GTA
All Springer/NP/PCP Air Gun Discussion General => "Bob and Lloyds Workshop" => Topic started by: lloyd-ss on June 18, 2012, 07:55:07 AM
-
I would like to do some testing of components at extreme pressures and think that using hydraulic pressure is a safer way do do it. There are hydrostatic pressure test pumps available but instead I made an air-over-oil pressure boost cylinder. Basically, a 1.18 dia piston with high pressure air on it, driving a .482 dia piston rod over the hydraulic oil. That is a 5.4 to one pressure multiplication. In the video I easily hit 6,000 psi, and actually, I hit higher pressures while I was working the bugs out and didn't realize an o-ring was blocking the oil flow to the high pressure gauge. Honestly, with all the work in this booster cylinder, purchasing a hydrostatic pump would have been so much easier, LOL. But the device really works. I'll eventually show the guts of it.
Now, what do I do with all that hydraulic pressure?
Lloyd-ss
P.S. All the prelim pressure work to proof the system was done behind proper barriers, and the safety margins within the system are adequate.
http://www.youtube.com/watch?v=q6xzS-Z7wRE# (http://www.youtube.com/watch?v=q6xzS-Z7wRE#)
www.airgunlab.com (http://www.airgunlab.com)
-
My thoughts are that at failure, the air in the primary side can expand and drive the hydraulic oil in an explosive manner.... I thought hydrostatic testing was done with NO air anywhere in the system so that upon failure there was nothing to suddenly expand and throw parts around....
The way I have seen it done at home is with a grease gun or the hydraulic pump from a port-o-power....
Bob
-
Isn't the air going to be on the upstream side of the large piston? Regards, Tom
-
My thoughts are that at failure, the air in the primary side can expand and drive the hydraulic oil in an explosive manner.... I thought hydrostatic testing was done with NO air anywhere in the system so that upon failure there was nothing to suddenly expand and throw parts around....
The way I have seen it done at home is with a grease gun or the hydraulic pump from a port-o-power....
Bob
Bob,
There are many ways to do destructive testing. An awareness of the possible failure modes is paramount, and I appreciate your cautionary notes. With air over hydraulic, the air and oil are seperated and do not come in contact with each other (thanks, Tom). The piston assy has a positive stop to limit its travel, and there is a finite volume of hydraulic oil that can be released from the failure point under greatly reduced pressure and at a controlled rate. This is controlled by the the available volume of oil in .482 dia cylinder, in this case 6 ccs (1/3 cuin), and by the size of the metering orifice within the booster that supplies the oil to the device under test. The only burst or "explosion" happens at initial failure, and then only if a catistrophic failure occurs. The oil does not expand like an air release would. Because the air is (1)properly contained in its cylinder in the booster, and (2)does not mix with the oil and (3) because the release of the oil is controlled via a mtering orifice, it will not contribute to any additonal uncontrolled energy release like air that was compresssed within the oil might do.
Probably a diagram of the system would have been helpful.
Lloyd
-
Just like with SCUBA equipment .. HYDROSTATIC testing of cylinders, Valves, handling apparatus etc are done with high pressure WATER with parts/pieces subjected to the HP submerged in water ......
Something fails of gives way it is just a splat of very little volume with the outward displacement of the failing part being absorbed into a vessel of water containing any fragmented parts/pieces displaced.
-
I understood that the air and oil were separated by a piston.... My concern was that on catastrophic failure of the component being tested, the piston (driven by expanding air) would expel the oil at high pressure.... By limiting the volume of oil, and including an orifice, you have minimized the amount (and speed) of the oil that can be released.... I still see the possibilty of a high-speed jet of 1000+ psi oil coming out of something like a crack in the test object (until the 6cc is gone), but as long as your containment looks after that I'm sure you will be fine....
I should have known you have thought through all the possibilities, but like you say, better safe than sorry.... You are much to valuable to lose to a silly acccident that could have been prevented....
Bob
-
I worked for a Special Effects company for a while that had some very smart MEs. They tested their pressure vessels with with a 6k nitrogen tank and water. The vessel and pressure hose assembly was filled with water and then attached to the tank. The hose assembly held very little water but was deemed adequate for the job. The only barrier was a sheet of plywood. A regulator with a flow restrictor was used and they just slowly cranked up the pressure. I did see one the these vessels fail at 1000 psi, the weld simply crackled open and the water sprayed out like a sprinkler.
Lloyd, I think your system is far superior and safer.
Tom
-
Thanks to everyone for the comments and concern. I promise not to do anything dumb. I originally considered getting a hydrostatic hand pump ($300), or using the pump assy in a bottle jack (didn't think it would hit 10,000 psi), so I decided to make my own, which is usually what I convince myself was the "best" way to do it anyhow. I think you already know this, but I would always rather make than buy, even if it can't be justified monetarily. And if I already have the parts and material laying around, so much the better. For this endeavor, I had to buy a gallon of hydraulic oil. Not too bad.
