Safety Alert please read!
 

I am a mentor for FRC team 3799 on Sat. 2/26 we had a serious situation occur while pressurizing our Clipper PVC air cylinders. Please read the below letter written by our program leader to our Regional Director.

Glen,
During robot testing today we had a catastrophic failure of two of the new Clippard Air Tanks simultaneously. Luckily no one was seriously hurt. We had them pressurized to approx. 110 psi and during leak checking they blew apart into small shards (see photo). I sent a note to Andy Mark (the supplier we purchased them from) earlier today. If there is someone at FIRST that should be aware of this potential safety issue please forward this to them. We were excited about using the new lightweight tanks but as a result of this have now decided to convert to an aluminum tank.

Scott

Attachment 13680

We don't really know what caused this but we just want to make sure everyone is aware of this and are taking all proper safety cautions while testing and using.

Re: Safety Alert please read!
 

Wow! We haven't had a problem with our plastic tanks from Pneuaire. This just serves as a reminder of the power of a compressed gas. Hopefully everyone was okay.

Re: Safety Alert please read!
 

Glad to hear that no one was hurt.
Wondering, did you use exercise force in screwing in the fittings on the tanks or mount them in a way that was putting more then need force on the tank?

Re: Safety Alert please read!
 

The tank looks pretty marked up. Was it dropped beforehand?

Re: Safety Alert please read!
 

This is an interesting failure.

Just out of curiosity, with the recent cold weather, what were the odds that the tank was exposed to temperatures outside of it's operating range before the failure?

Re: Safety Alert please read!
 

Most ordinary industrial pneumatic parts are designed to burst at over 300 lb/in²; the working ratings (here 150 lb/in²) are lower so that you never approach the burst pressure, even in a dynamic load situation, and are consequently very conservative in an essentially static case.

Were you running any liquid in them, or running them in a liquid?

Did the tanks ever get exposed to any chemical agents whatsoever? Markers/paint/organic solvents/adhesives/etc.? (The only thing¹ you should clean them with is water: not even isopropanol is guaranteed to be safe.)

¹ Edit: Unless you've properly researched the cleaning agent's compatibility with polypropylene, and particularly its tendency to either cause crazing or to dissolve the plastic.

Re: Safety Alert please read!
 

I'm interested in the fact that one of the fittings on the end has been split in two and the other is completely intact. The fitting is a pretty beefed up area that shouldn't be under a lot of stress from the pressure. I'm leaning towards someone overtightening the fitting on that end and starting a crack or creating a highly stressed area where a crack is likely to form. NPT threads are tapered, so it's pretty easy to crack a fitting by overtightening. I've seen guys at work crack brass fittings sometimes, after all.

Re: Safety Alert please read!
 

We installed fittings on these tanks yesterday, and I can absolutely see how this happened.

As a student and I were tightening the NPT fittings, we wondered "when should we stop?"

The plastic ports didn't offer increasing resistance as you tightened fittings into them, like metal ports do. They just spun and spun and spun with the exact same resistance as we went.

We stopped when we still saw 1-2 threads exposed on the fittings.

They still felt loose in comparison to when you put fittings on metal components, but they did not leak.

WE VERY EASILY could have tightened them so much that the fitting bottomed out, and only then would it feel like the turning resistance increased. By then, you're pre-loading the thread, and could fracture the end of the cylinder...

This is speculation of course, but for any teams using these tanks:

Tighten your fittings until you see 1-2 threads exposed. That is tight enough to NOT LEAK, given you've properly applied teflon tape/have sealant already applied to the fitting. Most fittings in the pneumatic KoP bag have white sealant already applied to the threads.

Don't risk tightening them any more, if this is "tight enough."

It is our position that PVC pipe shall not be used as a means of transporting compres
 

There are numerous OSHA references on the internet that say;

"It is our position that PVC pipe shall not be used as a means of transporting compressed air. "
http://www.osha.gov/dts/hib/hib_data/hib19880520.html

The danger is that it shatters upon failure. There are types of plastic pipe rated for compressed gas use but I do not think the standard PVC pipe should be used for compressed air.

