3D Printing Compressed Air Tanks

Using PVC pipe as a pressure vessel for compressed air can be a fun and enjoyable hobby. It’s safe, too: while there are are reports of PVC pipe being the cause of accidents, these accidents include a black powder potato gun[1], and welding too close to a PVC pipe containing compressed air[2]. Compressed air stored in a PVC pipe is never a proximal cause in any accident, and the OSHA’s Fatality and Catastrophe Investigation Summaries bear this out; there was no industrial or occupational accident recorded in these summaries where a pressure vessel made out of PVC was the cause of any injury or death[3].

Although PVC pipe can be a perfectly safe, effective, and cheap pressure vessel for hobby applications, it’s not always the best choice. A group of students in Renens, Switzerland are building autonomous robots for the Eurobot competition, and this year’s robot uses pneumatics. That means compressed air, and that means a pressure vessel. Since just about everything else on this robot is 3D printed, they asked the obvious question. Is it possible to 3D print a tank for compressed air?

The tank for this robot would only be used up to about 4 bar (400kPa), and after a few quick calculations, the team discovered the wall thickness – even in a pressure vessel with corners – would be pretty low. The first prototype, a 40mm cube with 20% infill and a hole drilled in the side, held 6.5 bar (650kPa) for an hour. This success didn’t last, though: he second prototype, a 65x40x80mm rectangular prism printed without as much infill, exploded at 5.5bar (550kPa).

The third time’s the charm, and with filleted ribs inside the tank, the third prototype was able to hold pressure up to 6.5 bar. Of course no 3D print is perfect, and the third prototype did leak, but a bit of acrylic spray paint applied to the outer surfaces held the air in.

While it’s not as fun, easy, cheap, rewarding, or safe as using PVC pipe as a pressure vessel, the team did manage to build a 3D printed pressure vessel with a custom shape. You can’t do that very easily with round pipe. And 3D printing opens up all manner of internal structure to experiment with. We’d like to see this developed even further!

Sources: [1], [2], [3]

106 thoughts on “3D Printing Compressed Air Tanks

      1. At the very least, they should disclaim this stupidity so people don’t inadvertently take this…. presumed hyperbole / nonsense for the sake of drumming up more views seriously?

        The overall outcome it has is actually probably quite opposite of the intended effect and instead causes irritation and consternation among the extremely small amount of people who actually comment by deliberately and callously playing the PVC = safe for air pressure meme for some kind of comical value (?) and page views. This deliberate attempt was extremely poorly thought through, offensive and quite frankly, dangerous.

        I expected better, Brian.

          1. Water is nearly incompressible, so it is not good at storing potential energy. Imagine using a bike pump to pressurize a tank. If you used the pump to pressurize water, you could pressurize a large tank to 100 psi with like half a pump. Air on the other hand is very compressible and might take fifty pumps to get to 100 psi. This works in reverse as well, so when a pipe full of water bursts you get a little dribble of water at the crack, and even that is mostly caused by the pressure vessel relaxing. When a pipe full of air bursts, you get an explosion as the air expands.

          2. Yes, very much so. Compressed gasses are just that – compressed, such that they occupy a smaller volume than they want to. Pressurized liquids are not compressed. Even if all the water in a pipe escapes, it still occupies one pipe worth of volume. When a pipe full of air escapes, it vigorously expands to occupy a lot more space. The danger is that it can take plastic shards along for the ride, as shrapnel. It can also be pretty rough on your hearing.

      1. NO, even with a 10:1 “safety ratio”, it is never safe. The danger isn’t over-pressurization, but the fact that when pressurized a relatively light impact could cause the pipe to burst (especially in cold weather), if the pipe bursts now it is a pipe-bomb and sprays shrapnel everywhere.

        Also consider that if you are using any solvent along the lines of acetone or lacquer thinner, the vapors will attack PVC. Now imagine 6 months after installing this widow-maker system, vapors have attacked the PVC. You walk out, pressurize your shop, start working and suddenly BANG the pipe next to your bench explodes in your face and you lose you sight to it; a permanent injury caused by trying to save a few bucks.

        OSHA is right, even in a home shop their warnings/rules still apply; you just don’t get fined for breaking the rules.

