Need a simpler solution to piston valve problem

[McThunderThighs]

Hey all! A bunch of guys and I got an assignment to build a potato gun (formally a tennis ball gun) and since we have a couple months and it is classifiable as homework, we are deciding to go all out.

That being said, here was our initial design:

Two separate reservoirs are attached to two separate piston valves, which feed into the barrel (which has it's own modifications and fancyness's but these are negligible). Our goal was to have a couple of choices, fire one reservoir at a time, allowing for a faster firing time, or fire both at the same time for maximum power. Sounds cool right (and way over the top!!).

We were originally testing the valves with 3/4in ball valves, which work great! But there lies the question of how to operate both valves simultaneously, and not be right behind several thousands of newtons and risk something going wrong (in other words, operating both ball valves at once is out of the question).

So we said hey! Let's replace the ball valves with solenoid sprinkler valves! However, using 3/4in sprinkler valves, we can't let the air out quickly enough to release the piston.

THE PROBLEM===========================

We want to simultaneously release two very large (4in ID) piston valves at once!

Solution 1: Use sprinkler valves = failed, not enough airflow.

We came up with a new solution. Pilot both piston valves with another, smaller piston valve, that is piloted by a sprinkler valve.

This adds some ridiculous complications, so can anyone think of something else that might work?

Thanks in advance. I hope to let you see some shots of the triple piloted double potato cannon some day!

Hasta!
-MTT

[Launch-Tek]

Sounds like you equalization holes in your pistons are too big. 3/4" sprinkler valves should be enough to fire them. 1" would be better. I would suggest twin modified sprinkler valves from www.spudtech.com.

[brplatz]

your cannon seems much like these guys one.

See if any of that applies to you

[frankrede]

Mod the sprinkler valves.

[jackssmirkingrevenge]

Any pictures of your setup? Sounds like your pistons could use a little optimisation in the efficiency department, tighter fit and smaller pilot volume. What's the ratio between barrel and chamber diameter?

[McThunderThighs]

@ Launch-Tek & frankrede: Both valves I am using have been modified.

@jack: Yes I agree there. The pistons fit VERY tightly (they were sanded to just barely fit when all gooped up), but there initially was a HUGE pilot volume (the 4in piston would move back about 6in). To see if the volume was the problem, we filled the back with a tostidos can filled with water (and made sure it didn't affect the piston's motion) and tried. In this case the piston did come back earlier, but we had to let out quite a bit of air before it did. (Working with the sprinkler valve here)

The ratio between barrel and chamber is 2.5x4, in the valves it is 2x4.

I just want it work like it did with the ball valves dang it!! (Nervous meltdown ensues).

Thanks!
-MTT

PS I can put up some pictures in a couple of hours!

[psycix]

Solution for best performance:
Build smaller piston valves to pilot the larger ones. Focus the build on fast-acting, lightweight pistons, as opening time is important to a pilot valve.

Note:
A very tight fitting piston is still not airtight, and on 4" diameter, that means theres a HUGE gap in total.
Use o-rings to make them seal.

[McThunderThighs]

Ok.

I have succeeded in this: A sprinkler valve pilots a 2" piston valve that pilots the 4" piston valve.

Works every time. Now time to build a second 2" piston and put the system together!

@psycix: I don't want the piston to seal, otherwise I can't fill the reservoir.

The other thing I would have liked was this: Pilot both valves off of one smaller piston. Can't be done, I would need 2 compressors to supply the air volume in order to simultaneously seal the valves.

[frankrede]

Ok.


@psycix: I don't want the piston to seal, otherwise I can't fill the reservoir.

You need a better piston. Even with a tight fitting and well lubricated piston air will leak past.

[Velocity]

For a 4" piston, you should aim to have it moving back only 1". If you have lathes and such to meet closer tolerances, you can reduce the pilot volume even more to have faster actuation time (its generally diameter/4 for pilot volume)

[man_o_brass]

You wouldn't need two compressors if you put a cutoff (ball valve) on the pilot of main valve. That would isolate the pilot valve from said chamber and allow them to be filled one at a time. The only problem might be a sudden pressure fluctuation when you opened the valve, but that could be remedied with a small hose run between chambers (also with a valve) to equalize pressure before the pilot cutoff was opened.

Kinda vague, did that make any sense?

[McThunderThighs]

Ok!! We were fully operational!! For about 3 seconds...

Then one of the piston seals broke and we ran on one cylinder for the rest of the night (just to keep the neighbors awake).

Pictures later, right now, my rebuttals:

@frank: I don't think you are understanding. I am using a coaxial(x) valve, the air to push the piston forward and to fill the reservoir comes from behind the piston, seals the piston with a gasket against the barrel (or the stuff to the barrel) and then goes around the piston to fill the rest of the reservoir. If the piston sealed to the sides of the valve, then the reservoir would never fill and then what's the point?

@Velocity: You are probably right. I don't move the 2" piston back more than a half inch, and it works great. The 4" still comes back 6" though, it's too late to start ripping stuff apart and rebuilding it.

@man_o_brass: I think I am already doing that. Check out the pictures and look for the mini-flex pvc tubing running around (its attached to the orange fill hose). The pilot behind the pilot valve and the big valve are equalized with this to allow for simultaneous filling. Works great!

Now for some picks, lets start with a glory shot!!


Here is the valve assembly from the back. You can see the exhaust for the 2" pilot piston valves.


The valves from the side:


And finally the damage that prevented a full power test. The rubber foam I was using as padding for the gasket allowed air though, with eventually caused the gasket to fail:

Yes I know it is blurry, it was dark and hard for the camera to focus.

A few more pictures can be seen in my flickr:
http://www.flickr.com/photos/james_hale

Tell me what you think of my overly complicated solution to a very simple problem!!

[MrCrowley]

You're using DWV, which is not pressure rated. I'm sorry to say but this cannon is very unsafe and should not be used again because of the forces involved with the pressures you are using and the piston valves, which put a lot of stress on the joints when fired.

You might want to read this: http://www.spudfiles.com/spud_wiki/inde ... _Rated_PVC

[frankrede]

@frank: I don't think you are understanding. I am using a coaxial(x) valve, the air to push the piston forward and to fill the reservoir comes from behind the piston, seals the piston with a gasket against the barrel (or the stuff to the barrel) and then goes around the piston to fill the rest of the reservoir. If the piston sealed to the sides of the valve, then the reservoir would never fill and then what's the point?

No no no, I understand completely, I'm just saying that you need an almost airtight fit around the piston.
Air will leak past.
Your cannon isn't a coaxial, its a barrel sealing piston valve in a T.

[jackssmirkingrevenge]

For a 4" piston, you should aim to have it moving back only 1". If you have lathes and such to meet closer tolerances, you can reduce the pilot volume even more to have faster actuation time (its generally diameter/4 for pilot volume)

Worth clarifying that the piston needs only to move back around 25% (maximum theoretical efficiency) to 50% (allowing a bit of tolerance for piston bounce) of the barrel/seat diameter for maximum efficiency.