Checked AM and the Clippard air tank is listed as made of polypropylene plastic -not PVC but given how it shattered I would be wary of using.

Re: Safety Alert please read!
 

I have to ask this question. Do we know for certain that these tanks are PVC?
The failure definitely looks like a PVC failure . Is there any chance that these are made of another material?

I would be very surprised if these actually were PVC! Providing those cylinders as FIRST CHOICE items, if they are PVC, is an egregious failure in itself.

Re: Safety Alert please read!
 

They are polyprolyene...

http://www.clippard.com/part/AVT-PP-41

Re: It is our position that PVC pipe shall not be used as a means of transporting com
 

I suspect that with the inclusion of plasticizers, a PVC pipe could be produced that did not exhibit this failure behaviour.

It's an overreach to take the position that "PVC pipe shall not be used as a means of transporting compressed air" without qualifying the pipe's composition, manufacturing process, installation and operating conditions. Despite their strongly-worded statements, I hope that OSHA is recognizing competent engineering as the appropriate way to circumvent their prohibition.

Re: Safety Alert please read!
 

The Cy-Ranch Cy-Borgs have had several issues with the PVC-esque tubes for pressure, one had a hairline crack in it that ran down a good length of the piece of tubing, thankfully we found it and marked it out of the crowd, we have quite a few pieces of that stuff labeled to go on our robot, but I'm going to send our lead mentor this thread just to be safe! So sorry for your setback, best of luck on recovery! :yikes:

Re: Safety Alert please read!
 

What's the part number of the component? (Are you talking about the plastic air reservoirs or the plastic pneumatic tubing?)

Re: It is our position that PVC pipe shall not be used as a means of transporting com
 

PVC pipe manufacturers also say do not use PVC pipe for compressed air systems. The big problem is that it shatters on failure. That warning along with the OSHA statement is enough for me not to use on compressed air systems. For OSHA competent engineering means following their warnings.

One possibility is the fitting was overtightened, starting the failure.

Re: It is our position that PVC pipe shall not be used as a means of transporting com
 

That's PVC pipe. I believe Tristan's point was that PVC the plastic could be made in a non-shattering formulation with added plasticizers. After all, I can get high grip gloves that are coated in PVC. I'm pretty sure that formulation of PVC isn't going to shatter at anything above dry-ice temperatures.

In other words, banning "PVC" from compressed air usage is about the same as banning "steel" from being used for something. Both words encompass many materials with a broad range of physical properties.

Re: Safety Alert please read!
 

I expect your gloves will not hold 90 PSI either :) The warning is specifically about PVC & compressed air systems. I suppose a mfr could find a formulation, test it, test, certify it. But then it would not be the commonly available PVC that the warning applies to. The warning also allows covering the PVC with a shatter proof material covering it. That would be an engineered solution I suppose.

BTW I have several applications that I don't use "steel" for as well. ;)

Re: Safety Alert please read!
 

Scott,
I hope you kept the pieces. I suspect Andy and Mark would like to take a look at these and contact the manufacturer. I certainly would like to see them.
The OSHA statement comes from an industrial accident in Texas at a chemical plant or refinery, I can't remember which. The plant was plumbed for pressure lines between buildings and process positions. As an employee was walking past one of the pipes, it burst sending shards into his arm. With a little research you can find the original report and findings of the resulting investigation.
From the pieces shown, it is possible that the first failure occurred at the fitting but the resulting failure of the rest of the tank is unusual. Was the tank chucked up in a vise at some point to assist with fitting insertion? Is it possible that the two tanks were close enough that the failure of one caused a failure in the other? What compressor were you using? The large Thomas compressor can get up to 150-160 psi if the pressure relief valve is not calibrated but that should be well below the failure point for these tanks.

Re: Safety Alert please read!
 

This reminds me of an experience I once had with some lexan sheet. Lexan is normally a very resilient material very resistant to a brittle fracture. On one of my projects I had used lexan panels as body parts on a vehicle. A friend of mine who was helping out decided to clean them using aerosol brake cleaner. After wiping them down with the brake cleaner and a rag, approximately one hour later the entire panel shattered into tiny pieces. We deduced that the chemical agents in the brake cleaner had been absorbed into the plastic causing it to become incredibly brittle. It is possible that a similar process may have affected these two tanks.

