SpudFiles

Hi All, my first Pneumatic launcher is well under way (a 3" Piston cannon ) but after browsing the forums for a while I discovered a few design concepts that I wonder if anyone has tested

the first of which is how far the barrel should be extended into the piston housing. I drew a simple diagram that compares how two different barrels are positioned, one going farther into the housing while the other does not pass the bottom socket of the tee

based on my knowledge, it seems like design 1 (the shorter one) would get better airflow than the other design. does anyone have experience that would confirm that?

It also seems that if the first design was used, the piston would have to travel farther to open fully, which leads to my second question about piston travel distance. I've heard of the d/4 rule for coaxials and the d/2 rule for barrel sealers (which is what I have), but I'm not sure which diameter it should be based off of (the port hole or the piston face?) . I've got 2" sch 40 PVC for my barrel, which makes the diameter of the port hole just over 2" and my piston face is just under 3". Also, would the barrel's position affect how much travel distance the piston should have?

Hopefully I haven't brought up questions that have already been answered, but I've looked around quite a bit and found nothing decisive

thanks

Of the two, I'd do a compromise. Have the barrel protrude exactly halfway I'm that example.

I remember this being discussed before.

The 1/4 D rule uses the ID of the barrel for the Diameter. It assumes a coaxial design where air can enter equally from all directions. In a T, most of the airflow arrives from one side, so the 1/4 D rule is undersized and the 1/2 D rule is used instead as flow enters mostly from 180 of the 360 opening. This is not based on barrel or chamber sealer. It is based on coaxial with the chamber surrounding the valve, or built into a T with most of the flow arriving from one side.

My opinion for piston seal placement. What matters most is the unrestricted flow into the barrel when the valve is fully OPEN. This rules out the gap off to the back end of the T as the flow has to enter the T to the rear where it is fully surrounded by a Wall, then make a 180 U turn to enter the barrel over a narrow edge, the valve seat.

With the closed valve oriented with the seal oriented exactly in the center of the T is a batter choice, but still not optimum as when the valve is fully open, the gap between the barrel and piston is entirely off center to the flow entering from the barrel, again causing a flow restriction.

Your Orange location is about right as you want the piston to move at least 1/2 the barrel ID Diameter away from the breech of the barrel to be considered fully open. When you are fully open, you want the gap to be centered in the widest location in the T for best flow from the chamber into the barrel.

The basis on the 1/4 D rule for coaxials is so the cross section area for flow entering the barrel is exactly the same cross sectional area produced by the gap between the face of the piston and the face of the barrel valve seat.

I personally don't like that rule as it places exactly two flow restriction gaps in series one right after the other, the piston to barrel gap and the barrel ID area. I prefer to open the valve beyond this minimum gap so the first flow restriction becomes much less. I generally like the piston travel to be about double the minimum, so a coaxial will open at least 1/2 the barrel ID and in a T move a full barrel ID distance. To do this in a T this places the end of the barrel exactly at the edge of the pipe from the chamber in the T and the piston when open moves to the other end of the pipe from the chamber into the T, so viewed from the chamber the gap has neither barrel or piston protruding into the T so the T in flow closely approximates an elbow to the flow with no restrictions.

If you add the pipe from the chamber at the same diameter as the barrel, you can see the orange location is about right. Using a larger pipe into the piston housing is even better as you have it drawn.

I know I have rambled some but I hope this answer makes sense.

I like coaxial designs as a shorter lighter piston can be used for faster operation is it is fully open in about 1/2 the distance of a T design.

Thanks technician. I'd thought that 1" piston travel seemed a bit restrictive. I just had seen a post where Jackssmirkingrevenge advised someone to reduce travel to d/2 to reduce pilot volume. I want to find a good balance between air flow and pilot volume. My current plan is to put the barrel about where the orange one is, and give the Piston 2" of travel. I should have gluing underway tomorrow

If you have a 2" barrel in a 3" tee, there is a greater cross sectional area for the gap between barrel and chamber than inside the barrel so no real flow obstruction.

In that regard, I'm a bigger fan of your design 2. It also allows your piston to be smaller, therefore lighter, meaning it will open faster and put less stress on your bumper.

After measuring the speed of piston valves, I would bet the flow is more improtant unless you are using really short barrels and very lightweight projectiles. The difference in performance in 1.5 vs 3 mS to open is less than the loss adding a 90 degree elbow on a launcher with a >15 mS barrel eject travel time.