        All that said the idea of 3D printing a pressure vessel is interesting, no matter how hazardous it maybe.

        1. I know someone who had PVC air headers in his garage- after less than a year they burst. There were shards of PVC embedded in the walls and ceiling as well as all over the floor.

          1. It gets better – PVC won’t show up on an x-ray and can’t be fished out of an eye with magnets, it’s a recurrent problem in emergency rooms after idiots play games with potato guns and compressed air projects.

          2. Thinkerer hit the nail on the head about why you shouldn’t make a PCV pressure vessel. It’s not that PCV won’t hold pressure – it can take withstand as much air pressure as it can withstand water pressure. The problem is what happens if you do exceed its maximum pressure. First, PCV tends to shatter – a metal pipe has some ductility to it and can sort of rip without sending fragments all over the room. Then, if anyone gets hit by the fragments, you can’t fin them on an X ray, making it a lot harder for the ER to treat you. For anything other than very low pressure, please use metal.

          3. I’m not going to comment on the rest of this, but PVC definitely shows up on x-rays (and ultrasound). I’m not sure where that myth started, but I’ve used x-ray’s and CT scans to localize fragments, and ultimately remove them.

        1. Well someone must…Benchoff says there has never been a reported fatality from an exploding PVC “pressure vessel” and it “can be a perfectly safe, effective, and cheap pressure vessel for hobby applications”. See for yourself, build PVC tank fill it to 100 psi and drop it on a concrete floor…it won’t seem so safe afterward.

          Anyone thinking about doing this, do yourself a favor and do it right. RapidAir makes cheap airline kits for any size shop (I used them when I plumbed a 10000sqft machine shop) and if you need a cheap storage tank goto Harbor Freight and buy one for $40….it’s way cheaper than going half-assed and paying the piper another way.

          1. And Benchoff is trolling you intentionally by playing with semantics.

            PVC undergoes brittle fracturing when it fails suddenly, which is the fundamental reason why it is unsafe for pressure vessels or pressurized lines for gasses of any pressure above normal. It doesn’t need to fail spontaneously to be dangerous, because an external source of failure, like dropping or bumping the pressure vessel or having a spark from a welder puncture it still causes an explosive decompression and shrapnel to fly all over the place.

            It’s made even more brittle with age, UV exposure, and many common solvents, which makes PVC a pressure vessel just a bomb waiting to happen. That’s why the offical recommendation is to use ABS instead. Brian is just playing with words by pretending that it’s not dangerous because the cause for the eventual failure is some external factor.

  1. NASA used to publish plans for a pressurized air PVC canister paper rocket launcher, similar to the designs in Make magazine. However, in about 2012 they pulled the plans down with a warning against using PVC for holding pressurized air.

    Apparently, PVC under pressure can shatter explosively. Exposure to sunlight, heat, vibration, and impact can weaken PVC and make it more susceptible to shattering. http://nasawatch.com/archives/2012/07/nasa-yanks-air.html

    While I haven’t yet found any mention of PVC compressed air rockets causing damage (other than that NASA article), there are more than one instance online of people who have been severely injured from compressed air PVC t-shirt cannons. I suspect that one of the reasons that there are no industrial/occupational accident reports for using PVC under pressure is that it isn’t supposed to be placed under much pressure at all.

    1. Sure, but you still have the old problem of uniformity between parts which means you need to lower your maximum pressure level to well below the worst case within an accepted range of variation then include a safety value that prevents the tank going over that level. But who decides what an “accepted range of variation” is? Perhaps it does not matter if it is a one off and only the builder is exposed to it, but anything else and you have legal liability and insurance nightmares. This is why we have government mandated standards for a huge range of products and materials.

      1. As long as you use enough Carbon fiber that the failure mode is hiss and not boom, I think this is safe enough for personal projects. A rubber balloon inside a cloth bag can take far more pressure that the balloon alone and when it pops the shrapnel is contained.

        1. Carbon fibre also shatters, it is just that the pieces at so small that they generally do far less damage. What about just dipping the entire assembly in a synthetic rubber solution?