Re: Safety Alert please read!
 

That's just the thing: apparently some state OSHA offices prohibited PVC pipe outright (e.g. Arizona, from which the quotation at issue was sourced) without specifying whether they were only talking about the conventional PVC water pipe. Unless they are confident that no safer PVC pipe can be made, they should only restrict the specific products that pose the hazard.

Exactly: that better-engineered solution conflicts with the prohibitions described by the Arizona, Washington and possibly Nevada OSHA offices. The main OSHA memorandum is better, because it recognizes appropriate ways to use the pipe.

Re: Safety Alert please read!
 

Fox,
The same thing happens if Loctite gets on the Lexan.

Re: Safety Alert please read!
 

There is plastic pipe suitable for compressed gas use and they are typically labeled as such. They are designed to split open upon failure and not throw fragments.

I mentioned this issue for 2 reasons;

1) A lot of folks do not realize that it is not a good idea to use the rigid PVC water pipe from the hardware store for compressed gases (unless buried or encased) because it shatters upon failure.
2) The air tank in question is listed as being made of polypropylene but the way it shattered upon failure is cause for concern.

Re: Safety Alert please read!
 

Update on the exploded tanks, I had time today to do some analysis of the fragments, first of all this is of my own opinion and I have reviewed this with some of my colleagues to confirm my findings. This had nothing to do with the handling of tanks all marks on the tanks came from the parts being scattered across a machine shop floor. It also had nothing to do with temperature the tanks had been at room temperature for a few days. Like most before looking we were thinking damage by over tightening but this was not the case. What I found was a elongated air bubble in the material. The bubble looked like a worm hole around part if the threaded openings base expanding in size and thinning out the threaded wall. My assumption is the fitting stressed the wall enough to open a crack to the bubble that ripped apart when under pressure. When the tank exploded the fragments caused the tank next to it to blow apart.
The one thing I would like to know would be does the manufacturer do any type of vision inspection either by ultrasound or x-ray of the tanks. Because if they don't how many more of these tanks could have hidden rupture points.

Re: Safety Alert please read!
 

I will make sure the folks at Clippard see this thread first thing Monday morning. They are one of our team sponsors.

Re: Safety Alert please read!
 

We at AndyMark have contacted the manufacturer about this issue, as has FIRST. More news will come from this today or tomorrow. Hopefully, we will have explanations and a direction of what to do.

Andy Baker

Re: Safety Alert please read!
 

That failure really surprises me, I would have not expected it to shatter. I've seen burst PVC pipe (schd 40) where people used it to make pneumatic air cannons. They charged and then dropped the cylinder and it exploded. This failure looks EXACTLY like that.

My guess is that this cylinder had been dropped at some point on a corner or edge which can stress fracture and weaken it a lot. Never EVER drop a charged cylinder - very bad things happen. Another possibility is that it's been in the sun for a long while at some point in it's life, that can weaken most plastics a lot too. Some weakening combined with over-tightening and stressing the threads == kablowie!

-Mike

Re: Safety Alert please read!
 

What type and part number fittings did you use? Did you inspect the tanks after screwing the fittings?

Re: Safety Alert please read!
 

:eek:

When we first got those things the first things that came to my mind were, 'someone is gonna over tighten a fitting and crack it or strip the treads out' and 'I'd hate to get one with a defect'.

Re: Safety Alert please read!
 

Thank you 3799 for bringing this to our attention. After reading about this Sunday morning we decided to change out our 2012 plastic tank and install a metal one in it's place. We felt that for safety reasons it was better to wait and see what the final determination of the problem is before going back to plastic tanks. Our tank is a different production style but why take any risks with the students.

Re: Safety Alert please read!
 

I've been mulling this around in my head after seeing the kids had been experimenting with pneumatics this past weekend. In the interim, if teams insist on using the plastic tanks rather than switching to metal, what about placing some sort of containment device around the tanks? I was thinking a heavy sock with tie wraps around where the pneumatic tube connects to the fitting. At the very least it should help contain any shrapnel or at least reduce its kinetic energy should one rupture.

