Simplified Balanced Valve

[mann]

If I make the chamber smaller or shorter it will fill faster and close the valve faster? Also helping tuning range ?

[rsterne]

Mark, making the chamber smaller does not make the valve closer faster, per se.... It allows the chamber to fill faster, preventing it from blowing open as easily.... That requires a bit more hammer strike at the top end of the power range.... and extends the tuning range as well.... The closing is primarily from the diameter/area of the stem, and that has a major effect on the hammer strike required to maintain dwell.... Generally, more dwell equals more velocity and FPE....

It's the old story of finding the best compromise.... If we reduce the hammer strike too much, then we lose the ability to use that hammer strike to vary the velocity.... At the limit, we end up with a valve like I had when I started (or a Cothran valve)…. ON or OFF, and no tuneability.... It is becoming increasing more obvious that you can't make the chamber too small, or the vent too large, IMO.... although past some reasonable limits you may not gain, either.... Mike is trying to define those limits, so we know when to stop trying for "more"....

Bob

[mann]

Thanks for explaining Bob helps me understand better on how it works

[PikeP]

Nice testing rig Bob. I do similar when testing
 to reduce hpa volume to nearly nil.

I'm confident Bob's current version even if needing tweaked like I advised with the spring swap is basically full proof for high powered big bore. Next tests will be small bore version and then anyone else's variation such as .1563" stems or other balance ratios.

[rsterne]

Mike, is there any point to looking at the barrel length in CALIBERS instead of inches, and the effect that has on the efficiency, dwell, and FPE ?.... Naval guns, and artillery pieces, are measured in calibers.... The 16" guns on the Iowa class Battleships were called a 16"/50 because the barrel was 50 calibers long (66 ft. 8 in.)…. A 25" barrel on a .25 cal is 100 calibers long, and the same ratio on a .50 cal would be 50" long.... I looked at this a while back and discussed it with Lloyd, and we thought it had a bearing on the efficiency of the shot.... A length of 100 calibers is probably about as long as we really need, I think....

Bob

[MJP]

You have now come to the point we where some years back, changing just the upper on a working lower.
All of the guns use the same valve, from .50 cal to .223

Now you found out for yourself that there is no disadvantage of the higher port volume, what you call wasted volume.

Nice work on the valve to get it bellcurve, something that I found no need for in my application.

Marko

[PikeP]

Bob I do think there is merit in your line of thinking with caliber length. So much on my mind I'll have to let that marinate for awhile.

After thinking about it..I think optimal dwell can be simply in the neighborhood of barrel length / bore diameter / 40...but the issue with using caliber lengths is very small barrels can take advantage of more dwell than really long ones..but it seems to scale sensibly even if 40 isn't the best constant. Between 30 and 40 seems reasonable. My current method of measuring dwell by metering air seems accurate for both low lift and full lift conditions regardless of the type of valve..which has a huge advantage to using newtonian mechanics to attempt to resolve it. The caliber length method is great for optimal dwell approximation but not low lift dwell / reduced power approximations unless we find some other scaling factor. Just my opinion. Dwell is the time it takes a dialating orifice to eject a specific mass of air and I dont think there's any better way to measure it than to run simulated computations for the dialating orifice down to the hundredth of an inch for every 100 micro seconds to build a real flow profile that determines lift within .001" and dwell within .01 MS. For now I use ejected mass / lifts average flow rate + time to create appreciable air flow at near .09 ms.

I ran some interesting throat/transfer plenum fill rate calcs..seems to be in the neighborhood of .25ms to get equal pressure into the transfer plenum due to choked flow...or .22 ms for a 1 cc .25 cal throat compared to .27ms for a 1.56 .357 throat. Which likely adds some micro seconds to our fill times in the pressure chamber, some of which are currently accounted for...but not all currently. For example the first .1 ms of lift has such choked flow that I'll likely add at least  75 to 100 micro seconds to the current fill times..I've come this far with precision might as well!

[rsterne]

The concept of "optimum dwell" concerns me a little, it depends on what you mean by optimum.... It is certainly valid for trying to optimize the balance between power and efficiency.... That is why for years I have been telling people that the valve should close before the pellet reaches halfway down the barrel.... which would be a divisor of 50 on a barrel length of 100 calibers (typical-ish for small bore airguns)…. Using Lloyd's spreadsheet, it appears that having the valve close when the pellet is 1/4-1/3 of the barrel length is more common.... and for very efficient PCPs (those approaching 2.0 FPE/CI) it is much sooner.... In fact the efficiency peaks for conventional valves at extremely short dwell.... Here is some data from a modded 2260 regulated at 1600 psi, varying only the hammer strike.... Yes, that was nearly 200 shots at about 620 fps on a 13 CI tank.... 



