets an ANSTI machine AND a surface test machine.
They use the surface test to certify regs for rescue divers and the like. I've attached the manual for their surface test machine, but here is the nugget:
"The high flow demanded by the CSTF approaches the peak flow requirements equivalent to a depth of 50 metres and 62.5 litres/min ventilation."

And the flow they're talking about is 1000 lpm, or 35.3 SCFM.

That means there IS a correlation between surface tests and a WOB loop performed at 170 feet.
So our tests last night at 16 SCFM were equivalent to a diver breathing 2.3 CFM (62.5 lpm) at what, 25m?
Maybe the differences in those curves IS significant at recreational depth...

Very interesting! I have often wondered how gas density and and the ip/ambient ratio factor in to reg performance. A 1000 l/m 'approaching' peak flow requirements for 50m@62.5l/m aligns with some ideas I have on the topic.

1707825464028.jpeg


This cool chart from the NEDU technical manual lists a peak flow of 3.27 l/s (196.2 l/m) at 62.5 l/m RMV, which at 50m is 1177l volume of gas atmospheric. (Does NEDU use an ANSTI machine? If not numbers might differ)
A regs flow rate reduces linearly with depth. Achim Schloeffel the guy from the 'InnerSpace Explorers' youtube channel tested an A700/MK25 at 130m (430 fsw) with air, it had a flow rate of 2200 l/m or 157l/m at that depth (it's advertised flow rate is 2000l/m), not quite adequate for 62.5 RMV but that's somewhat too deep for an air dive even for some of the yokels around these parts.

I think surface tests with high flow rates to simulate depth can be used to definitively confirm a reg will deliver enough air at a certain depth and get a general idea of it's breathing characteristics. They don't however account for the marked decrease in venturi effect at depth, be that due to gas density or expansion ratios I do not know but I would like to.

I have recently discovered your reg geeks presentations on youtube and they are beyond great! Which topics are in the running for the next one?
 

Attachments

  • MK25,A700 test.pdf
    78.1 KB · Views: 102
Thank you for resurrecting this dormant thread!
Well, there's a lot to unpack in your post, and this NEDU data plus Achim Schloeffel's test gives us some more to go on.
I'm hoping an engineer with expertise in gas flow can weigh in.
There are several variables I now see at play in trying to compare surface tests to an ANSTI loop:

What is the "x" factor to be applied to RMV in order to equate it to a surface test?
Reynold's numbers, gas viscosity and frictional components all play a role. While 62.5 lpm RMV is equivalent to 1177 l of gas at 1 atm, I'm not sure we can just run 1177 lpm thru a reg at the surface and look at our inhalation resistance and Venturi effects.

But the coolest info (for me, at least) in the NEDU chart is your peak flow data!
To know that 62.5 lpm RMV (which is an average of 1.04 l/s) actually has a peak flow more than three times as large(3.27 l/s) now puts a max slope on that sinusoidal curve. That validates my "back of the envelope" calculations in which I suggested that 80% of the RMV inhalation flow was accomplished in half of the inspiratory cycle, or 1/4 of the total cycle. And 62.5 x 0.8 x 4 is 200 lpm peak flow, or 3.33 l/s. So far, so good.

You have added a whole new wrinkle to the argument, however, in pointing out the diminishing gas expansion at depth! Venturi flow augmentation is a function of delta-pressure. And as you correctly point out, that changes drastically between the surface and 100 feet.
Gas at 150 psi absolute IP at the surface (135 + 14.7 psi) goes to 14.7 psi at the second stage valve, which is an expansion of 10:1 that generates its own velocity wholly apart from inhalation flow.
However, gas at 195 psi absolute IP at 100 feet (135 + 60 psi) now only has an expansion of 195/60, or 3.25:1. That's much less expansion velocity.

As you so nicely point out, on one side we have diminishing expansion velocity, suggesting reduced Venturi effectiveness.
On the other side we have 4x as many molecules passing through the system, which might (depending upon Reynolds numbers and stuff I know little about) tend to augment Venturi effects.

If the two cancel each other out exactly, then you could use surface flow tests (and the measured inhalation resistance, a la Pete Wolfinger) to approximate regulator performance without a $500k ANSTI machine. The CSTF machine mentioned above suggests a rough equivalence between 1000 lpm and 62.5 RMV at depth. But despite talking with them on the phone, I could never convince DiveLab ( You are being redirected... ) to do some ANSTI tests at increasing depths up to standard test depths, in order to quantify this.

If the two don't roughly cancel each other out, or if there's a whole 'nother factor I haven't considered, then Pete Wolfinger's "A.I.R. Test" in Regulator Savvy is irrelevant (and a lot of testing on my part has been a waste of time).

