Help With Spud Gun

[nomansland]

Hi Guys. I've created a pneumatic spud gun that fires by electric actuation. The chamber and barrel are made of PVC. The valve is a sprinkler valve. The gun actually works, and fires fine. The problem is that I'm running impact tests for a school project and I can't seem to get consistent forces from the gun. I fire the gun to hit deadcenter a force hammer which is to measure the impact force. When I fire at a constant pressure (say 50 psi, compressed air), I can get forces ranging from about 400lbs to 1000lbs. The tip of the barrel is about a foot away from the hammer. The hammer also works fine, I tried hitting it against the ground and got reasonable forces. Therefore, I'm thinking that I'm not getting consistent speeds out of the gun. I would like to achieve 80 mph from the gun, buy when I measured the speed, I would get about (very roughly) 50 mph at 80 psi. Above 80, it seems as though the valve chokes. To measure the speed I used a radar gun.

Does anyone have any suggestions on how I can get the gun to be as accurate as possible in its speed. Also, I really think I need to get a valve that has a larger aperture (whole that the solenoid covers) so I can dump the air faster. Maybe this will get me better results? Also, if you have any recommendations on good devices to measure speed, that would be nice too.

Thanks for the help in advance.

- nomansland

[THUNDERLORD]

Are you firing a projectile in the test or not?
If so what is it?
First guess would be YES, change the pilot to pneumatic actuation. (Modded for blow nozzle) They pilot faster without pneumatic hose btween them too
There's a "sticky" at top of a catagory on how to do that.

BTW, welcome to SF's forum, Looks like you have some interesting stuff to add to the site!
Reminds me of the hurricane debris tester a member here who's professional at that made. (I'll try to find a link)... 8)

EDIT: Are you using wadding or a sabot? If so what?
Also the barrel has some flow restriction through the valve. Maybe try a barrel diameter smaller for less flow restriction(???)

[nomansland]

Thanks for the Welcome. I hope to get a lot of help from this forum (and give help where I can, still a noob).

Yea I forgot to mention, this is a baseball launching setup. Some person came to us and asked us to test out a chest pad they "invented" for little league players. So I am firing a little league baseball.

I also contacted the maker of the force hammer and they recommend using an impact load cell (makes sense..) so I'm going to get that as soon as possible.

Since I am new to some of these terms, what is the pilot? I'm guessing its the device that opens the aperture that lets out the air from the valve? In my setup, I hit a button that completes the electrical connection for the actuator and thus opens the valve to release the air.

I was thinking that the aperture is too small on the sprinkler valve, so I need to get a bigger valve.

I'll check out the sticky though.

Thanks for the help.

[nomansland]

I used some napkins to cover the baseball when loading it in the barrel. I did this to try give the baseball a good seal in the cannon.

[THUNDERLORD]

Might be a good idea to weigh each baseball to make sure, or use the same one each shot (If you aren't already)...
Baseballs probably aren't optimal for seal, maybe try plastic shopping bags for the wadding (check for weight consistency too).
They Might seal better than paper napkins.

Also, that way, You could find something to lube the barrel and shot wad like IDK synthetic motor oil, baby oil etc. I use water on plastic bags sometimes, but seems like something else would work better.

I was going to suggest maybe it's not striking center or the same spot, but it's firing from so close...

Another member will hopefully have more ideas for you...Although a lot are busy for the contest coming up soon... 8)
EDIT: Sounds interesting testing a chest plate for safety!

[jimmy101]

Looks like a fun project.

Like Nomansland and Thunderlord said, you will probably need proper wadding, or even a sabot. The baseballs have fake stitching on them? That'll make sealing kind of hit or miss. You might see if you can find a clear plastic cup that can be used as a sabot. Insert the cup with the open end towards the breech and the closed end to the muzzle so the cup expands when it is under pressure.

If sealing isn't your major problem than a faster valve should help. Or, add a burst disk between the valve and barrel. That will essentially convert a slow valve into a fast one. Search the forum and the SpudWiki for info on burst disks. I'm not sure how consistent burst disks are.

