Timberwolf resurrection project

[Ragnarok]

In the interests of starting to cross some of my long list of unfinished projects off my to-do list, I'm starting to look back at the Timberwolf project - euphemistically, "HEAL 2"

Given costs and difficulties, the plan had been changed to achieve things on a step-by-step basis. A lot of parts were compatible with (although not identical to) HEAL, so it was possible to substitute in parts from HEAL and get a working end result until the upgraded parts could be finished.

The current step is to get Timberwolf's valve and chamber working and fitted to enough of HEAL to produce a working result. Technologically, this is still going to be a breech loading pneumatic (rather than the semi-auto Timberwolf was envisioned as) but it's still a pretty big upgrade.

Timberwolf's valve design has been strengthened, so should safely be able to handle 40 bar (580 psi) rather than the 25 bar I had limited HEAL's valve to, and has also been made somewhat more efficient; similarly, the chamber has all the fittings modified to remove hard edges in the flow, which should improve the flow characteristics to an extent.
It should therefore pack a decent punch for the calibre.

~~~~~

The task list is some aesthetic things like cleaning/varnishing the stock and finishing off the scope mount, a few more structural things like the stock clamps and barrel mountings (rather than using electrical tape ) and then a few functional things - finishing the assembly of the parts I've actually got and working through a moderate shopping list.

Things like swapping to 6mm OD tubing (rather than 8mm) between trigger and the pilot QEV, which should reduce the volume by about half and speed up trigger response a little. That's not too hard a decision.

The bigger questions are:

Piston construction:
HEAL's piston is a "bolt" piston, based on M6 studding, nuts and large size washers - all steel, save for the rubber part, for a weight of 70 grams.
It's possible to buy aluminium studding and nuts instead and while getting the right size of washers is more of a challenge, those can be easily fabricated from plate; a complete materials replacement would let Timberwolf's piston weigh about 30 grams.

However, as I theorised recently, piston weight doesn't make a catastrophic difference to either performance or piston energy*.
*Actually, I think in certain performance ranges, piston valves should be able to outperform burst disks. Burst disks either create debris or fold open into the flow, neither of which is going to be beneficial for maximising the flow. In this case, the cleaner flow is probably worth more than a faster opening speed.

So I'm primarily more concerned about piston strength. I'm thinking a compromise might have to be made on mixing in UHMW (or similar) pieces to help transfer forces better. I'm not possessed of the lathe that would be ideal for that work, but I think something could be improvised.

This is might be about the ideal I can think of:

Bolt_piston_400.jpg (27.03 KiB) Viewed 5452 times

... not to scale, mind you.

Loosely, red is rubber or similar, blue is aluminium parts, grey UHMW, black copper/brass and green is the modifications to the fittings.

From the right, the clusters of rubber washers are the sealing face, check valve, piston bumper and valve bumper.
As rubber distorts rather than compresses under force (a layman's way of saying "Poisson's ratio of 0.5"), the piston bumper can also work as a friction brake, expanding out as force is applied to the rear of the piston and braking against the valve walls.
(The valve rear bumper is more likely to be sorbothane than just a more general rubber though).

40+ bar air source:
I'm not well off for fridge compressors - our local dumps are funny about people taking things away, and replacement ones come with a bit of a price tag - and given they don't necessarily hit 40 bar anyway, I'm thinking the best solution here is to build a two stage track pump.

I'm thinking a ~26mm ID up-stroke pre-compressing for a ~13.5mm ID down-stroke. This should shift a fair bit of air at a time and, for my body weight, should convincingly hit 60 bar. It'll be a workout, but it should shift air at an appreciable rate and pressure.

Eventually, the plan will be to adapt 2 or 3 of the CO2 extinguishers I have around into an air reservoir; that'll be a REAL workout, but should be able to support something like half a dozen regulated shots and will be a step towards the semi-automatic goals of Timberwolf.

Pressure ratings:
I'm not too worried about most of the pressure ratings I will be exceeding, given that the ones I'm going over by that great a margin are mostly exhaust valves and thus conservatively rated due to the noise and potential risk of debris.

The one thing that has got me on edge is the check valve on the filling setup, which is only rated to 10 bar. I've run it up to 25, but I've had problems with similar ones in the past.
The choices I can run up for a reasonable price are either pneumatic check valves in the 16-17 bar range or hydraulic check valves that carry much bigger ratings (like 350 bar).

I know pneumatic fittings are naturally far more conservatively rated than hydraulic fittings, given that air is more energetic and more "leak finding", and so the functional "for air" rating of a hydraulic valve won't be as high as that, but have little experience of how not as high. I'm guessing the hydraulic valve is probably still the safer option though.

The QEV is theoretically only 10 bar rated as well, but I'm fairly sure I recall that similar valves have been tested a good way past 40 bar by others. It's probably the "piston" that's the limiting factor here, although I'm not immediately well off for a way to upgrade it.