Like I said, I wish we could get a gas engineer into the discussion.
But congratulations on introducing the relative gas expansion factor into the Venturi discussion, @Michelle Louise !
Now I have to go find that YouTube video from Schloeffel. Maybe there's more data than in the attachment. His little article touting 2200 lpm gas flow didn't specify what gas he was using!! So the contribution of gas density to the Mk25EVO/A700's performance is unknown (air wasn't listed on his list of gases for the test, although he did say "2200 liters of air").

Oh! And the next Regulator Geeks will probably be on the Poseidon designs. IMO, the XStream is perhaps the finest diaphragm design in the world.
Screenshot_2016-06-07-08-22-40.jpg
 

Thank you for resurrecting this dormant thread!
Well, there's a lot to unpack in your post, and this NEDU data plus Achim Schloeffel's test gives us some more to go on.
I'm hoping an engineer with expertise in gas flow can weigh in.
There are several variables I now see at play in trying to compare surface tests to an ANSTI loop:

What is the "x" factor to be applied to RMV in order to equate it to a surface test?
Reynold's numbers, gas viscosity and frictional components all play a role. While 62.5 lpm RMV is equivalent to 1177 l of gas at 1 atm, I'm not sure we can just run 1177 lpm thru a reg at the surface and look at our inhalation resistance and Venturi effects.

But the coolest info (for me, at least) in the NEDU chart is your peak flow data!
To know that 62.5 lpm RMV (which is an average of 1.04 l/s) actually has a peak flow more than three times as large(3.27 l/s) now puts a max slope on that sinusoidal curve. That validates my "back of the envelope" calculations in which I suggested that 80% of the RMV inhalation flow was accomplished in half of the inspiratory cycle, or 1/4 of the total cycle. And 62.5 x 0.8 x 4 is 200 lpm peak flow, or 3.33 l/s. So far, so good.

You have added a whole new wrinkle to the argument, however, in pointing out the diminishing gas expansion at depth! Venturi flow augmentation is a function of delta-pressure. And as you correctly point out, that changes drastically between the surface and 100 feet.
Gas at 150 psi absolute IP at the surface (135 + 14.7 psi) goes to 14.7 psi at the second stage valve, which is an expansion of 10:1 that generates its own velocity wholly apart from inhalation flow.
However, gas at 195 psi absolute IP at 100 feet (135 + 60 psi) now only has an expansion of 195/60, or 3.25:1. That's much less expansion velocity.

As you so nicely point out, on one side we have diminishing expansion velocity, suggesting reduced Venturi effectiveness.
On the other side we have 4x as many molecules passing through the system, which might (depending upon Reynolds numbers and stuff I know little about) tend to augment Venturi effects.

If the two cancel each other out exactly, then you could use surface flow tests (and the measured inhalation resistance, a la Pete Wolfinger) to approximate regulator performance without a $500k ANSTI machine. The CSTF machine mentioned above suggests a rough equivalence between 1000 lpm and 62.5 RMV at depth. But despite talking with them on the phone, I could never convince DiveLab ( You are being redirected... ) to do some ANSTI tests at increasing depths up to standard test depths, in order to quantify this.

If the two don't roughly cancel each other out, or if there's a whole 'nother factor I haven't considered, then Pete Wolfinger's "A.I.R. Test" in Regulator Savvy is irrelevant (and a lot of testing on my part has been a waste of time).

Like I said, I wish we could get a gas engineer into the discussion.
But congratulations on introducing the relative gas expansion factor into the Venturi discussion, @Michelle Louise !
Now I have to go find that YouTube video from Schloeffel. Maybe there's more data than in the attachment. His little article touting 2200 lpm gas flow didn't specify what gas he was using!! So the contribution of gas density to the Mk25EVO/A700's performance is unknown (air wasn't listed on his list of gases for the test, although he did say "2200 liters of air").

Oh! And the next Regulator Geeks will probably be on the Poseidon designs. IMO, the XStream is perhaps the finest diaphragm design in the world.
View attachment 826887

What is the "x" factor indeed. You are right greater gas density almost certainly augments venturi now that I think about it, and partially makes up for the loss of it due to lessening expansion. I do not think they cancel each other out completely, but that is just a feeling.

1708027321234.jpeg


Relative gas expansion is interesting and no one pays any attention to it. It's the reason I'm comfortable ascending from depth up to 40m or so quite fast, and very slowly once I switch to EAN50. The same concept holds true for ip/ambient and it's alleged meddling with our reg performance.

What other variables does increasing density/viscosity/reynolds introduce? Adiabatic cooling? Probably a few others, perhaps of little significance. I might go searching for some niche engineering forum =) Comparing existing ANSTI data to surface test results of the same reg could be of some value too.

Schloeffel didn't make a video about his test, he just wrote it up for a German forum, in the original German he specifies it was air.

I do think it is fair to say that that a regs flow rate follows boyles law, perhaps it is a bit more but nonetheless a tight correlation. Which lends credence to surface tests. If we take a look at regs in the 1987 NEDU tests, compare when their gas delivery fails to it's manufacturer specified flow rate, things line up.

