Hydraulic testing air tubes to 30,000 psi

[Machinist]

The thought of producing my own pressure tubes makes me cringe.  I worked in a hydraulic shop for 3 years and made lots of cylinders, pistons etc. that held 3,000 psi, BUT... the cylinders often had like a 1/4" or even 3/8" wall.

With the lightweight gun stuff, I'd be afraid I got a bad batch of material, or the material was innocently mislabeled, or some other problem that causes a kaboom. Material really needs to be certified no? We rest our faces against these pressurized tubes!

As far as volume, when we tested large cylinders, we had a mack diesel engine driving a pretty big pump.  Maybe you would like a small pump driven by a 1/2 hp electric motor? Could get you some more volume. I think you could put something together pretty cheap if you have a free electric motor and a tank lying around.. No more pumping. Just a thought.

These guys http://www.surpluscenter.com/Hydraulics/Hydraulic-Pumps/Gear-Pumps/ have a pretty good selection of hydraulic stuff.

Thanks for saving us from ourselves Lloyd.

Steve

[lloyd-ss]

Steve, if the thought of making your own cylinders makes you cringe, it just shows that you have a healthy understanding of what you are making.   Understanding the necessary material strength calculations is imperative, and there are at least 3 different ones that must be done.  I buy my tube material from Online Metals and unless the pieces are shortys, they are always labeled, supplied with certs, and the tubes themselves have the certification stamps along their length.  I worked just about forever (so it seems) in manufacturing facilities that produced military equipment, so I understand the importance of documentation, certifications, traceability, etc, etc, LOL. 
Instead of the pumping, how about an easy and versatile way to purge all the air out of the system?  Laying the assembly on its side is the easiest way to handle it, but I have about 4 different bleed ports to deal with to get most of the air out.  And you can tell in the video by the way the tube popped, that I didn't have 100% of the air purged. Maybe use the electric pump with a continuous purging loop to get the air out, and then seal that off and use the hand pump for the final pressurization.  When it is free of air, it doesn't take many pumps to raise the pressure all the way up.
Lloyd
P.S.  Thanks for the link.  Looks like neat stuff!

[Sfttailrdr46]

Steve, if the thought of making your own cylinders makes you cringe, it just shows that you have a healthy understanding of what you are making.   Understanding the necessary material strength calculations is imperative, and there are at least 3 different ones that must be done.  I buy my tube material from Online Metals and unless the pieces are shortys, they are always labeled, supplied with certs, and the tubes themselves have the certification stamps along their length.  I worked just about forever (so it seems) in manufacturing facilities that produced military equipment, so I understand the importance of documentation, certifications, traceability, etc, etc, LOL. 
Instead of the pumping, how about an easy and versatile way to purge all the air out of the system?  Laying the assembly on its side is the easiest way to handle it, but I have about 4 different bleed ports to deal with to get most of the air out.  And you can tell in the video by the way the tube popped, that I didn't have 100% of the air purged. Maybe use the electric pump with a continuous purging loop to get the air out, and then seal that off and use the hand pump for the final pressurization.  When it is free of air, it doesn't take many pumps to raise the pressure all the way up.
Lloyd
P.S.  Thanks for the link.  Looks like neat stuff!

A continuous loop with purge port would be the best option to eliminate all the air and a recirculation of the  fluid at low or no pressure and it would allow the use of an automatic purge vent then close the isolation valve and begin the full pressure test

[lloyd-ss]

A continuous loop with purge port would be the best option to eliminate all the air and a recirculation of the  fluid at low or no pressure and it would allow the use of an automatic purge vent then close the isolation valve and begin the full pressure test

Don, yes, that system sounds like a good way to do the purging.  What were we talking about?..... going down the rabbit hole and never coming back out?   Hmmm, now if the place MichaelThomas uses only charges $75 for a test, by the time I am finished with this, I could have bought a lot of tests.  Too bad I am so hard headed and always want to do it my way, right now!. It costs me.
Lloyd

[Sfttailrdr46]

A continuous loop with purge port would be the best option to eliminate all the air and a recirculation of the  fluid at low or no pressure and it would allow the use of an automatic purge vent then close the isolation valve and begin the full pressure test

Don, yes, that system sounds like a good way to do the purging.  What were we talking about?..... going down the rabbit hole and never coming back out?   Hmmm, now if the place MichaelThomas uses only charges $75 for a test, by the time I am finished with this, I could have bought a lot of tests.  Too bad I am so hard headed and always want to do it my way, right now!. It costs me.
Lloyd

Yep !! welcome to the world of R&D and the progression of improvements in both test equipment and end product

[lloyd-ss]

Don, every hear the expression, "It's time to shoot the engineers and move on with production."?