According to Lloyd's spreadsheet, at the peak of the efficiency curve, which occurred at about 620 fps, the dwell was only about 0.4 mSec. and the valve was closing when the pellet was only about 2.2% of the way down the 24" barrel (or roughly about 1/2" from rest) and only going about 200 fps.... All the rest of the velocity was obtained through the expansion of that tiny sip of HPA....  …. At 800 fps, with efficiency 1.50 FPE/CI, the valve dwell would be about 0.9 mSec. (10% of barrel length) and at 900 fps, with efficiency 1.30 FPE/CI, the dwell would be about 1.25 mSec (19%)…. At 990 fps, with efficiency of 1.05 FPE/CI, the dwell is about 1.6 mSec., with the valve closing when the pellet was about 1/3 of the way down the barrel.... As you can see, I did not have enough hammer strike available to get to the plateau, but Lloyd's spreadsheet predicts the efficiency would be down to about 0.4 FPE/CI at that point (valve open until pellet departs the muzzle), and the dwell to do that would be over 3 mSec.… I have found with numerous PCPs that if the valve closes when the pellet is about halfway down the barrel, the efficiency is about 0.7-1.0 FPE/CI.... The efficiency is DEFINITELY related to the pellet position when the valve closes (ie the dwell)…. Once the pellet is halfway down the barrel, the efficiency tanks.... it takes a huge increase in the amount of air released to get the last few fps.... Many PCPs only lose about 3% of their velocity when the valve closes at 50% compared to what they can do with a "dump shot".... Conversely, a .25 cal MRod with internal regulator tuned for a "40/40" shot string (40 shots at 40 FPE) will have a dwell of less than 1 mSec. and the valve will be closing when the pellet has travelled less than 3" down its 20" barrel.... In PCPs, the efficiency (in FPE/CI) peaks when the valve closes before the pellet reaches 10% of the barrel length, for many guns probably less than 5%....

Bob

PS, don't fall into the trap of thinking there is choked flow at the beginning of the shot cycle, when the valve is opening.... The pellet is not moving, or barely so, and therefore the velocity of the column of air is near zero.... All that is happening is that the molecules of air are filling the "wasted" transfer port space, in the same manner they are filling the vent to the balance chamber.... The limiting factor, as I understand it, is the molecular speed of the air molecules.... With a 1" long transfer port (valve seat to pellet base) and 1650 fps (19,800 in/sec) as the "average" molecular speed, it only takes about 0.05 mSec. for the air to reach the pellet base.... The valve is physically open quite far by then.... lots of area and no real flow velocity, so no choking....

Adding this.... At 0.1 mSec. the valve is open about 20% of its total lift, let's say 0.016".... and the pellet has only moved about 1/32" and moving about 40 fps.... I don't see how there is any possibility of choked flow....

RBS

[mann]

I switched hammers from 45 grams to a peek 31 gram and I may give the tss a try also with a few different springs

[PikeP]

Optimal dwell can be found and I have a lot of theory behind the mystery of dwell but not enough time spent pondering it to feel worth sharing..but I sense there's a lot of mystery surrounding dwell for you as well...pretty difficult to quantify. If you load the hammer with less spring it moves slower..poppet should also move slower.. sufficient air pressure to base of pellet to begin its movement is slower..the pellet itself moves slower because slower pressure build up...you see where I am going with this...meanwhile the difference in lift may only be. 02"...care to do that math? Btw I'm not implying lower energy shots have longer dwell than high energy I'm implying the difference in dwell between the two is less than you think but still present..imo

At .1 ms I get 20 mg per micro second flow rate through my valve and 158 mg per microsecond flow at .2 ms...and my throat is 450 mg...tell me again how enough air makes it to the pellet in .05ms to begin its movement..10 out of 450 mg is 2 percent...I think .1 ms pellet movement is much more likely in your average pcp...that's 100 micro seconds..now if it's really half that at 50 micro seconds..color me impressed. Care to run the flow numbers at that time resolution to tell me the pressure at the pellet at .05ms?

[mann]

How can the spring be way easier to cock and give me almost a 100 fps over the other spring this flat 13 lb spring cocks so easy

[mann]

Ok I switched to the lighter hammer backed off the spring and got the same speed out of the valve it sounds quiter I think the heavier hammer was really smacking the rear of the valve . Set right now at about 850 fps with 34 gr mk 2  13 lb flat spring which cocks easier than a 10 lb spring I need to see how many shots I can get

[PikeP]

Air to the base of the pellet traveling 1650 fps at 19.8 inches per ms takes .076 ms alone to travel the 1.5" in my valve...that's at a equal pressure not the rising throat pressure..that also assumes air accelerates from 0 to 1650 in 0 microseconds..