      2. “but you still have the old problem of uniformity between parts”

        No you don’t. If you are mass producing goods using a 3d printer you are doing it wrong! 3d printers are good for one-offs or maybe very small runs at sizes where you are probably going to individually test each piece anyway. If you are doing a production run then use more traditional tools. A 3d printer is going to be a waste of electricity and time plus will probably need frequent repairs.

        Use a 3d printer for porototyping or for home hobbyist projects. Don’t use it for volume manufacturing!

          1. You know the point needed to be made, the tanks cannot necessarily be fabricated any other way, so one-off or product it does not matter. The reason that some rocket engines are 3D printed is the same, the design does not change, they are a product, but their complexity requires that they be fabricated that way. The advantage people like NASA and SpaceX have is that they can CT scan the finished item to ensure it’s integrity.

            So lay-off the snarky retorts unless you want me to demonstrate how inferior your intellect is.

    2. It’s fairly common in with high end paintball guns to see fiber wrapped HPA tanks running 4500psi…so I would think that wouldn’t be a problem if engineered correctly

  2. Somebody check my math, but the energy stored here is roughly equal to a 22LR as it comes out of the barrel. Also, the geometry is bad… one of the reasons that PVC can be passable is that it is in hoop stress, and we don’t pressurize rectangular sections of it. Check out Roark’s to quantify the relative badness of the selected geometry.

  3. OSHA didn’t record any injuries from PVC pressure vessels because PVC is not used for compressed gas pressure vessels. There is no industry where you would be allowed to use PVC for a compressed gas. As a hobbyist you are on your own for deciding what risks you want to take, but in the workplace you need to work in a way that isn’t going to kill or injure other people.

  4. My understanding is that the main problem with PVC in a pressure vessel is the failure mode: it creates shrapnel that explodes out at high speeds. I have heard that wrapping the PVC in duct tape is enough to contain the shards at pressures seen in most potato cannons/water rocket launchers (say up to 120 psi). Anyone care to chime in on this? Good solution, or bad idea?

      1. I wonder could ultrasound pick them out (with a lot of pain killers). Still the risk of microscopic shards of PVC inside a body does not sounds like an ideal failure mode.

    1. I had the idea of wrapping the PVC pressure vessel with carbon fiber. You’d get better strength, rupture resistance, all the while having a lower weight that a similar metal vessel would have.

        1. Those and most CF dive tanks have a think aluminum lining both for shape & to give fittings something substantial to thread into.
          A now ubiquitous water bottle would serve nicely.

    2. Duct tape certainly isn’t enough, it’ll rip, it’s designed to rip pretty easily too. I’ve built many spudguns before and in the community, newbies would sometimes suggest duct tape but we would always say it’s useless.

      1. Thanks for answering my question nicely. I’m sure as others have commented that carbon fiber is better. And there are better solutions in general for pressure vessels. Having built a water bottle rocket launcher from PVC (filled from a bicycle pump from 15 feet away), I’m trying to figure out just how bad a parent I am :)

        Thankfully, it has never had any problems. But I’ll reconsider using it from now on!

  5. Please, please don’t use PVC for any serious pressure. Spend $16 and buy a real tank. http://www.andymark.com/product-p/am-2477.htm

    I’d guess that the advantage of wrapping pipe in duct tape will be when it explodes and the tiny pieces enter your body, the tape will close the wound long enough for you to get your shredded body to an urgent care center.

    High pressure air isn’t the time to throw safety to the winds.

      1. Well, there you go. @Hackaday @ Benchoff for the sake and safety of people who read this later, but skip the comments section, please consider including a retraction/disclaimer at the top. I don’t think it’s a good idea to suggest that PVC for compressed air is “safe.”

        1. Shrapnel is always going to be a dangerous failure mode. At least with steel there is also a level of plasticity that is more likely to see it peel open in a less dangerous way, most of the time. What if you put the PVC inside a more elastic plastic so that if the PVC fails the outer skin balloons and contains the air enough to slow things down? Perhaps the PVC would just cut it’s way out and not slow down much? Sounds like a lot of testing would be needed to be sure, a Masters Degree worth of work of work probably…

  6. Interesting, could this also add a bit of stiffening to parts while keeping them lighter? I mean using a little bit of over pressure in the voids to put the plastic shell under tensile stress to stiffen it?

