WTF is wrong with me?

[Technician1002]

I've only skimmed this thread, but you should find the links on this page to be extremely helpful D_Hall: http://trettel.org/bags/other-models.html

You might find this to be the most interesting (and may even have a printed copy already available to you): http://www.dtic.mil/srch/doc?collection=t3&id=AD0475660

The high muzzle velocities will require more complicated modeling much like what I plan for BAGS 2 and what Rag has done with Apocalypse. The simplest way is to assume velocity, density, pressure, etc. varies down the length of the barrel (this is 1D).

There are a number of ways to solve an unsteady compressible 1D problem... the books I have listed primarily use the method of characteristics, as does Rag's Apocalypse. I'm been thinking about going that route myself, but I really don't have the patience for it. Ragnarok must have significantly more patience than me because the MoC is a PAIN IN THE ASS to work with. I'm sure that when it works well, it works well, but it's another layer of complexity in an already complex problem. I can give you some good references in addition to what I already have listed if you want to go this route.

For that reason what'll be released will likely use what's called a finite difference of finite volume scheme... which aren't too difficult to implement but you run into the problem of the scheme being a "black box" in that you don't really know how it works or when it's working well. So I have much highly mathematical reading ahead of me. The end result surely will be worthwhile.

@ Jack, The 3 gallon launcher would consume a huge amout of gas for a 200 PSI shot. GGDT does make wonderful predictions for it. The 7 gallon at 200 PSI would perform even better but use about 12 cubic feet of gas per shot. It may get the ball supersonic, but doesn't answer the question, can I do it with just air?

The links are great but some of the links on the pages have changed. I have found the NASA pages with some of their simulations. They are not as easy to use or as intuitive as GGDT.

The modeling is very complex, but to assist I've been modeling the acceleration of a projectile piston in GGDT to give an acceleration curve, then reversing the process to find the chamber size necessary to reverse the process lossless in say 1/4 the barrel length. This gives a higher pressure than I started with in a smaller volume. I then use these values to see the "temperature drop if it was launching and look to that as a temperature rise instead.

Then I take the hot gas temperature with the small virtual chamber and high pressure and use that to model the ejection of the small projectile out the small barrel with the high pressure high temperature "light gas" in the chamber. Next projectile weight diameter and such is adjusted to give the exact same time to launch as the first piston takes to stop in the compression end.

I know there will be massive losses due to the temperature changes, friction, piston blow by, recoil causing incomplete energy transfer etc etc, but the preliminary numbers are looking real good.

Taking 70 degeree air at 200 PSI and dumping it on a 3 inch piston gives it energy, which then comes to a sudden deceleration in 1/4 the distance compressing air to over 800 PSI which heats it a lot, which transfers energy to the projectile 1/4 the mass of the piston as a light gas at theoretically over 3X the speed.

A 500 FPS piston into the light gas tube in theory can launch the golf ball at well over 1500 FPS. My burning desire is to be the first here to try it.

Since I've launched apples at over 700 FPS on only 100 PSI the required energy on a relatively massive 3 inch piston at 200 PSI is easily done. Imagine a golf ball with an 800 PSI light gas behind it for 10 feet. This is the type of figures I am getting.

Keeping it from blowing up will be the trick.

[btrettel]

Two of the links were broken. The problem seems to be that the NASA technical reports server is down. Thanks for letting me know.

As these reports are in the public domain, I've reposted them:

http://trettel.org/bags/New%20Higher-Or ... 20Guns.pdf
http://trettel.org/bags/The%20Light-Gas ... uncher.pdf

I'm not advocating using these simulations (I am not aware of the code for any of the light gas gun simulations I have listed being released), rather, I'm advocating understanding how these simulations work. Reading about this simulations has been enlightening for me, and I suspect it would be too for D_Hall, especially if he's writing a new one.

GGDT should be reasonably accurate for relatively low velocities with the procedure you've detailed Technician. But the point of light gas guns is to go supersonic. GGDT's method is inadequate for this. None of the technical papers about light gas guns I've linked to use a similar approach.