We keep doing calculations about material strength and safety factors.... which is definitely good.... but I have never seen anybody post empirical data regarding the airguns we take for granted every day. Maybe I have missed some thing that has already been posted, but if I haven't, does this lack of empirical data bother anybody else? Or at least make you curious. I am 90% sure there will be o-ring and seal leaks long before any metal parts come close to failing. Probably very anticlimatic. It is quite likely that a material failure will have to be forced by bolstering the built-in safety failure points that occur at much lower pressures. My guess is that the airguns we trust every day, when used properly, are made with a comfortable safety margin.
Just my train of thought,
Lloyd
-
I think it great that you are taking on the empirical testing. Like you said, we do stress calcs. all the time, but wouldn't it be great to see the actual failure mode? A whole new level of insight will be obtained. Simply invaluable.
Tom
-
The only test I am aware of was by Walter Glover on a 2240 tube.... He installed a plug in place of the valve, using three 8-32 low profile SHCSs (200,000 psi tensile) with the head set into a shallow recess in the plug (and hence the head was partially in shear), so he wasn't testing the "Disco style" fasteners.... The other end used a stock CO2 cap with a plug sitting against it.... Both plugs were sealed with O-rings.... He reached 10,000 psi (the limit of his pump) without catastrophic failure, even though the roll-stamped area of the tube was pressurized.... We don't, however, know if the roll-stamp was a shallow one or a deep one, they vary a lot....
The only "failure" noted was the tube deforming slightly around the "valve" screw heads at about 7,000 psi.... It yielded until it took a "set" where all three screws were under equal load, and then quit moving.... He never repeated the test with a stock Disco valve attachment, unfortunately....
Bob
-
Just like with SCUBA equipment .. HYDROSTATIC testing of cylinders, Valves, handling apparatus etc are done with high pressure WATER with parts/pieces subjected to the HP submerged in water ......
Something fails of gives way it is just a splat of very little volume with the outward displacement of the failing part being absorbed into a vessel of water containing any fragmented parts/pieces displaced.
Cheaper than hydraulic oil and a lot easier to clean up too........:)
-
Bob,
Thanks for letting us know about Walter's work. He's a pretty solid source, and I am glad to hear that he was able to max out his set-up without a failure.
William,
Yes, with using the oil for the testing I have to set-up outside, which is probably a good idea, anyway. Because some of the fits I am dealing with aren't exactly perfect, I think the lubricity of the hyd oil will be a good thing and might prevent any possible galling, and I think oil needs a bigger hole to leak thru than water. Does anybody know if that is true? Plus, rust just really bothers me. :(
Thanks,Lloyd-ss
-
Lloyd maybe try looking at the home page of koni shocks and tokico shocks. There aftermarket adjustable shocks for cars and some of the best in their industry. Neways I'm willing to bet that somewhere on their website they have a picture of a shock that's
"cut away" showing to some degree the hole sizes they use. They won't give you measurements but I'm sure you of all people can scale them to fit ur need by getting at least a look at them.
-
Using a 1" diameter aluminum tube for the P-rod double kit gave me the incentive to do some pressure testing with this rig.
The 2024T3 aluminum is rated such that the tube should start to distort at 9,500psi, and then if everything held together, burst at 13,300psi.
My rig, or at least the pressure gauge, only goes to 10,000psi. I went to a lot of trouble to set everything up on the back deck of my shop so that if an o-ring popped or anything came loose, at least the hydraulic oil cleanup would be outside. I am glad I played it safe, but.......
Lloyd-ss
www.airgunlab.com (http://www.airgunlab.com)
http://www.youtube.com/watch?v=kjYX39htLIM# (http://www.youtube.com/watch?v=kjYX39htLIM#)
-
Have fun with this bud. Air over oil is a common way where we work to test fittings and valves to failure. In school we used pure hydraulic power to test tubing. Really looking forward to your results!
-
Great video.... Did you measure the diameter of the tube before, during, and after the test?.... I would expect that the diameter during would be slightly increased, but if the diameter afterwards was any larger than the initial diameter, you would then have proof of yield.... I understand that is how SCUBA tanks are tested, and that they have criteria for the diameter change during and after.... Not necessary for your purposes, just of interest....
Bob
-
Bob,
I will the tube diameter tonight to see if a permanent deformation occured, but I really don't want to touch it when it has 10,000 psi on it during the test.
vrotsos991,
Well, with no failure, I think I am done with the aluminum tubing, and honestly I am very glad it didn't fail. I was expecting maybe an o-ring to blow out, but nothing happened at all. Kinda boring! There is a chance that air would have leaked before hydraulic oil, but an actual failure of the tube should be at the same pressure, whether tested with water or oil or air. (But failure when testing with air would be really dangerous.)