~~~~~

I'm open to suggestions on any of these fronts though.

Really, the goal at this stage is to reach the power potential of the Timberwolf valve/chamber - semi-auto components can come later.

[velocity3x]

I doubt your front barrel seal will survive beyound the first shot because of the explosive decompression it will suffer at 40 bar. One of my designs shoots 70 bar x 32mm barrel. I devoted much time to finding the simple materal that would survive the self-destruction of launch. Use a hocky puck for the front seal. Hocky Pucks are machineable, can be turned to your reuired diameter and they will not self-destruct from explosive decompression due to their 95 durometer.

[Ragnarok]

I'm not quite sure what you're suggesting is going to happen?

The piston face going walkabout down the barrel seems doubtful, as the design will have it physically clamped between two washers (although the front one will likely be larger than I've shown it in the diagram, close to the barrel diameter). It'll probably also be generously epoxied to solve that typical bolt piston problem where air leaks through the bolt threads.

There's a lot holding the piston face in place - my concerns are more that the valve seat could go cookie cutter on it. If anything is likely to take a trip on firing, my suspicions are more directed at the valve bumper, which won't be as well held in place.

[DYI]

I believe velocity is referring to the common problem with high pressure seals where a rapid decrease in pressure does not allow sufficient time for gas previously absorbed by the seal to flow out, causing the seal to literally explode from internal pressure. Generally not a problem at PVC pneumatic type pressures or in low pressure hybrids (in the second case, there is insufficient time spent at high pressure for the seal to absorb gas), but it will become problematic at some point in higher-pressure pneumatic launchers.

Luckily, it's a well characterized problem in industrial applications and there's no shortage of available data on it.

[Ragnarok]

I see. Well, that's not specifically a problem restricted to the sealing face then.

[jackssmirkingrevenge]

So I'm primarily more concerned about piston strength. I'm thinking a compromise might have to be made on mixing in UHMW (or similar) pieces to help transfer forces better. I'm not possessed of the lathe that would be ideal for that work, but I think something could be improvised.

You are, however, a member of a forum whereseveral folk offer machining services at reasonable prices, no excuse to have something bodged out of nuts and bolts

[Ragnarok]

Having some parts properly machined is being considered, but as far as the piston, I'd have to be struck by a grand idea for a hugely better design for something other than a "bodge" of nuts and bolts to be worth it.
As is, the friction braked element of HEAL's piston proved itself to the point that it's something I definitely want to keep, but I've not yet thought of a way to achieve it in a radically different piston design.

Mind you, I am also excruciatingly exhausted at present.

[velocity3x]

Having some parts properly machined is being considered, but as far as the piston, I'd have to be struck by a grand idea for a hugely better design for something other than a "bodge" of nuts and bolts to be worth it.

It may not be as "Grand" as a "bodge of nuts and bolts" .... but it works exceedingly well.
http://www.spudfiles.com/pneumatic-cann ... er#p324300

[Ragnarok]

Point #1: Please don't misquote me.

Point #2: A dashpot could provide a useful solution to replacing (or partially replacing) the valve bumper, but I'm considerably happier transferring at least a portion of the forces to the valve walls.
The pipe is always going to be stronger than the rear fitting, and while I do trust the rear fitting to hold up, I'd also rather have a redundancy to increase the safety factor (given the piston is the part of the launcher that actively fires itself at my shoulder).

In an ideal world, there is a much better design for a rear fitting, but it would require machining of both internal and external pressure-holding threads; and in the long run, there's only so much money that's worth throwing at an alternative design unless it's going to dramatically improve piston mass and endurance*.
*But I am looking at the problem of sealing face materials as a separate matter.

A piston of 70 grams has already worked very well - the parts to get that down into the 30 gram region go pretty cheap. There's only so much money worth throwing at further advancements.

To have it my ideal way would start with the raw material costs of large diameter brass round bar that needs some moderately precise machining; even before you get to the piston costs it starts to look very not cheap as compared to a solution that, while not the prettiest, will do the job almost as well.

That said, you have given me ideas for things to change, but I'm not yet seeing a motivation to throw a larger budget at this particular problem.

[dart guy]

sorry never saw the timberwolf, mind posting a link or somethin?

[tigerblues28]

His other H.E.A.L cannon was sweet, Dart.

Ill be pleased to see how this turns out, Rag.


Nick

[Ragnarok]

Timberwolf is an as of yet uncomplete project, so there's no link; However, it is based on HEAL to the extent that several parts are interchangeable.

Visually, the first stage of this project will look a lot like the current configuration of HEAL, but will still be a major step up in power.

In the long run, it will receive an integral six shot rotary magazine (based on the Webley Fosbery auto-revolver but pneumatically, rather than recoil, operated) and become a semi-auto.
The design of the magazine does limit total shot count but should be able to feed a wide range of different ammunition types (unlike most of the semi-autos around here) and doesn't increase the dead volume. And when it comes down to it, even a mere six shots will make quite an impression at this pressure and calibre.