      1. It is when it’s so obvious that only a fool would think that the author was being serious. It becomes even more funny when all of the self-righteous jerks in the world come out to shout at how wrong the author is.

        1. ” only a fool would think that the author was being serious. ”

          Without prior knowledge of the issue, it’s very very difficult to detect sarcasm. Don’t assume everyone knows and understands what you do. A fool is always someone who doesn’t know that which you learned five minutes ago.

          This article is just trolling in the same sense as how 4chan trolls instructed people to “grow crystals” by a reaction which actually produces a deadly gas http://i.imgur.com/9XrBc.png

        2. Then I am a fool – and I know not to use PVC for a pressure vessel!

          This is akin to claiming bleach and ammonia combine to an excellent universal cleaner and that all accidents related to the mix is only due to external factors. Or that acetone peroxide is a reliable explosive and associated accidents are due to making it wrong, storing it wrong or handling it wrong.
          Just as bleach and ammonia produces a toxic gas and acetone peroxide is so unreliable that it can explode due to vibrations, it’s own weight, sunlight or due to crystallization (and that includes 100% pure products!) PVC is inherently unreliable as a pressure vessel and have a dangerous failure mode.

    1. I think the issue is that the article is written in such a way that, if you didn’t already know how stupid the suggestion is, it comes across perfectly straight. Like those trolling “life hacks” that have people microwaving their iPhones to charge them, those who know can tell it’s a stupid idea but those who don’t have no clue it’s a joke.

      Admittedly I have little sympathy for people who microwave their iPhones, but this PVC pipe shit can easily lose someone an eyeball, innocent bystanders included.

  7. If weight is a factor for your hobby application, use a PET soda bottle. They are designed to handle pressure. If you need higher pressure in home materials, use steel pipe. Steel doesn’t tend to shatter when damaged and doesn’t get brittle with age/heat/sunlight like PVC.

    If you’re going to use the pressure vessel seriously, invest in a safety valve to blow off overpressure.

  8. Any pressure vessel should/must be contained. One type of container is a net main of thin aircraft cable. An external web does not keep the vessel from exploding (don’t try to use an external web to keep a pressure vessel from exploding!; leave that to the pressure vessel itself.), but it does keep the larger pieces from flying. The `porosity’ of the webbing should not be so small it restricts the expanding air, but the `porosity’ should be small enough to contain larger pieces. I would always be leery of using plastic for a pressure vessel. The suggestion of a copper pipe with end-caps seems a much better solution — particularly for compressed air up to four or five atm. (~100 psi). It’s the stored energy in compressed air that’s the problem in entirely pneumatic systems as well as in air-over-oil hydraulic `accumulators.’ A system with no pressurized gas can still burst, but it far less `dramatic’ than a bursting air vessel. In the case of a crack in a `no-gas’ system you still need to be wary of high pressure liquid — that’s the stuff of a water jet, for example, that can cut an inch or more of steel plate — and many more inches of flesh — depending on the pressure.

    One other comment on using plastic: It does not take a very high temperature to soften plastic. In this case, a plastic accumulator will expand, and with the resulting thinner wall and larger area, will almost certainly burst.

    Everywhere I have said `air’ understand it can be any gas that does not react with fluid and/or other parts of the system.

  9. WTF Brian? I’m not even going to try to be a smart ass here. You are not only posting garbage, but you are acting like an authority on something you know nothing about and giving advice based on your 5 minutes of internet research. Crap like that is going to get someone hurt.
    It’s one thing to promote kickstarter scams and make shit up, it’s another thing to pretend to be an authority and to actively tell people it’s OK to do something dangerous and not to worry about it.

    Your attempt to back up your crap position with sources is bullshit and misleading. It’s also flat out wrong as other posters have pointed out above:
    https://www.osha.gov/pls/imis/accidentsearch.accident_detail?id=200262848
    https://www.osha.gov/pls/imis/accidentsearch.accident_detail?id=201142718
    The second OSHA report clearly states:
    “PVC pipe is an inappropriate material for compressed air lines.”

    If I had to take a wild ass guess as to why there are not more OSHA reports of PVC as a pressure vessel causing more injuries, it’s because… Most people are smart enough to just use the correct tool for the job professionally, and OSHA doesn’t cover some idiot blogger in their garage that blow their fingers off because they read something stupid on Hackaday.