Re: Safety Alert please read!
 

I wonder whether mere tensile strength is sufficient, or whether the shearing action of those shards in motion is a consideration? And do small shards pose less of a risk due to their small mass (like birdshot compared to buckshot)?

Although your suggestion of a sock sounds pretty good, what about versions of the mesh bags used to contain onions and other similar vegetables? (Those might be easier to incorporate into existing designs.)


This is an opportunity for a good discussion on risk, and what drives prudent decision-making. How much of the hazard and the associated uncertainty are realistically transferable between the specific tank that failed, others in its lot, others of the same model, others in the product line, and other product lines of similar manufacture? Or is this strictly a matter of perception being everything—action is being taken, so outsiders will perceive it as the responsible course of action?

I'm curious about what specific factors are driving your decision—not particularly to criticize you, but to discuss the proposition that a drastic response often isn't a methodologically valid way to approach a spectacular failure.

After all, you (and everyone else) obviously take plenty of risks with the students. It's our responsibility to try to manage those risks in a sensible way.

Re: Safety Alert please read!
 

2 tanks burst at the same time? What was the 'leak checking' procedure? (tightening fittings or spraying soap?)

Were the tanks connected together with a common brass fitting, or a flexible hose between them?

Re: Safety Alert please read!
 

Guys,
The OP posted that further examination showed what appeared to be a manufacturing defect in the plastic. Please remember that these tanks have been used on FRC robots in the past without failures.

Re: Safety Alert please read!
 

In addition to Al's statement above: Clippard, FIRST and AndyMark take safety VERY seriously and all of those parties are currently evaluating the situation. I believe they will have a statement and course of action very soon.

Re: Safety Alert please read!
 

Al, correct me if I'm wrong, but I was of the impression that these Clippard tanks were new, and not before tested in an FRC environment, but that they were an answer to similar model tanks from another manufacturer used in the last couple FRC seasons by some of the heavy hitters like 217 and 1114.

Re: Safety Alert please read!
 

Yes the Clippard tanks are new and not what we used last year. We will be harvesting the tanks from last years bots until this is resolved.

Re: Safety Alert please read!
 

Would it make it safer to put a wrapping of duct tape around them, at least for testing and until we are given an actual course of action? I realize this would likely render them illegal, but might help to contain any shards since dict tape usually splits on a seam.

Re: Safety Alert please read!
 

The chemicals in the adhesive of duct tape could also potentially weaken the plastic, I would be wary of this as a solution.

Re: Safety Alert please read!
 

Theoretically, same materials, similar specs. I would suspect we are going to have a response soon. Please be patient. Use the tanks from last year in the meantime.

Re: Safety Alert please read!
 

Indeed they should, but all too often people who don't understand the science behind a decision are the ones who make it, especially when the government is involved...

I'll never forget teaching in California when a law was passed (unanimously by both houses and signed into law by Governor Davis) that outlawed "chemicals made from dangerous elements" in science classrooms... Further investigation indicated that "dangerous" included anything "toxic" or "flammable".

Not withstanding the absurd misunderstanding that toxicology is a matter of dosage (and that even water ingested in sufficient amounts is toxic), the ignorance of even middle school level chemistry was quite astounding.

Among the things unintentionally outlawed in California science classrooms: matches, table salt, water, air, humans.

Re: Safety Alert please read!
 

My thoughts around using the sock was that they would be elastic and absorb the energy of an expanding tank while allowing the compressed air to permeate the fabric and release pressure. The onion bag idea could work as well but are they not a relatively non-elastic material?

I think since the duct tape would be wrapped around the tank tightly it would tear as the tank expanded. I feel it would also contain the compressed gas which could cause it in itself to burst like a balloon.

Re: Safety Alert please read!
 

I'll just leave this here...

Kevlar Biaxial Sleeving

I do not advocate teams reinforcing their plastic tanks.