In my experience, spudguns are typically only reproducible to about +/-10% in muzzle velocity (+/-20% in kinetic energy). There are just too many variables to do much better than that.

My personal approach for measuring energy would be to drop the load cell and just use muzzle velocity. Velocity is lot easier to measure than is force. If you know the velocity and mass (weigh each round for each shot) then you can calculate the kinetic energy. In general, time is the always the easiest, and easiest to get accurate, parameter to measure. Measuring force or energy is usually a lot harder and less reliable.

If you've got ~$70 you can buy a shooting Chrony (check Ebay). Or, for perhaps $10 you can build your own with a couple phototransistors and a PC or Laptop. You use the PCs soundcard to log the data, plugging the phototransistors into the MIC input. Here is one example. You can also search this forum for other rigs that work about the same. At the velocities you are interested in, either chrony should be accurate to a percent or so.

Technician (another denizen of this forum) uses a pretty slick inductive pickup that uses a storage o-scope to record the data (it should also work with a PC soundcard as the data logger, or use the storage o-scope with phtotransitors and a battery). Just need to add a strong magnet to the ammo to get a signal.

[cannon monkey]

maybe not getting the same weigh to each baseball with them covered
but a neat thing i learn if u got few inchs of pipe same as the barrel put 3-4 plastic bags from the store like windixie or what ever and force them into small section of the pipe against the floor makes great seal to put behind ( in your case) behide the base ball

[Technician1002]

I am compelled to toss in my 2 cents worth. It may cover some of the low speed and variability between shots.

A sprinkler valve is very quick in opening. It's why they are popular for air cannons. When open, the chamber pressure rapidly drops. Blow by affects this in a random manner with baseballs.

Moving forward a step, the diaphram position is affected by the differential pressure across it.

The pilot side has a rate it can drop pressure. The supply side (chamber) has a rapidly dropping pressure which most likely can exceed that rate. This can cause the valve to re-close several times during the shot.

Recommendations:
Big fast pilot valve. Ditch the factory pilot. It was never designed to open the valve quickly.

Big valve for better flow when open.
Speeds in excess of 100mph are possible. I launch apples at over 600 FPS.

[nomansland]

Hi guys. Thanks for all the inputs. As for measuring kinetic energy, the problem is I would need to measure the energy when the ball leaves, and then again after it rebounces off the pad (The project is to measure force reduction from these chest pads). It would be very difficult to try to measure speed on the bounce back as the ball can come back on an angle.

I am highly inclined to just buy a bigger valve (I guess that also includes getting a bigger pilot valve, I still don't understand what the pilot is...)

The only valve I can find online that is bigger than my sprinkler valve is this:
http://www.spudtech.com/detail.asp?id=39

Does anyone know any other place I can buy a bigger valve?

Thanks.

[MountainousDew]

Mauler Valve!
Its more compact, Less expensive, and Moar Powerful!

[mark.f]

Well, you can also try http://www.launchpotatoes.com/, or you could build a piston valve.

[psycix]

-Pneumatic actuation, more power, possibly more consistent.
-Make sure the projectiles are consistent.
-Make sure your force hammer is accurate. I'd recommend to use a chrony.
Maybe the force hammer can't take the projectile forces properly.

I'm not sure that a bigger valve will help much.
Pilot.. The pilot volume is the volume behind the diaphragm or piston. When emptied, there is still pressure on the other side, pushing the valve open.
The faster you drain the pilot volume, the better, so large pilot valves are recommended. Note that electrical pilot valves have a very small orifice, and are thus very bad pilot valves.
You can "mod" a sprinkler by putting another valve (preferable a blowgun or ball valve) in the pilot area.

[jimmy101]

Hi guys. Thanks for all the inputs. As for measuring kinetic energy, the problem is I would need to measure the energy when the ball leaves, and then again after it rebounces off the pad (The project is to measure force reduction from these chest pads). It would be very difficult to try to measure speed on the bounce back as the ball can come back on an angle.