Lloyd-ss
-
That was impressive. I don't guess I've ever considered aluminum as a viable pressure vessel. My prejudice never allowed me to consider it even though I know scuba tanks are aluminum. I'm reviewing strength of materials all over again, I have future project that this would be wonderful on.
Bill
-
Well, with no failure, I think I am done with the aluminum tubing, and honestly I am very glad it didn't fail.
Nice vid but a bit boring. LOL. When will we see one taken to failure?
Done with testing? or done as a viable material for some reason?
If you have a height gauge with a dial indicator attached you could measure the yield dynamically. Just lay tube on a plate an bring indicator down on mid-tube center. Pump and watch :)
Tom
-
That was impressive. I don't guess I've ever considered aluminum as a viable pressure vessel. My prejudice never allowed me to consider it even though I know scuba tanks are aluminum. I'm reviewing strength of materials all over again, I have future project that this would be wonderful on.
Bill
Bill,
I really like 7075 for machined "blocky" things. It machines beautifully and tapped holes are very strong. 2024 is good for tubes and as strong as many steels.
Lloyd
-
Did we really exceed the yield point of the aluminum tube as calculated ??? ::) Well, actually, yes we did :o
The original tube dia was .999 to 1.000. I checked the tube with calipers and decided to get the mikes out to check it more carefully.
Both of the ends of the tubes are isolated from the pressure by the o-ring on each end plug, and they did not change size.
The middle of the tube swelled to about 1.0035 to 1.0045 So that is 0.4 percent (less than one half of one percent) permanent deformation. ???
Yield offset is normally .2 percent, so this is barely, but definitely, into the permanent deformation region. But the material is behaving exactly as it is supposed to, so I am happy ! ;D ;D If all follows as it is supposed to, the tube wouldn't burst until over 13,000 psi, and I don't know anybody who has that kind of air or nitrogen easily available.
So my opinion is that the 2024T3 aluminum airtubes, when properly designed, manufactured, installed, used, and maintained, are very safe.
Lloyd-ss
-
EXCELLENT RESULTS !!!.... It is actually very reassuring that you found some permanent expansion in the diameter, it shows that the material is on spec and that our math is good as well.... Thanks for measuring it up after the test.... I would agree that your tubing choice is a good one for the purpose, in fact for anything running at 3000 psi or below.... just over a 3:1 safety margin to yield....
Bob
-
hey Lloyd,
I have to weigh in on your pressure step-up. You've still got a substantial air spring compressed. I would suggest that for failure testing you use a hydraulic pump on the low pressure side. Something like a Porta-Power's pump...keeps stored energy to a minimum, which IMO is a very good thing whilst running pressure testing.
cheers,
Douglas
-
Douglas. We discussed this earlier in the thread, and I have designed the system with adequate precautions. i appreciate the concern. But if you can loan me a porta.power,that would be even better! LOL
Lloyd-ss
-
Douglas. We discussed this earlier in the thread, and I have designed the system with adequate precautions. i appreciate the concern. But if you can loan me a porta.power,that would be even better! LOL
Lloyd-ss
On the pump, I'll have to agree with you; they are not easy to come by inexpensively...:) I'll let you know when I finally find one( which loaning to you might be more productive than keeping it ).
cheers,
Douglas
-
The aluminum airtube passed with flying colors at 10,000 psi. To get it to fail I had to machine the O.D. of the tube down to make the tube wall artificially thin so that it would rupture below 10,000 psi. And it did fail rather nicely after machining 1/2 the wall thickness away..
Lloyd-ss
http://www.youtube.com/watch?feature=player_embedded&v=AwVgPtogySA# (http://www.youtube.com/watch?feature=player_embedded&v=AwVgPtogySA#)
-
that was cooool. I was wondering if you would take the wall thickness reduction across the threads or if you'd leave the threads( and o-ring seal ) full and have a wall burst. Nice vid BTW.
cheers,
Douglas
-
Douglas, to be perfectly honest, I was not anticipating that failure, and thought I had about 2k more to go before the wall ruptured. But after seeing it and doing the calcs, the thread root was the weak link for sure. When I first saw the failure, I thought the threads pulled out. The material stretching at the failure line is very easy to see with an eye loupe. You should have seen the barricade I was behind for that test, LOL.
lloyd-ss
-
Lloyd, I'd say you need an apprentice just to clean up after you on that one. ;D
-
Bursting tanks are not your only concern.
Petroleum oil aspirated into the air is deadly.
You are talking about a lot of pressure which can easily accomplish this.