If you want a demonstration, blow as hard as you can through a drinking straw. Now try again with a 3/4 inch copper pipe the same length. Restrictions to flow hurt performance in high speed air flow.

Technician1002 wrote:After measuring the speed of piston valves, I would bet the flow is more important unless you are using really short barrels and very lightweight projectiles.

Very true, but it depends heavily on your launcher as you say, projectile weight, chamber and barrel proportions etc.

Did some quick modelling with GGDT's default generic spud launcher firing a 100 gram potato at 100 psi.

With a full 1.59" diameter seat, I got the following for opening time vs energy:

20 ms - 510 ft lbs

15 ms - 540 ft lbs

10 ms - 560 ft lbs

5 ms - 568 ft lbs

0 ms - 568 ft lbs

Keeping opening time at 5 ms, but reducing the diameter of the seat, I got the following for diameter vs energy:

1.59" - 568 ft lbs

1.5" - 550 ft lbs

1.2" - 457 ft lbs

1.0" - 368 ft lbs

Clearly both parameters are important, and ideally you have a high flow and fast valve. If you have to compromise ( which in this case given the suggested dimensions, you don't ) then model the individual launcher and see what works out best.

Sounds like it would be a good Idea for me to get GGDT and mess around with it some. Unfortunately I'll have to wait until my dad gets home from work because my computer user has no Administrative Privileges to download things

For now, I'll make it my goal to leave flow unrestricted and take other measures to speed opening time (remove other dead space, make piston lighter...)

Thanks for all the input

Well, I got GGDT, and have been fiddling with it for a while. It seems like with the piston I have now, opening time will be about 5 ms. I was also experimenting with other aspects of the cannon, and the strange thing is that when I increase the dead volume (the space after the valve and before the projectile, correct?) the muzzle velocity and energy increase. Shouldn't that decrease the power/ efficiency of the cannon? If someone could explain why, that would be very helpful

Could it be due to heat from the compression of the air behind the projectile before the projectile starts moving?

Actually, depending on th piston OD vs seat diameter, a larger pilot volume is an advantage..

I'll explain. If your vent trigger is relatively small such as a blowgun, and the pilot volume is low, as the piston opens, it compresses the gas in the pilot stopping the piston from opening all the way.. until the pilot can finish venting the volume.

With a really light pistion EG Diaphragm or sprinkler valve, this is known as a honking valve.

With a larger pilot, iit may take longer to vent the volume to a low enough pressure to unseat the valve, but when it unseats and chamber pressure hits the face of the piston, it moves further back quicker, resulting in a higher energy launch.

This knowledge was the inspiration for the design of the QDV. Pilot is vented all the way to 0 PSI before the piston moves, so when it does move by mechanical trigger, it blasts itself all the way open for sub 3ms opening times to full open.

This only works where the pilot ID is only slightly larger than the valve seat diameter so the pilot pressure has to bleed to a low pressure before the piston valve cracks open.

In a QDV the valve seat diameter and the pilot diameter are the same.

Some cannons to compensate for a small pilot and large diameter piston simply use a huge fast pilot valve such as a QEV or sprinkler valve to quickly vent the pilot.

Alanstone wrote:Could it be due to heat from the compression of the air behind the projectile before the projectile starts moving?

This was a theory expressed by the creator of GGDT himself a few years ago: http://www.spudfiles.com/spudtech_archi ... =5&t=13463

Some cannons to compensate for a small pilot and large diameter piston simply use a huge fast pilot valve such as a QEV or sprinkler valve to quickly vent the pilot.

As they well should

well, these explanations seem to make sense to me. Hopefully the GGDT results will prove accurate (although I'll need to find a chrony somewhere).
My plan is to pilot the piston valve with a 1" sprinkler valve, although I could use a 3/4 qev that I'm working on

Thanks for the help. My cannon should be up on display tomorrow

I appreciate it too. I don't apply a whole lot of science to my projects; besides things that are crucial, I generally eyeball most things and just do what "Looks" right. Kinda ironic considering how much of a critical thinker I typically am.

Blitz wrote:I don't apply a whole lot of science to my projects; besides things that are crucial, I generally eyeball most things and just do what "Looks" right. Kinda ironic considering how much of a critical thinker I typically am.

You are just using a little common sense that has been instilled from previous failures. It is called learning and becomes common sense.