Here is the original G250, old catalog says flow rate is 920 lp/m:
1708027406909.png


Plug in the NEDU peak flow figures from my last post and we get:
75RMV,50M --> 2.75x60x6=990 lp/m
90RMV,40M --> 3.93x60x5=1179 lp/m

The only reg that achieved 90RMV @ 300 feet (2385 l/m peak flow), was the Poseidon Odin/Jetstream, 2100 l/m advertised flow rate which checks out.
1708028093678.jpeg


Here is the PDF link to the NEDU tests for those interested

@rsingler A Poseidon deep dive would be fascinating, the Xstream was the first significant evolution in first stage design in decades. Servo seconds could be interesting too, as you have done barrel and center balanced. I'm looking forward to it and appreciate the time/effort.
 
Oh I have been thinking about this a bunch -- well in between dreaming about/hunting the next reg -- since I first saw the reggeeks sieries back in last Dec.

And @Michelle Louise you have certainly added to the dimensionality of it -- my brain is out of RAM 😅

Excellent post!

In a previous lifetime I used to be a hudraullics/pneumatics engineer (at least through uni & grad school), funny enough not since as I moved into software, and you have certainly taken me back

I think the toughest part is seperating the various (numerical, as well as literall) "manifolds" that represent various ranges of dynamics at play (and I am probably overthinking parts of this):
1- Abs (IP) pressure gradient between the 1st & 2nd stage, has a tiny (negligable) amount of drop that is best modelled as a 1-D pressure wave {tho probably thsi is more negligable that the HP/IP change in the 1st stage as this is a behaviour that takes effect at really high pressures like in diesel injections sytems: ~2K+ bar}
2- this (1) probably (I have no numbers to crunch here) is dependant on the ratio between IP and ambient pressure, as well as gas (density) and ambient temprature; and from it adiabatic cooling comes also into play.
3- Factoring in 1&2, Reynolds number would wildly vary; and we still didn't go beyond the jam nut on the second stage
4- we finally reach the oriface, where the 1st significant venturi phenomena (on IP level) happens due to transition from barrel to orriface, then the diameter diff. on the orrifice sides itself; then right after movig back into the barrel
5- Now we have restrictions to consider, the poppet and the spring -- and depending on the Re# we are either stabilizing transitional laminar flow (I really doubt that) from going full turbulant; or just crating more turbulance if we are in the Turbulant flow range (more probable)
6- we finally leave the barrel, and start the next venturi phenomenam the big one, getting really turbulant in the case
7- the last 2 ventuir effects, the flow vane (deflection), and the mouthpiece (constriction)

All along that path we are changing absloute IP, since air is indeed very compresable, and I believe that air/gas leaves the mouth piece at Ambient pressure we have in extreme cases (100m) around a 9~10bar drop between 1st stage and mouthpiece

I think this is a very chaotic chaotic system, so your obsorvations are on point, they just wouldn't corelate in a linear fasion (or even a "simple" non linear one)
Probably why the built these expensive machines -- unless we have very detailed and complete simulations/calculations for the system in its entierty + a supercomuter we can't fully tell the factor at which the results "dampen"

on the other hand, it porbably can be "linearized" in a principal component analysis-- and probably that "factor" would look like an inverse or maybe (scaled) exp decay 🤷🏽‍♀️, and probably can be simplified to something like that NEDU table you have shared.

we just have to have access to some ANSTI test data (and time) to do that

I wish I could have put something more helpful than that, but again I haven't touched that stuff since 2016

EDIT: I will look into some of the specifics that are dull in my brain, maybe something more helpful comes up

Thanks again, you've taken me back to seomthing dear that I haven't thought about in years, also reminded meof one of my favourite "poems":
"Big whorls have little whorls
Which feed on their velocity,
And little whorls have lesser whorls,
And so on to viscousity" -Lewis F. Richardson (foreword of Strange Attractors ch., Chaos, by James Gleick)
 
5- Now we have restrictions to consider, the poppet and the spring -- and depending on the Re# we are either stabilizing transitional laminar flow (I really doubt that) from going full turbulant; or just crating more turbulance if we are in the Turbulant flow range (more probable)
Avoiding 5 is where the advantages of Mares' VAD is supposed to start.


The other advantage is that by channeling the higher pressure gas away from the diaphragm (instead of emerging right next to it), you get a relatively low pressure area adjacent to the diaphragm which lowers (or zeroes) the effort needed to keep the lever depressed once the inhalation has begun. And it does this without the need for Venturi vanes.

The theoretical disadvantage is that the lack of Venturi vanes means you can't tune the feel of the inhalation other than by adjusting cracking pressure.
 
https://www.shearwater.com/products/perdix-ai/

Back
Top Bottom