Though the added weight a reinforcing a plastic tank might make metal tanks weight-effective again. They will certainly be safer.

Re: Safety Alert please read!
 

From the Tristan

"I'm curious about what specific factors are driving your decision—not particularly to criticize you."

No criticism taken. Here is the reasons for my actions. Another school had an accident a few years ago and we are under a microscope. A failure of one of our tanks after being the alerted to the problem would cause us serious trouble. Our students run our machine shop and welder. I do not want to risk that program. The type of tank is not critical for out prototype and the definitive answer should be out in time for the final robot. While we all manage risks all the time because of our situation this risk is not needed or worth taking.

Re: Safety Alert please read!
 

FYI, here is an email I just received.

Re: Safety Alert please read!
 

You really can't beat that for customer service!

Re: Safety Alert please read!
 

We have at least 8 of the Clippard Tanks...until we know for sure if they're actually bad, I think that if anything we'll just ask for the minimum number of replacements that we actually need on robots this season. If they are indeed safe, the weight savings are worth it.

Re: Safety Alert please read!
 

Does someone have the email address to request replacement air tanks from Clippart?

Thanks!

Re: Safety Alert please read!
 

One of our air tank's was cracked on the "rim" right out of the box. We're glad we spotted it before we put it on to our robot all thanks to this thread. Thank you, and were glad everyone's okay!

Re: Safety Alert please read!
 

Spencer,
I bet that the manufacturer will want to know or see that defect.

Re: Safety Alert please read!
 

One legal way to provide some measure of protection of your tanks from external damage, or shrapnel containment would be to use expandable mesh sleeving. This is the stuff usually used for wire bundling, and which also can be used to make pneumatic muscles.

McMaster Carr has it in various colors in a size suitable for slipping your tank inside.
ID range: 1 1/2" to 3 1/2"
Part number: 9284K419
Price: $14.53 for 10 feet.

Re: Safety Alert please read!
 

Todd,
While this product might contain the broken parts, I would not want to mislead people into thinking this a good fix. This mesh is designed for something entirely different.

Re: Safety Alert please read!
 

this picture describes how my team are handling the air cylinders after this tread ;)


If you have definitive decisions please let us know, we are in Brazil ... swap or buy new cylinders would lose more than 2 weeks

Jeser
1772

Re: Safety Alert please read!
 

I know this thread is old news, but I want to caution teams with these tanks still on hand. One of our mentors attempted to install a 90 degree fitting (plastic -- from FIRST Choice) on one of our tanks today, he was only able to turn the fitting about 400 degrees before the rib on the port cracked open. It was a hairline crack, hardly even noticeable to the eye, however it extended the entire length of the threads and was about an 1/8" deep. Since he was watching for this to happen, the flaw was discovered before a real problem occurred. Our plan had been to test each tank with shop air (~140psi) prior to use, after seeing how fast these crack we are moving on to other tanks. If you are going to continue using these tanks, please do so with caution.

Re: Safety Alert please read!
 

Yikes! Glad everyone is safe!

Re: Safety Alert please read!
 

Did FIRST include this information in an email blast? I was visiting a team earlier this week and they didn't know about this issue. They were in the process of setting up a pneumatics breadboard. I showed them the blog post and a bit of pneumatics safety. They decided to plumb everything as training, but were not going to use these tanks in a live system. I am working with a different subgroup when 20 minutes later when one of the pneumatics kids came over with a cracked cylinder. They had followed the blog installation instructions and had put a crack that is vertically the full length of the threading, out through one of the ribs, and about 1 cm down the outside of the tube. They said they probably would not have noticed if I hadn't pointed out the blog post and they were concerned about doing it correctly and looking at it closely.

The point is that teams still don't know, and tanks are still having issues.

Wetzel

Re: Safety Alert please read!
 

I haven't seen a FIRST email blast in two weeks-unusual during build season.

Just the Blog warning.

And a cryptic notice to look at the blog warning, on the KOP Technical Resources webpage.

Re: Safety Alert please read!
 

The email blasts have been going out. I do not recall see a warning about the air cylinders on them though. There are also on first's web site if anybody wants to look through them.