test rig to burst small plastic cylinders 3

[NiZa]

Hello everybody,

In order to describe the behavior of several composite polymer cylinders under burst pressure (explosion) i would like to design a small test rig.

Has anybody already seen something like this?

Are there existing standards for a "burst test" like this?

Thanks for any input!

N

 

[dcasto]

a hydaulic pump and gauge. The typical problem is getting air out. Make sure you have the bottle in a cage or something for when the pieces fly. I think I've seen MythBusters do something similar.

 

[NiZa]

That means you would do it with water rather than air?

 

[jte]

That means you would do it with water rather than air?

Ummm... yes.

If you test with a liquid, you'll get a little squirt of water when the vessel fails. Since liquid is significantly less compressible than air, the stored energy is substantially lower. The same burst with air will be much more violent. Ever seen a balloon pop? Ever seen a water balloon pop? The sounds are quite different even though the rubber is under the same loading.

jt

 

[NiZa]

I agree with you jt, water would be much less violent. But for the real case, the cylinders are also filled with gas, and in case of over pressure that is what will exactly happen, a violent burst. We want to test different materials in order to find one which fails fairly safe. Therefore we want to built this small test rig.

 

[moltenmetal]

If you actually want to test articles of significant size to pneumatic failure, you're asking for a world of hurt if you don't do it properly!

First: what's the volume, and the expected burst pressure range? How much potential energy is there in that stored gas? Start with this calc- it may give you pause.

Second: are you just testing them to failure, or are you inducing failure by introducing a known flaw, stress concentration point or the like? Real articles seldom rupture merely because of applied pressure- we tend to engineer that sort of failure out of the design. But tanks do get dropped, knocked over, have objects dropped on them etc.

Third: how will you get sufficient pressure into them to cause them to rupture rather than merely leaking? Most connection types will leak vigorously long before the container that you're testing will rupture.

Answer a few of these questions for us and maybe we can help you avoid serious injury or death before you get it right.

 

[NiZa]

thanks for your input moltenmetal.

to 1.

for the testing i thought of a small cylinder with an outer diameter of 100mm and a thickness of 2mm. With a yield stress of roughly 80MPa Barlow's eq gives me a pressure of 32bar.
(just used for estimation, will use FE later)

to 2.

we planed to test them to failure without creating a known flaw. that means we might need more pressure.

to 3.

we wanted to use a pipe and seal the ends with some steel caps, using o-rings to seal it. The two caps will be connected through rods and therefore held in place without applying additional compressive stress to the pipe...

General note:

We are extremly cautious and would never do something without an appropriate safety cage.

 

[MJCronin]

Ninza,

I am a little confused..

Are you stating that the mode of failure is significantly different for a "pressurized-air" destructive test than a "pressurized-water" destructive test ?

Are you saying thay you will somehow "learn something from this " ??

I have trouble accepting this....

In over thirty years of design, I have only heard of pressurized-air system testing being accepted for systems that will not tolerate water.

Again, why are you doing this ?

-MJC

 

[NiZa]

Thanks MJCronin,


The test was thought out in order to see the fracture behaviour of fibre reinforced composite tubes under internal pressure.

The reason why i wanted to use air is the the high expansion of the air when decompressed. The fracture should be more rapid, and therefore more severe. Do you agree?
(more small parts, rather than just one big crack) But clearly i have no experience on that.

The test is just a baseline test in order to asses the material.

We have a reference material and would like to see performance of new materials in "close to live" conditions.

Thanks again!

 

[moltenmetal]

MJCronin: I think NiZa wants to see if his pipe disintegrates into high-velocity schrapnel or simply tears open and leaks. The latter will indicate that the reinforcement system is doing what it should.

If you test the assembly under pure hydrostatic conditions, it is likely to merely open and leak regardless. The test is VASTLY safer but not as meaningful to NiZa's design intent as I understand it. A proper forensic/failure investigation engineer could give you meaningful advice in relation to whether or not such tests could give you sufficient evidence of the expected failure properties under pneumatic conditions without the need for pneumatic testing- I can offer you no informed opinion in this regard. Certainly that's more desirable IF it can give you the information you need.

I'd suggest that you do hydrostatic testing FIRST, regardless, to establish the expected burst pressure range. Once you know that, you can design a test rig for pneumatic conditions- with great caution. From what I've heard of your plan, I still think you'll have difficulty loading this thing to failure without significant leakage at the end connections long before rupture occurs, but maybe I'm missing something.

First, given the size and burst pressures of your test specimen, you'd be far safer doing these tests outdoors. Second, consider that your "cage" will not only need to be strong, but also must have enough open area to ensure that it is not ruptured by the shock wave generated by the bursting cylinder. Third, make sure the entire thing is remotely operated and monitored. Don't put people's lives at stake for the sake of a test. And fourth, don't forget about the energy stored in all the other components of the test system, which will be released during the failure. Keep the volumes of stored gas as small as possible, and the line connecting the test specimen and the test rig of a small bore to limit the release rate from the test apparatus, and well restrained so it doesn't whip around. No point in making this any more dangerous than it has to be.

 

[NiZa]

Moltenmetal,

thank you very much for your input. The situation is exaclty like you said and i really apprechiate your much better explanation.

As i am not so sure myself how much difference a hydrostatic test compared to a pneumatic test would make i also decided to start of with a hydrostatic experiment.

From the point of difficulty to load this cylinder to failure i also thought of changing a bit the original idea.
Maybe an even simpler geometrie will help.

I thought of a disk clamped between two cylinders where the top cylinder will be pressurized with water. Until the disk fails.

Is there an existing standard for a hydrostatic burst pressure test for composites? Are usually only pipes tested, or are there also other geometries?

 

[jbwick]

NiZa a good source for testing information would be a local fire extinguisher repair facility. They pressure test the extinguishers before filling them. Whoever provides that test equipment might even have a packaged solution for you.

Burst tests do exist for cylindrical objects if you live in the US check with the nearest DOT office or on the web. There should be something there pertaining to carbonfiber wrapped aluminum bottles.

Another small point once you have a safe test cage you should have no problems making an air tight seal at 32 bar.

 

[EdStainless]

One method if you have to test with gas is to place the units in a water tank. That way the fluid will absorb a lot of the energy at rupture.
Unless you are concerned about the inner liner failing and gas leaking into the composite and causing a failure I would stick with hydro.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[Compositepro]

NiZa has valid reasons for testing with a gas. How the vessel depressurizes when it fails is a significant safety issue and key factor in whether a material is suitable for making pressure vessels. There is a big difference between developing a crack or tear and venting slowly or bursting into fragments. If there are fragments the size is important to safety. My initial reaction was to recommend hydrotesting but now it is clear that that will not provide the information NiZa is looking for.

 

[NiZa]

Compositepro,

thanks for your input, i agree with you it would be much better to test with gas, but due to safety reasons and simplicity at the moment i still tend to switch to water.

I also changed the initial geometry of the specimens from a cylinder into a disk shape.

I will clamp the disk between two cylinders, and pressurize one cylinder until burst.

At the moment i am calculating the dimensions of the cylinder and the flanges. According to the standard the flanges for 100bar have impressive dimensions.

Thanks again!