Simplified Balanced Valve

[rgb1]

"The average speed of air molecules at room temperature is ~1650 fps, independent of the presssure.... "


rms molecular velocity (denoted by lower case c) is pressure dependent
just like sonic velocity (a) and they are related by


       
      c/a = sqrt(3/gamma)


Hope this helps, Bob.


                                                                                 Ron

[PikeP]

Thanks Ron. My sheet isn't using the air velocity to calculate much currently so I havent put focus or time into it...but I put brief thought into it this morning..although with the distances I'm calculating even large velocity shifts aren't changing my calculatios more than .05% because air velocity is not the largest factor for my current formulations...

I changed to the following to get velocity based on KE and mass...however it may not be the best or most precise I think it's good enough for what I'm doing..thanks again!


Sqrt(2*ke/mass)

This puts atm air doing 1650~ and air at 4500 going 2850~...roughly

In most pcps air flow rate limits your pellet velocity more than air velocity...so it's not THAT important here as it would be in a theoretical maximum velocity event..which I'm not currently considering ..not saying air velocity plays no role it's just much less of one until you start going for absolute peaks..meaning air flow rate causes more delay through small orifices and at lower lifts. Where as air velocity + ke have more impact at peak flow rates with large orifices..

I'm calculating fill rates from 100 to 1500 micro seconds through orifices from .01 to .05. Air could move 1000 feet per second and still not effect the end result by much..going from 1400~ fps to 2700~ fps for example changes my estimated fill rate by .05%.

[rsterne]

Hey, Ron.... good to see you again.... I have a question about that formula.... If the quantity c/a can be determined by sqrt(3/gamma)…. or for that matter any quantity, fixed or variable, lets call it X = sqrt(3/gamma).... doesn't that mean that c is proportional to a ?.... What I mean is, the equation could be rearranged to....

c/a = X …. or  c = aX …. or  a = cX …. So, if you don't know the value of c, you can't know a.... and vice versa.... or am I missing something?.... We know that gamma varies somewhat with pressure, and we know that a also varies somewhat with pressure.... but is that because c is constant?.... Without more information I'm still a dummy.... 

I always used this calculator for the molecular speed....

http://hyperphysics.phy-astr.gsu.edu/hbase/Kinetic/kintem.html#c4

I don't see pressure or density there, just temperature and the atomic number (molar mass).... What am I missing?....

Bob

[PikeP]

For a giggle I ran calcs as low as 10 micro seconds. With my assumption that nearly no air flow occurs in the first 10 micro seconds you have to add that to the time pellet starts movement. From there I let the air go full speed 1650 fps...which is probably way high...ever catch a leak on your pcp and see bubbles slooowly grow when you use soap?

1650 fps eh?..Anyhow. Short distance .065ms long distance .091ms... I personally feel confident in these numbers and the margin of error considering the scale of time involved. I certainly didnt add any losses to the above times either with the exception of 10 micro seconds added to generate lift that can flow any appreciable air over time..and that is being very giving. And in air resistance the very low flow profile encounters among other losses and the long distance looks better and better.

[rsterne]

Mike, if I set Lloyd's spreadsheet up for 860 fps with 154 gr. bullet in a 27.5" barrel, using 108 cc of 3000 psi air, at 73% efficiency (actually pretty high, but that's what it takes to balance), using 2 lbs. of bullet drag and 5 lbs. of breakaway, with a 1.64 cc wasted transfer port volume, I get 2.44 mSec. for the dwell at 1.32 FPE/CI, and the valve is closing at 36% of the barrel length.... That's as close as I can get to my operating conditions for the .357 with 2T of gap on the SSG....

If I change ONLY the dwell and leave the efficiency at 73% (which certainly isn't the case for higher power, longer dwell shots, they get less efficient), I get the following results....

4.45 mS.… by adjusting the efficiency down to 68%, this is the dwell when the valve closes at pellet exit (100%), @ 0.66 FPE/CI.... reasonable assumption....
4.0 mS.… 921 fps @ 0.77 FPE/CI, valve closes at 88%.... To keep the velocity at 890, efficiency must be reduced to 69%....
3.0 mS.… 897 fps @ 1.07 FPE/CI, valve closes at 53%.... Velocity a bit too high, using efficiency of 71% I get 883 fps, which agrees with observed results at zero gap....
2.0 mS.… 810 fps @ 1.58 FPE/CI, valve closes at 25%.... Pretty good agreement with observed....
1.5 mS.… 725 fps @ 1.94 FPE/CI, valve closes at 14%.... Valve has gone "over the cliff" at this point, no valid test results....
1.22 mS.… 662 fps @ 2.16 FPE/CI, valve closes at 9.5%....   ditto.... This would be 150 FPE....
1.0 mS.… 605 fps @ 2.34 FPE/CI, valve closes at 6.4%....   ditto....
0.5 mS.… 467 fps @ 2.62 FPE/CI, valve closes at 1.6%....   ditto
0.1 mS.… 398 fps @ 2.63 FPE/CI, valve closes at 0.1%....   ditto

It is pretty obvious that Lloyd's spreadsheet doesn't work once the dwell is on the order of the lag time to fill the transfer ports, as it assumes the pellet is being accelerated from the get-go.... Perhaps if we added 100 uS (ie 0.1mS) to all those dwell times to allow for that, it might be a much better model.... Lloyd and I have discussed that, but as of yet that hasn't been done.... I hope you find the above useful....

Bob

PS.... I understand now that the 28% is not always applied, only for "good" PCPs.... makes perfect sense, thanks....

