SpudFiles

Hey guys,

I have been working on a cannon for a few weeks and now its crunch time to get it working. Its for a university project, and I would really appreciate some guidance. Here are the requirements:

Barrel ID needs to be greater then 155mm

Needs to eject a artillery shell shaped projectile (155mm) at an exit velocity of greater then 100 m/s. The projectile weights 4.5kg

Here is my current plan:

I have sourced a pressure rated pipe with a 159mm ID that is 6 meters long. I plan on cutting this down to 4m for logistics.

In line with the pipe will go a pressure rated flange with alluminum foil in between the two pieces. Inside the flange is a linear push solenoid glued to a side, right next to the foil. The seat diameter will be 159mm.

The chamber is a 140L LPG tank (actually two welded together and certified by a pressure vessel dude). This will be charged at 120 psi via a air compressor.

This is all in a line, supported by a welded frame.

My question is am I in the right direction? Would you be using pneumatic design with that style valve? I have run it all though GGDT with good results.

Really appreciate your help, about to drop a load of cash on this (nearly 2k AUD) and I really would love some general confirmation from some experts.

Regards,

Stuart

With that mass of projectile a spring loaded butterfly valve may be an option. Will the solenoid have enough force to pierce the foil? For less space than a solenoid, take an idea from your bathroom sink and look at the pop up drain stopper. This lever/ball seal can remove most of the cross section of a trigger device inside the barrel for better flow.
A double burst disk might be another option to consider as a trigger device.
Coaxial configurations work very well.
If you have the machining ability, a velocity cone between the tank and burst disk can reduce the drag of the edges between the tank and pipe. A reducing cone from about 180mm to 160mm can improve flow. GGDT is not set up to show this. To allow for some gas blow by past the projectile, it is wise to engineer above the bare minimum to compensate for losses that would be unaccounted for.

An engineering challange I took part in reached it's goal at 35PSI when we had a 100PSI limit. It is better to exceed the goal than to miss it.

My original design was a coaxial dual valve but I went away from that for simplicity sake, and pressure rated T bends and 90 degree bends are very expensive.

Thanks so much for the tips about the velocity cone, great idea, will get that made up.

In my head I imagine finding the right number of layers of foil to hold in the pressure, then have the solenoid with a sharp tip just rupturing the foil. I have looked into butterfly valves and the super valve but wouldn't the foil setup result in a faster release (rupture) with no flow restriction?

Stu

In GGDT do a graph of your projectile motion in the barrel with the burst disk. How far does it move in the first 1, 2, 3ms. Compare that to how fast a spring loaded butterfly valve can open. A valve 1/2 open with a projectile mostly at rest next to it will have little pressure loss on it. By the time the projectile has appreciable velocity, the valve can be fully open to provide the high flow required. Your projectile is 4.5kg. A lighter projectile can take advantage of a faster valve.

As mentioned earlier, the solenoid can be located outside the barrel to remove it's profile from the high speed airstream and have just the piercing knife and rod in the barrel.

As Tech has suggested, a butterfly valve may be a better option. If you plan on firing this thing more than half a dozen times, a proper valve is really beneficial. It can be annoying playing around with burst disks to get them to seal and set up correctly.

4" butterfly valves are common and could be had for $50AUD without much trouble but you'll need a 6" valve which are much harder to find for bargain prices on auction websites:
1
2

Hey guys,

I completely agree about the multiple run benefits of the butterfly valves. However as we will only be firing once or twice per day I was edging towards the burst disc.

Whats your thoughts on the C/B ratio for this design? Should I be edging towards 4:1?

I never complicate myself with C:B ratios, I run various simulations through GGDT until I reach a design that is reasonably close to a size I want while providing the required performance.

Keep decreasing chamber diameter and length and barrel length until you still have good performance with an efficient design. Efficient in terms of materials used and compressed air.

GGDT will tell you the pressure drop in the barrel which will give you a good idea whether your chamber is too large or not most of the time. For example, a pressure drop of only 10PSI would indicate your chamber is probably larger than it needs to be. If your projectile is long gone out of the barrel and there's still 90PSI in the chamber, all that air is more or less wasted. Instead of having a large chamber, a slightly longer barrel can make all the difference.

This goes back to what Tech was saying about reaching a design goal with using only 35PSI instead of 120PSI. If you're using 120PSI, your chamber can be much smaller than a cannon that has the same performance using only 35PSI.

stukearney wrote:Needs to eject a artillery shell shaped projectile (155mm) at an exit velocity of greater then 100 m/s. The projectile weights 4.5kg

Without rifling you do realise that it's just going to tumble as soon as it leaves me muzzle, right?

In line with the pipe will go a pressure rated flange with alluminum foil in between the two pieces. Inside the flange is a linear push solenoid glued to a side, right next to the foil. The seat diameter will be 159mm.

I wouldn't go this route, you will have incomplete disk rupture and the solenoid will create turbulence that will reduce power.

I'm assuming this is not something where rapid fire is important, personally I would go with simply having the burst disk fitted, and pressurising the tank until it bursts. Simple, clean and if you're pressurising at known rate with consistent burst disk material, the burst point can be predicted with reasonable accuracy to avoid surprises.

Another option (used by NASA, no less...) is to use a mylar burst disk (architect's tracing paper is a good source) and sandwich a loop of nichrome wire between et voila, electronic ignition. Here's an example I had built on a smaller scale.

Also don't forget the triggered burst disk option.

Brilliant! Thanks for those links. Love the idea of the mylar and wire.

Without rifling you do realise that it's just going to tumble as soon as it leaves me muzzle, right?

Yep, we have the COG far forward on the projectile (relative to the COP) and it has spring loaded deployable fins.

stukearney wrote:Yep, we have the COG far forward on the projectile (relative to the COP) and it has spring loaded deployable fins.

Sweet! Excalibur style?

Hahaha. Yep. If your interested in what we are doing..Basically its more like the MIT WASP (independent origination of the idea though). We are sticking a whole UAV in the shell and getting it to deploy out of the cannon. Thats the main reason I don't want it spinning, the parachute will get all tangled up.

stukearney wrote:Basically its more like the MIT WASP (independent origination of the idea though). We are sticking a whole UAV in the shell and getting it to deploy out of the cannon. Thats the main reason I don't want it spinning, the parachute will get all tangled up.

Ah, nice! Good luck with it

Another gun launched UAV that might be of interest if you haven't seen it yet is the Oto Melara Horus.

OMG
cool project


While I find it awesome and would like to see an amateur tube launched UAV... I have to ask:
is there any reason for tube launching it rather than some more typical method ?

normally ppl use something like these:
[youtube][/youtube]

[youtube][/youtube]

As you can probably know this method makes stuff a lot easier.... both if you consider the UAV design and

logistics

not to mention VTOL capable UAVs, I find this method particularly interesting due to its simplicity:
[youtube][/youtube]I imagine that tube launching already puts some serious constrains on the design. So I guess you might as well add a more powerful engine and go for VTOL

Hey mate,

We are eventually scaling up to full sized artillery. The idea is to get the UAV 80km away near instantly. The spudgun is just for proof of concept

I imagine that tube launching already puts some serious constrains on the design.

You have NO idea how right you are. It's been such an epic 2 years getting the UAV working when its gotta fit inside of a 155mm cylinder. Not to mention the parachute system, avionics, ground station etc etc.

so basically something like this -> clicky

hmm why 155mm ? That suggests a howitzer but that doesn't make any sense