[PikeP]

More in regards to my lift equivalent port theory...which may be wrong. If you change transfer port size, it doesn't matter how hard you hit the valve that transfer port will be one of the variables determining energy output and that's due to its flow rate..same applies to throat and lift...you hit a valve only so hard to create an opening either equal to or a % of your transfer port and you get the expected energy output. I'm all about K.I.SS and the above theory works out on paper and in my head...I'm certain it's as close as I can get without using an order of hundreds of combined formulations and simulations, or as Lloyd is doing with real world tests...but even his data will only be as good as the equipment and method.

[mann]

Changing the transfer port to smaller size can and does control fps and lower es I've used that exact way to control fps and es on my Marauder 22 with a Cobra valve . Took some trial and error but it worked well

[PikeP]

Here is a simple dwell and lift model I made. Note that the average flow is 430~ mg/ms which incidentally is half of the lifts equivalent port flow rate(870)..because the average port flow rate matches the average predicted lift and dwell... it all ties together really well..doesn't mean its correct but it's the best I've come up with..this is an estimated 56 fpe. 25 cal dwell and lift model. The air mass of 795 mg ejected and the 874 mg per ms flow rate of equal porting are the only known variables in this equation..


If your average flow rate numbers dont match predicted dwell and lift which present a beautiful graph like below something isn't right. Predicted dwell here is 1.82ms

The theory behind this dwell model carries the same principle that got the balanced valve tunable so if it's wrong I guess I got real lucky solving balanced valves.

[rsterne]

As I understand it, the air molecules are already moving at 1650 fps, so they don't have to accelerate.... Now granted, only about half of them are moving in a direction that would take them through the valve.... so maybe that 0.1 mS is a better number, I haven't a clue, which is why all this stuff is still a mystery, right?....

OK, so here is a bit more data for you to chew on.... I hooked my .224 cal gun up to a regulator at 1830 psi and shot the 30.6 gr. NAA bullets.... All the way from zero gap to 4 T, the velocity was 1016-1018 fps (no real variation, random over those 4 turns)…. At 8 T gap it was 1010 fps, plus or minus a few fps, and at 6T "in between", again +- a few fps.... So basically solidly on the plateau, out past the gap where the valve quit working at 3000 psi.... Now it gets interesting.... At 10T gap, the ES started to pick up, I got anywhere from 970-997 fps.... At 12T gap it was unusable, the velocities seemed to alternate between 770-775 fps and 845-860 fps, every other shot.... This is the typical unstable operation I have seen with the Cothran and SS valves when you are too close to the cliff.... and indeed at 13T gap there was no shot, and even at 12.5 T it wouldn't fire.... So basically the valve reverted to untunable at 1830 psi.... In addition, I tried shooting a string at 10T gap, and got two shots, 995 fps and then 944 fps, ending at 1370 psi.... It would NOT shoot a 3rd shot at the same gap.... 

I was hoping I could just use less pressure in order to shoot a lighter bullet, but that did NOT work.... Possibly with a lighter hammer spring, so that the SSG gap was much smaller I may have been able to get it working.... but I could not just dial it down.... Any idea what the explanation is?....

Bob

[rsterne]

Nice model, and I get where you're going.... but I admit I have no basic understand of flow rates or the math behind it.... so you could be spot on, or dead wrong, and I wouldn't be able to input anything useful, one way or the other.... My understanding is that in a conventional valve, the lift to dwell curve was essentially a parabola, but if the lift exceeded 1/4 the throat diameter, the flow was clipped, like this.... In all the diagrams below, the mass of air flowed is roughly represented by the area under the curve.... (ie below the curtain limit)….



I further understood that the lift was proportional to the residual hammer energy, and the dwell was proportional to the residual hammer momentum.... residual being that which was left over after the valve cracked off the seat and the stem area x pressure was the primary (and relative constant) closing force.... Something like this happening if you change hammer mass M or velocity V.... Force F is closing force....



If those concepts are completely in error, then I will have to start at square one on my understanding of PCP Internal Ballistics.... Since I'm too old to go back to college, that isn't likely to happen.... so perhaps I should forget even trying to understand it at all.... 

Bob

[PikeP]

Heres a 20 fpe model same rifle. Notice how the average flow rate is half of the lifts flow rate of 470~ ejected mg at 244 mg/ms...1ms dwell..

I think this model accurately represents lift and dwell times thanks to the added z axis of air mass ejection...you cant just model lift and dwell with x and y without the z making sense of it all. Just my 2c...if I didn't scale the graph the lift and flow rates would be overlapped perfectly...

[rsterne]

The one thing I get from those two diagrams is that the area under the lift curves (or the flow rate curves) is ROUGHLY proportional to the FPE.... At least that much I seem to understand.... I would love to see them both on the same graph, plotted to the same scale, with the same instant of valve opening.... What is the throat diameter, and where does the curtain limit (D/4) kick in?.... or does it?....

Bob