[mann]

So I tried a .150 transfer port Which is .050 smaller than what I had in place it seemed to give me more tunability tuned from 690 fps up to 800 fps is where I stopped I have to readjust the gap which after trying more tension on the ssg got to wide and I didn't have any more time to keep testing

[rsterne]

WHOA !!! …. Back up a minute, Mike !!! …. Where did THIS formula come from ? ….

Sqrt(2*ke/mass)

This puts atm air doing 1650~ and air at 4500 going 2850~...roughly

If you have some way to calculate maximum air velocity of 2850 fps at 4500 psi I need to know about it.... 

Bob

[rgb1]

I always used this calculator for the molecular speed....

http://hyperphysics.phy-astr.gsu.edu/hbase/Kinetic/kintem.html#c4

I don't see pressure or density there, just temperature and the atomic number (molar mass).... What am I missing?....

The answer is right there at the top ......."for an ideal gas"........

[PikeP]

WHOA !!! …. Back up a minute, Mike !!! …. Where did THIS formula come from ? ….

Sqrt(2*ke/mass)

This puts atm air doing 1650~ and air at 4500 going 2850~...roughly

If you have some way to calculate maximum air velocity of 2850 fps at 4500 psi I need to know about it.... 

Bob

Well it better explains the current obtained velocities no? Do you think 2200 + fps was obtained with no losses in air speed? Idk...and that 2850 at 4500 psi is theoretical max before any adiabatic cooling occurs which slows air speeds no?...its not what you'd ever obtain from a projectile...

I'm definitely open to different methods of obtaining airs *average* velocity and *peak* so I'll check out your link.

Dropping temp just a small amount of degrees reduces speed greatly, reducing KE due to losses takes that 2850 down further...see where I am trying to go? If your peak 2k+ fps shots were recorded when air was moving its peak velocity I'd be quite impressed! I'll calculate a few losses and report back with a new practical maximum out of a pcp..

[PikeP]

Bob, thank you very much for all that data. I'll try to tie a lot in to my sheet/brain and maybe make a dwell model from one of your tested shots comparing flow rates with dwell and lift so I can attempt to match lift/dwell somehow by simpy knowing flow. I'm going to have my work cut out for me...but I think itll help me understand the relationship between lift, dwell and flow rate better.  I think my guess at 1.5ms wasnt top far off for just doing a napkin guess..if I put eff to what lloyd predicts I get 1.37ms currently without tweaking a thing...how crazy is that? Two entirely different methods of obtaining lift and dwell in almost perfect agreement..within 10% or less of eachother... remarkable.

[PikeP]

Another thought on practical tune range..notice at very little close distance you still get lots of energy..the ejection rate can only be practically reduced x amount from peak before either valve doesn't open or acts very funny with huge swings in velocity when well below optimal velocity..but as you get to larger calibers that suggest you cant reduce power by near the % from peak compared to smaller calibers due to flow rates..unless we greatly choke off the transfer port or as you have on a build, at the bolt probe. So I think the tune range we found on yours is near perfect..certainly if you were up to my last revision. Gotta say I'm still beside myself how well this valve is operating.

[PikeP]

So I simply scaled lift to 25% more than the required lift to flow a set amount of air or rather KE at a specific efficiency...and my dwell times are all within a few micro seconds...I arrived at 2.87 vs 3.0 ms...and 1.32 ms vs 1.22 ms...next I will model both with flow and see what makes sense so I can tweak further. Thanks again Bob. I also get 14.8" valve close on the 270+ fpe shot now where lloyd predicted 14.84... 😁

Pretty crazy to arrive at the same figures using different methods..seems to solidify things a tad.

[rsterne]

Conventional valves can be tuned wayyyyyyyyyy down, even in a .457 cal.... They just don't exhibit the drop off like we see with any of the balanced valve when we back the hammer strike off too much.... I sure wish we could get our valves to open like a balanced valve and be adjustable like a conventional one.... 

Yeah, I want to have my cake and eat it too....   

Bob

[rsterne]

Don't forget, Lloyd's spreadhsheet runs strictly on Newtonian mechanics.... It includes the air mass that has been released by the valve with the mass of the projectile to be accelerated.... A later version, which I don't use, has that fraction adjustable.... IIRC, DomingoT said that using 1/3 the mass was a closer match to reality.... but that makes the air work like a spring (which it does after the valve closes)….  but during the valve open stage Lloyd found that including the air mass previously released worked the best, as that air, plus the mass of the bullet, is what is going to have to be accelerated in the next 10 uS (Lloyd's spreadsheet does a calculation each 0.00001 sec)....

BTW, you can toggle including the air mass on or off.... it makes a HUGE difference.... For the 2T Gap shot at the beginning of my last long post, which is 253 FPE, if you toggle the air mass off, and just accelerate the pellet.... you get 335 FPE.... The mass of gas in the barrel for that shot was 57.16 grains....

Bob

[mann]

I gained adjustability from using a smaller transfer port but it is no where near a conventional valve I wonder why any ideas Bob

[mann]

Is the tunability due to the valve not letting enough air threw to be able to fill the chamber. That why Travis said in order to get more he had to make the chamber as small as he could

[mann]

So really only way to get the adjustability would be to change the chamber size or to change the transfer port size . Need a adjustable transfer port like in the original mrod valve then you could tune it to where ever you want it

[mann]

Couldn't  a person just drill a hole into the valve transfer port like 8/32 or larger and use a long set screw to adjust enough so adjustability would be there

[mann]

I might do that so I could run my .200 port and not have to change to a different port just turn the screw to be able to take it to what ever fps you want to

[PikeP]

Yes man, if you read one of my messages up top I touch on transfer port restriction s furrher helping...if I were you I'd squeeze out volume like travis did...you have 3 other solutions if that isn't feasible but others require more work than the former. GL it's very achievable