[Ragnarok]

Something else I've decided needs to be looked at is a suppressor.

The easily accessible sizes of suitably strong PVC pipe (given that Timberwolf will have fairly generous pressures at the muzzle, I'd prefer to have pipe that at least pretends to be strong) are 75mm (68mm ID) and 90mm (81mm ID).

Now, the general rule with suppressors is volume (in the sense of capacity, not sound) makes for an effective result. Downside is that 90mm pipe will be considerably heavier and a Photoshop mock-up of a suppressor that large isn't very visually appealing.

So I'm probably thinking 75mm pipe and extend a vented length of barrel through it; probably off-centred, so that most of the suppressor's volume is below the bore axis.
Make it about 30cm long and it'll still have a litre of capacity; this is a little more than chamber and barrel put together, so it'll have some effect on pressure.

Still, I've been trying to find ways to make that 75mm pipe punch a little above its weight.

One point was packing materials. If I recall, common wisdom has been that packing materials like steel wool are a bad choice for pneumatics, as the gases are below ambient temperature.

However... I've poked a calculator a few times. At least in the case of Timberwolf, the adiabatic compression of the gas already in the suppressor actually causes the temperature to rise above ambient temperature by a margin of around 100K. Effectively, that immediately makes a quarter of the suppressor's volume disappear as compared to keeping everything at ambient temperature - so it may be worth putting some of that internal volume to use as a heat sink.

At this point, I started wondering if it's worth looking at Aluminium pipe, to better absorb the heat. 76mm Alu is a bit more expensive than 75mm PVC, but it has thinner walls, so it's no heavier, and actually increases suppressor volume by 10% or so. Seems like a plus.

The other idea I have in mind is reed valves. It's a bit off the wall - I know it's not used in firearms, but the (comparatively) low pressures and high gas volumes here may actually give it potential. It has two goals:
1) Being able to release the gases from the suppressor cavities slower than they got in there is quieter.
2) Because the gases won't flow back into the barrel, the earlier cavities could contain more gases than they would otherwise (the later cavities can reduce barrel pressure further and potentially allow back-flow).

The flaw is obviously that reed valves can potentially reduce the flow into the suppressor - but if it does get good flow and heat transfer, adding these two factors could be a big boost.
The assumption of perfect flow/transfer would allow dropping muzzle exhaust pressure by a further 60% fraction and duration by 40%.

It's all rather wacky, but it might just be worth an experiment.

But even if all the above conjecture is utter piffle, a suppressor is still going to be a good idea.

[jackssmirkingrevenge]

Had you seen this? http://www.spudfiles.com/spud-cannon-re ... 23931.html

Not exactly a reed valve but didn't have much luck with it.

However... I've poked a calculator a few times. At least in the case of Timberwolf, the adiabatic compression of the gas already in the suppressor actually causes the temperature to rise above ambient temperature by a margin of around 100K. Effectively, that immediately makes a quarter of the suppressor's volume disappear as compared to keeping everything at ambient temperature - so it may be worth putting some of that internal volume to use as a heat sink.

I don't think this is a factor, baffles are what you want. Baffles that create turbulence behind the projectile that traps as much gas as possible in each chamber.

[youtube][/youtube]

Another thing I've discovered is that it's not only volume that counts, but how that volume is distributed. A long thin silencer is more effective than a short fat one of the same volume, because the projectile takes longer to exit.

[Ragnarok]

Had you seen this? http://www.spudfiles.com/spud-cannon-re ... 23931.html

I hadn't. In any case, it's not quite the same as I had in mind; I am naturally reluctant to produce something that has the potential for blocking the barrel.

I don't think this is a factor

It is artificially reducing the volume of the suppressor though, and attempting to remove the effect is not mutually exclusive to other approaches.
I've got some aluminium mesh around, which can be put around the circumference of the cavities. Even if I'm really over-generous about how much is actually needed to absorb the heat fast enough, it shouldn't take up more than about 5% of the suppressor's volume, but it can unlock 25% or more.

Given it doesn't have to hit very much of its potential efficiency to be a net gain, I think it's worth going with.

Baffles that create turbulence behind the projectile that traps as much gas as possible in each chamber.

I looked at the idea, but I'm somewhat torn on the choice between this baffle shape (which isn't trivial to produce) and being able to extend the 22mm pipe through the suppressor (which has its own advantages*), as these aren't hugely compatible

* It's mostly of interest to my dart project - allowing the suppressor to also act more like venting.

Another thing I've discovered is that it's not only volume that counts, but how that volume is distributed.

Ultimately yes, but the 30cm planned length is about as long as I'm up for stomaching. Longer would make the launcher unpleasantly long and front heavy.