[Technician1002]

GGDT should be reasonably accurate for relatively low velocities with the procedure you've detailed Technician. But the point of light gas guns is to go supersonic. GGDT's method is inadequate for this. None of the technical papers about light gas guns I've linked to use a similar approach.

There is little modeling on using that slow of an initial piston speed in a light gas gun, which is a big part of the great unknown. Some flow models work both ways, some don't Due to the way temperature scales from absolute zero, I'm sure the temperature modeling is way off base for the pressure rise and temperature rise with it. There are a few other vastly unknown dynamics in this. Taking a wild guess at some of this stuff and trying to account for the energy transfer, losses, and thus a final energy of launch is exciting. A build and testing is the only way I know of to produce answers.

The only thing known is it will generate high pressure and high temperatures, possibly high enough for auto ignition of lamp oil. In reading literature on compression testing diesel engines pressures in the range of up to 600 PSI is normal.

http://www.tpub.com/content/constructio ... 64_105.htm

This looks able to reach a diesel compression stroke pressures and thus temperatures.

[btrettel]

I'm confused about precisely what you are talking about. You seem to describe what most call adiabatic heating. There are some good simple equations to describe quasi-static adiabatic processes that you might find good for an approximation.

[Technician1002]

I'm confused about precisely what you are talking about. You seem to describe what most call adiabatic heating. There are some good simple equations to describe quasi-static adiabatic processes that you might find good for an approximation.

In small spaces with high temperature variations between the gas and the chamber wall, some of the "work done" is energy lost from the gas to heat the tubing, projectile, etc. This is not returned to the gas as it moves on down the barrel to transfer the energy to the projectile. This is difficult to model in small diameters with reasonable precision.

The other model is the complex model of the piston slowing and compressing. At the same time the projectile is accelerating preventing full compression, so this moving size changing chamber of gas between the piston and projectile is difficult to model the final stopping position of the piston, the amount of piston recoil (energy not transferred) loss to turbulence in the transition zone, and heat loss.

Ideally the piston will advance and compress and come to a stop at the same rate the projectile accelerates and exits leaving the piston stopped at no pressure for recoil, but this is unrealistic and would require too low of a system pressure on the projectile for high energy transfer efficiency, so reality will be a partial energy transfer with the projectile exiting with significant pressure still in the barrel which will both exit at high loss and return energy to the piston as kinetic energy. The total dynamic is complex.

[btrettel]

Adiabatic means no heat transfer. The equations in the link should be treated as simple approximations (as I said) or a "best case" scenario. You are right that this is one reason to think these are approximate best, but...

Heat transfer isn't very substantial and is neglected in most simulations of light gas guns. The time scale is far too small. It's still mentioned in a few of those links and is not particularly difficult to model as far as I know.

The movement of the pistons and projectiles isn't difficult either. It comes down to force balances, Newton's laws, and numerical methods for ODEs.

Note that when I say model here, I mean with a computer simulation. If you're looking for analytical (i.e. plug numbers into an equation) solution, you won't find any accurate ones.

You might benefit from reading the old PDF I wrote about how BAGS works, but as I recall you don't know calculus so I doubt you could understand it all. Essentially, everything comes down to calculating changes in everything for small, discrete time steps.

Edit: I reread this and I come across as patronizing. Let me make it clear that this is not my intention. This isn't an easy task. If it were I'd have my own simulation like this. Go figure.

[Brian the brain]

Hmm...One of my next projects ( depending on time and money) would also be a dieseling piston gun...


these things travel in waves I tell ya...

[benstern]

I am pleased by this announcement.

[rikukiakuchiki777]

I've been reading up on 2 stage light gas guns, and they are an intriguing topic. I also came across this . It's an all-PVC 2 stage light gas gun!

As expected, it underwent catastrophic failure partly because it was compressing oxygen and hydrogen, and mostly because it used gunpowder as the piston's propellant... . But still, it makes me wonder - is something like that theoretically possible. As jack said, could a simple propane 1x mix or pneumatic means be used to propel the piston, and could this be done in PVC without destroying the launcher?

Apart from that, I cant see any reason not to create a 2LGGDT, if your up for it give it your best shot .