    You are not an engineer. You have a bullshit communications degree. Why don’t you go communicate and find a real materials engineer and get some real numbers on PVC’s pressure rating. You don’t have to speculate. Someone who actually knows what they are talking about has already done that testing and can give you hard numbers.

    Hackaday needs to retract this crap.
    If this was a real respectable tech media site, an editor that posted something of this level of dangerous misinformation would be fired. Seriously.

  10. People! We are talking about 4 bar = 60 PSI! This is NOT high pressure. Even a think coca cola bottle holds more than 90 psi! (cooled the pressure is < 40 psi). Calm down! A several mm thick PVC pipe will not breat at 60 psi.

        1. Easier than you think, because it is… *drumroll* …BRITTLE.

          PVC loses its impact strength sharply as the temperature drops below room temperature. In air pressure vessels, releasing some of the pressure causes the temperature to drop. Add an impact and you got a shrapnel bomb.

          1. Two words: heat capacity. There isn’t much in air (even at 60 psi air is only 5.2 kg/m^3) so you won’t really be able to cool much of the PVC.

          2. @carlos- The issue is the temperature of the environment and the sudden decrease in temperature of the air in the tank as it expands when some air is let out.

  11. couldnt you use steel pipe like what is used for bringing city gas into the house as a pressure tank?

    configure like mythbusters did when they did the seat launcher myth from austin powers

  12. Volume of gas at atmospheric pressure = (cylinder volume) x (cylinder pressure) / (atmospheric pressure)

    so for the 40mm^3 cube, that’s 0.064 litre * 6.5 bar / 1 = 416ml of air at atmospheric pressure

    that’s about a single breath, impressive for such a small plastic box.

  13. not an expect, but I do have a few 1 litre pet bottles pressured up to 45 PSI of air, for a few years now. maybe I should take some pictures and show the process I used to make them. I have seen many a youtube video of people intentionally blowing up pet bottles. there’s one already posted. As I pressurized those bottles (in my hands, not from 20+ feet away, again I was not trying to get them to blow up), I could feel them get warm as they got past 30 PSI. These are not the thin water bottles made today. I wouldn’t trust those with more then 5 or 10 PSI of air. I don’t trust the pet bottles past that 45 PSI mark. as they expand that plastic stretches as you empty them the plastic contracts, that process alone will shorten the life span of those bottles. Steel does the same thing only it will last a whole lot longer. PVC pipe will stretch and contract as well I’m sure and will help bring about that material fatigue. The larger difference of pressure you cycle any material will bring about that fatigue aka failure. But then there are many things that as you cycle them with large swings can bring about the failure quicker (i.e. batteries). It will be interesting to see and something those robotic guys should consider is not just the max PSI their containers can hold but just how many times can they cycle the psi range before the tanks give out.

    1. not an expert* also when the plastic stretches it gets warm very quickly. You can even feel the plastic get thinner as it stretches. I interesting thing is the tone of the bottle also changes. the higher the PSI the higher the note.

    2. 45 psi is nothing for PET. PVC is more brittle, and it’s properties change more with temperature.
      Even the little LEGO pneumatic air tanks (ABS I think) are safety rated at 70 psi.

    3. A friend of mine tested pop bottles to use as cheap portable air reservoirs, reckons they can often withstand 300psi and are good to store up to about 100 (think of a shaken soda bottle, it builds SERIOUS pressure).

      The failure mode is important though – PVC shatters and sends sharp shrapnel flying, PET bottles split/burst but don’t throw shards of death around.

      The engineering in a plastic bottle is amazing, not least because of the global shit-storm of lawsuits that would kick off if the likes of Coke or Pepsi bottles were injuring people.

        1. I have been drinking WHAT!?!

          Soda is so gross. Ssoooo mmuuch sugar! Crap food coloring. Sodium Benzoate (wtf is it?) MOAR SUGAR!!!

          Parents let them drink 2 bottles a day and then wonder why they are bouncing off the walls. Then the doctor prescribes ADHD meds. Uhgh!