[rsterne]

Lloyd, that is one of the COOLEST Videos you have made yet !! .... Watching the tube let go at over 16,000 psi was a real treat.... However, I did notice one thing, and I'm hoping you already saw it and I'm not being the bearer of bad news.... If you look at the slow motion part, at exactly 7:00 into the video, and look at the gauge, you will note that the needle on the gauge, when it rebounded to zero when the tube let go, travelled well past zero and stopped pointing several thousand psi below zero....

I hope that doesn't mean that rather expensive looking gauge is toast....  .... Please say it isn't so.... 

Bob

[lloyd-ss]

Bob, I managed to score that gauge on ebay for an incredible $75, and its brand new.  Very lucky on that count.  But it is a much more expensive gauge than what I paid for it, and as such, comes with instructions and is adjustable and repairable.  I am hoping the pointer just spun on the shaft, but I haven't delved into that yet.  We are having thanksgiving with relatives today so I having to sneak a minute here and there to post things, LOL.
Lloyd

[Machinist]

It's convenient to have the cylinders lying in the horizontal, but if they were vertical,  and had a bleed valve at the highest point, it would be easier to get the air out no?

Mount everything on a piece of plywood and tilt it at 45 degrees for bleeding, and have the bleeders in the corner? Then go back to horizontal for testing?

You don't have to send stuff out, pay for shipping, wait etc.  Plus blowing stuff up is fun!

Steve

[lloyd-ss]

Yield and Tensile
I didn't show it in the video, but I measured the O.D. of the tube before the test, and few times during the test, and after the test, to see if I could detect any permanent yield strain. This material has 72ksi yield and 87ksi tensile, so I was expecting to be able to measure some amount of expansion.  0.2% yield offset is almost .002", and I was measuring with mikes, so it should have been measureable. 

I stopped the filling at 11,000 psi, which is the yield point, and bled part of the pressure off and miked the tube.  No expansion at all. I pumped it up again to 13,000 ( I edited out these steps), backed the pressure off and re-measured.  No change.  Then after the tube failed, I measured again, and no change in diameter.  Rather odd, I thought, because at 16,000 psi the stress in the tube was about 100ksi, well above the yield point, AND the 87ksi tensile.

So I looked more carefully at the tube and that’s when I noticed the stretching in the threaded region of the tube.  The threads are 13/16-28, so that only leaves a wall thickness of .030.  With a 7000 pound axial load on the end plug at 16,000 psi that is about 87ksi axial stress on the thread root, and failure looked imminent.  So some yield is actually visible at the thread root area, but NOT where the end plugs actually engaged the threads.  Also, when I miked the tubes in the region of the strained threads, it measured about .001 less in diameter than it did originally, indicating some stretching. The A513 Type 5 tubing is supposed to have 10% elongation, but maybe not in all directions. The area of the failure shows some elongation in addition to the split.

[rsterne]

Lloyd, I'm wondering if a small orifice to damp the volume entering (and exiting) the gauge would prevent future damage.... and/or a one-way, spring loaded check valve, with a bleed screw, for the pressure between the check valve and the gauge.... That might have the added benefit of acting as a tell-tale for the highest pressure reached.... and the manual bleeding down (through a tiny orifice) might avoid damage to the gauge.... You would have to be careful not to use too strong a spring in the check valve or it would alter the gauge reading because of the force required to open it, of course....

Bob

[lloyd-ss]

Lloyd, I'm wondering if a small orifice to damp the volume entering (and exiting) the gauge would prevent future damage.... and/or a one-way, spring loaded check valve, with a bleed screw, for the pressure between the check valve and the gauge.... That might have the added benefit of acting as a tell-tale for the highest pressure reached.... and the manual bleeding down (through a tiny orifice) might avoid damage to the gauge.... You would have to be careful not to use too strong a spring in the check valve or it would alter the gauge reading because of the force required to open it, of course....

Bob

Bob, I plugged the output end of the hyd multiplier and pumped it up to check the condition of the components.  The big output gauge was reading about -2,500PSI (my estimate) and the small input gauge was reading zero. I pumped the input gauge up to about 1300 psi and the output gauge now read about 4500psi.  That really translates to 7000psi just as it is supposed to.  My guess is that the needle spun on its shaft,
I've have given a little bit of thought about a new snubber  or check valve to stop the oscillation but we'll see what happens during future sessions.  Hopefully, once the needle is properly secured on the shaft, the problem might go away.
Lloyd