The reason it is deadly is simple. It coats the alveoli prevent them from being able to effect a transport of O2 into the blood stream and the lungs can not clear it in time to prevent death.
-
Cliff.
Thanks for the concern and the info. During the recording, i was inside my shp and the test set-up was outside. the video cam was outside too
lloyd.
-
Douglas, to be perfectly honest, I was not anticipating that failure, and thought I had about 2k more to go before the wall ruptured. But after seeing it and doing the calcs, the thread root was the weak link for sure. When I first saw the failure, I thought the threads pulled out. The material stretching at the failure line is very easy to see with an eye loupe. You should have seen the barricade I was behind for that test, LOL.
lloyd-ss
The "pop" had more energy than I expected :)
With the data you now have, you could calculate the thread root area and see where it failed with respect to the material properties. This would give an estimation of how the stress concentration (notch sensitivity of the matl) from the national v form thread affects the tensile of the material. Hard to nail down with one test but a ball park fiddle factor for the threaded area would shed some valuable insight for future designs.
Tom
-
Tom, I was planning on doing that but it slipped downward on the to-do list. You pushed it back up, LOL.
The hole in the tube is tapped so that it only has one spare thread before the plug jams in the threads, and the failure occurred at the root of that one lone full thread. The tube is tapped 7/8-24 and I just miked the tap at .880 dia. The tube mikes at .910 O.D. so that gives .015 wall, or .0422 sq in.
The plug is sealed on the as-drawn I.D of the tube which is .810 (1" tube with .095 wall, verified) or .5153 sqin. It looks like the failure occurred at 5,100 psi which is 2628 pounds of force (pardon the significant digits), which calculates out to a stress of about 62,300 psi. The figures I have been using for 2024T3 are 50ksi yield and 70ksi tensile. So the 62,300 break seems appropriate for the stress concentrations at the root. In fact, considering the very thin .015 wall, and the cut thread, I'd say that is quite good.
What do you think?
As an aside, if the tube were full diameter of 1", that would give .1772 sqin, or a loading of 1804 pounds at 3,500 psi air, or a stress of 10,170 psi, or a safety factor of greater than 6 to 1. I think I can sleep peacefully on that one.
Lloyd-ss
-
Really great work Lloyd. The testing is great for the the advancement of PCP hardware. Amazing that there is aluminum tubing out there that can hold up to these pressures and be lighter than steel.
Gene
-
Lloyd I know that many of the BB makers are using 4130 CM . How does this compare weight wise as well as ding, dent resistance as well .Marvin
-
The CM is a stronger and tougher material, but significantly heavier. But the 2024-T3 aluminum that I used is a hard, high strength alloyy, that is used in aircraft airframe construction. It might not be as tough as the CM but it is strong and fatigue resistant.
For the 1" tube, I used .095 wall 2024 aluminum , and the safety factors (3 to one and better) allow it to operate at about 3200psi.
A CM tube of the same strength could be about .065 wall.
In comparison, considering a 16" long tube:
Al .095 wall CM .065 wall
Al 135cc CM 156cc
Al .41 lbs CM .88 lbs
Bottom line is, if you are rough on equipment (dings and bumps), and don't care about the extra weight, use the CM.
If you want the lighter weight, definitely go with the aluminum. I have no concerns about its strength.
Lloyd
-
The CM is a stronger and tougher material, but significantly heavier. But the 2024-T3 aluminum that I used is a hard, high strength alloyy, that is used in aircraft airframe construction. It might not be as tough as the CM but it is strong and fatigue resistant.
For the 1" tube, I used .095 wall 2024 aluminum , and the safety factors (3 to one and better) allow it to operate at about 3200psi.
A CM tube of the same strength could be about .065 wall.
In comparison, considering a 16" long tube:
Al .095 wall CM .065 wall
Al 135cc CM 156cc
Al .41 lbs CM .88 lbs
Bottom line is, if you are rough on equipment (dings and bumps), and don't care about the extra weight, use the CM.
If you want the lighter weight, definitely go with the aluminum. I have no concerns about its strength.
Lloyd
Lloyd Thank you for the info . I do not question the safety of your material or work . You at least take the time to test what you use . Most do not . Just trying to gather knowledge . Some of them blue and some paint so that might be part of the reason for their material choice as well . Marvin
-
So the 62,300 break seems appropriate for the stress concentrations at the root. In fact, considering the very thin .015 wall, and the cut thread, I'd say that is quite good.
What do you think?
Yes, I agree. Very good. The added copper content of the 2024 really adds to its toughness.
How does this compare weight wise as well as ding, dent resistance as well .Marvin
A hard anodize or nickel will go a long way towards preventing cosmetic damage but hard knocks may leave dents, I wouldn't expect them to compromise the integrity of the tube though. With in reason of course.
Tom