  14. Please add a huge warning at the top of this. The materials suggested, even for the pressures suggested, are not suitable/fail-safe. Recall that the American Society of Mechanical Engineers (ASME) was essentially formed because pressure boiler failures were killing people on a nearly daily basis until suitable designs, suitable engineering, and suitable materials were investigated in detail and, in many place, put in place by force of law.

    As a counterclaim on the OSHA data, look at the number of people killed by elephants in industrial applications. Doesn’t mean it is safe to have an elephant in the house.

  15. “Using PVC pipe as a pressure vessel for compressed air can be a fun and enjoyable hobby. It’s safe, too…. Compressed air stored in a PVC pipe is never a proximal cause in any accident”

    I had to gasp when I read that. It’s an awfully broad statement for something that is clearly not broadly true. Any pressure vessel has risk, and the level of risk depends upon the pressures involved, design, and material properties. PVC is not durable against physical impact and UV damage and can fail quite catastrophically, sending shrapnel flying in multiple directions. Sure, at lower pressures the risk is less, but I still wouldn’t want one rupturing anywhere near me.

    Please dont give inexperienced hackers the kind of misinformation and false confidence that will get them hurt.

  16. “Of course no 3D print is perfect, and the third prototype did leak,…”

    “Of course”, did you say “of course”? Hey guys, there are folks who 3D print air screws for their copters. So, in their view, 3D printing has to be perfect, otherwise those copters would be kind of Russian roulette…

  17. It would be interesting to design a few and see how the infills affect the quality.
    I may print some up and hydrostatically test them for science sake.

    Another neat idea would be using it for no/low pressure fluid storage or making up custom fittings+hollow aluminum heatsink extrusions for inexpensive watercooling.

  18. Danger of pressurized PVC pipe bursting aside. I’m I only one that thought of using 3d printed (or metal welded) frame parts as a pressure tank to minimize weight and volume needed? For instance, if you know that your 3d printed leg on hexapod will have low infill but could hold some pressure you need, why not double it as a tank? Or other 3d printed parts? Just connect all with tubes to have a distributed tank.

    It might not be practical or usable at all, but might be something to consider.

    1. The only issue is capacity, I’ve seen off-road guys use bumpers, rock sliders, and roll cages as air reservoirs but when you work out the volume it’s hardly worth it compared to having a small tank.

  19. I just want to comment on the many who suggest to wrap carbon fiber to strengthen the vessel:
    carbon is brittle. so this is not a good idea.
    Use aramid fibers instead, very strong and in case of overload they will endure plastic deformation instead of shatter/break. That is why they are used in bullet-proof vests and in kayaks for example…

  20. Miss-information like this is a good way to get someone killed and a great way to get sued. While I do build pneumatic spud guns they are very unsafe when shocked while under pressure. I had a 3″x36″ SCH 40 pressure vessel leaning on a wall, charged to ~100PSI, it slid down the wall, cracked the reducer on one end which sent the vessel flying like a pneumatic rocked and put a large dent in the side of a corrugated steel barn, got lucky on that one.

  21. “We are routinely using PLA filament and this material has a rather high yield tensile strength. Doing a quick “back of the envelope” calculation indicates wall thickness can be pretty low.”

    They don’t say it, but I’ll bet they’re assuming the tank walls are homogeneous. The 3D-printed tank walls are going to be strong along the plane of printing and weak perpendicular to it. If they really want a custom-shaped tank, best they create a mold of the tank from the 3D-printed part then cast it some other material. That will result in a tank made of a homogeneous material. Then the back-of-envelope calculations they’re doing might hold water, err, air.

    Also: Compressed air can be dangerous. Are you sure you really want to store energy this way?

  22. “OSHA’s Fatality and Catastrophe Investigation Summaries bear this out; there was no industrial or occupational accident recorded in these summaries where a pressure vessel made out of PVC was the cause of any injury or death[3].”

    I know of at least one that was not reported. The mechanics on a ship were playing with pressurized PVC tanks, and they accidentally over-pressurized it. The tank exploded and hit several people with shrapnel, and the guy holding the tank ended up with a broken arm. OSHA never found out because they and the captain faked an accident and destroyed the evidence, but PVC can definitely explode and harm people.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s