Ah, that makes it more difficult to get the actual numbers you want. Still, a chrony will tell you if the gun is firing consistently. As things stand now it is hard to tell if the gun is inconsistent, or the force hammer and/or how the ball hits the force hammer.

One more easy way to measure the muzzle velocity would be to just use a microphone. I'm sure the gun gives a respectable bang, which occurs at the instant the round exits the barrel, and the ball makes a noise when it strikes the target. If you put the MIC equidistant from the muzzle and target you won't have to correct for the speed of sound. There's even a freeware software package for doing the data recording and analysis, google for SoftChrono.

[Technician1002]

Hi guys. Thanks for all the inputs. As for measuring kinetic energy, the problem is I would need to measure the energy when the ball leaves, and then again after it rebounces off the pad (The project is to measure force reduction from these chest pads). It would be very difficult to try to measure speed on the bounce back as the ball can come back on an angle.

Ah, that makes it more difficult to get the actual numbers you want. Still, a chrony will tell you if the gun is firing consistently. As things stand now it is hard to tell if the gun is inconsistent, or the force hammer and/or how the ball hits the force hammer.

One more easy way to measure the muzzle velocity would be to just use a microphone. I'm sure the gun gives a respectable bang, which occurs at the instant the round exits the barrel, and the ball makes a noise when it strikes the target. If you put the MIC equidistant from the muzzle and target you won't have to correct for the speed of sound. There's even a freeware software package for doing the data recording and analysis, google for SoftChrono.

This might not be accurate enough for him. The waveform is complex for this high speed event. For example below, what is the sound associated with both below. It is the same shot. Where is the projectile when you "detect a launch".

[jimmy101]

This might not be accurate enough for him. The waveform is complex for this high speed event. For example below, what is the sound associated with both below. It is the same shot. Where is the projectile when you "detect a launch".

A subsonic gun, especially one that is firing at less than 200 feet/sec, makes very little noise until the round exits the barrel, so the muzzle event shouldn't be any problem detecting. (His max velocity needs to be less than about 100 MPH in order to study the affects of a baseball.) The air being expelled out of the barrel in front of the ball won't make nearly as much noise as when the gun "uncorks" when the ball leaves the barrel. When the barrel "uncorks" that is what makes the bang and that should be easy to detect since it'll be a damped sine wave. The muzzle event is simply the first peak in the damped wave form.

If the target is 3' from the barrel then the transit time is about 20mS. The sharp "bang" when the ball hits the target should also be easily identified. Again, that will be a damped sine wave. The event time is the first peak in the wave. The farther the target is form the muzzle the simpler things will be but even at 3' the data should still be interpretable.

If the distance to the muzzle and mic is the same as the target to mic then both events are delayed by (1/SOS)(distance) and the delays cancel each other out.

Granted there'll be echoes but the two largest peaks will be muzzle exit and target impact.

A baseball moving through the air at ~100 MPH makes very little noise compared to the muzzle and impact events. Heck, at 50 MPH it won't make any sound at all. So there are really only two sound generating events, muzzle exit and impact. No other significant sounds are created during the firing event.

Edit:

For example below, what is the sound associated with both below. It is the same shot. Where is the projectile when you "detect a launch".

Launch is the time the round just clears the muzzle plus the (1/SOS)(muzzle to mic distance). You just need to know that it is muzzle exit that creates most of the sound.

If the mic is far from the muzzle then you need to take that into account. Same with the impact event. But both those corrections are trivial since they are just the (distance)(1/SOS)

If the damped sound wave of the firing event (plus perhaps echoes from nearby walls) causes a problem finding the target impact event then put the mic at the target (heck, tape the mic to the target). The launch bang will be much quiter but it occurs when everything is pretty quite so it'll still be easy to detect. If there is echo, or the muzzle bang is still present when the ball hits the target, then the much closer proximity of the target to the mic will make the target signal much stronger. Just need to correct the flight time for how long it takes sound to cover the same distance.

You could of course also use two mics, one at the muzzle and one at the target, but most PCs don't have two mic inputs.