Back in March, a small aircraft in the UK lost engine power while coming in for a landing and crashed. The aircraft was a total loss, but thankfully, the pilot suffered only minor injuries. According to the recently released report by the Air Accidents Investigation Branch, we now know a failed 3D printed part is to blame.
The part in question is a plastic air induction elbow — a curved duct that forms part of the engine’s air intake system. The collapsed part you see in the image above had an air filter attached to its front (towards the left in the image), which had detached and fallen off. Heat from the engine caused the part to soften and collapse, which in turn greatly reduced intake airflow, and therefore available power.
While the cause of the incident is evident enough, there are still some unknowns regarding the part itself. The fact that it was 3D printed isn’t an issue. Additive manufacturing is used effectively in the aviation industry all the time, and it seems the owner of the aircraft purchased the part at an airshow in the USA with no reason to believe anything was awry. So what happened?
The part in question is normally made from laminated fiberglass and epoxy, with a glass transition of 84° C. Glass transition is the temperature at which a material begins to soften, and is usually far below the material’s actual melting point.
When a part is heated at or beyond its glass transition, it doesn’t melt but is no longer “solid” in the normal sense, and may not even be able to support its own weight. It’s the reason some folks pack parts in powdered salt to support them before annealing.
The printed part the owner purchased and installed was understood to be made from CF-ABS, or ABS with carbon fiber. ABS has a glass transition of around 100° C, which should have been plenty for this application. However, the investigation tested two samples taken from the failed part and measured the glass temperature at 52.8°C and 54.0°C, respectively. That’s a far cry from what was expected, and led to part failure from the heat of the engine.
The actual composition of the part in question has not been confirmed, but it sure seems likely that whatever it was made from, it wasn’t ABS. The Light Aircraft Association (LAA) plans to circulate an alert to inspectors regarding 3D printed parts, and the possibility they aren’t made from what they claim to be.
That’s crazy honestly
I’m going to err on the side that the seller was just an enthusiast who designed and printed it themselves but they live in a cold place so the problem never manifested during their own use? Or they’re just not very familiar with material properties of 3D printable plastics?
Glad no one was injured too badly
Smart enough to create a very complex 3D model of that, but stupid enough to use PLA in a high heat environment? A 3D model somehow available on the net then printed by an idiot? The report gives no details other than it was purchased at an airshow in the US and Grok despite extensive searching produces nothing more than found is in the accident report.
The Pilot gets a Runner UP Darwin Award.
At this point; I believe this guy should give up building and flying his own planes before he gets Hurt.
His A&P guy is going to get him killed…..
It’s a homebuilt aircraft so the builder/owner isn’t required to use a certified A&P mechanic who is only installing parts that are direct replacement or have a supplemental type certificate (or whatever the UK equivalent of a US STC is called.) But this sort of failure is exactly why certificated aircraft require an A&P and traceability of repair parts. (But there are repair parts out there with forged traceability documentation, and that’s caused plenty of crashes and scandals.)
That glass transition temperature indicates PLA, not ABS with its much higher temperature.
From the report:
“The aircraft owner who installed the modified fuel system stated that the 3D-printed induction elbow was purchased in the USA at an airshow, and he understood from the vendor that it was printed from CF-ABS (carbon fibre – acrylonitrile butadiene styrene) filament material, with a glass transition temperature3 of 105°C.”
CF-PLA filament does exist, but who would have the capability to 3D print such a complex part so professionally based upon its external finish and be stupid enough use PLA in an engine compartment unless it was a mis-marked filament spool?
On second thought about a mismarked filament spool , to successfully print the part with PLA and not ABS, the printer would need to be set for PLA since the print head (hotend) temperature for ABS is much higher than PLA and using PLA at ABS settings would create a mess. So, someone intentionally used PLA.
Maybe they used a printer with NFC tags on the filament spool, causing the printer to use the correct settings for the material? Maybe the part was printed from PLA as a test, and either unknowingly sold as ABS, or the owner of the plane bought this one cheap, knowing it was PLA, but not understanding the implications, or thinking it would be fine because there is a constant flow of cold air through it?
That would prevent the part from overheating, until the throttle is set to idle while the engine compartment is hot.
It could be the sellers fault, the owners fault, or a genuine mistake.
It’s also bad luck. If the engine was a bit hotter, the part would have melt completely (and I guess PLA burns ok in this motor?), freeing the air intake and probably saving the plane.
Maybe I’ve only used dumb printers or something, are there printers that automatically slice a model based on the filament presented?
Slice a model not so much, modify the g-code that tells them the right temperature and extrusion coefficient of the spool they are coming from though isn’t all that strange. NFC style spool ID stuff has been around in 3d Printing for a long long time, but its not all that common.
Yeah, my first printer (it was heavily discounted) had a companion mobile app where you chose or uploaded a model and it would do all the slicing behind the scenes. Mine only “supported” the manufacturer’s PLA, but others have different systems for forcing you to use their first party filaments.
Another huge issue that people fail to understand is that plastics don’t have a glass transition temperature. There isn’t a magical temperature where it goes from a pure solid into something that deforms. It’s a range, and this can start 20 degrees lower then the common listed temperatures.
Which makes PLA pretty useless for anything structural.
It could also be outsourced printing to someone who didn’t even know the purpose of the part, but decided to cut some corners.
50-80 and 80-100° C seem like relatively low margins and temperatures for something that requires structural integrity. Why not use something with much higher tolerance? Weight considerations?
It seems crazy that these things could be pushed to their limits by direct sunlight or something as pedestrian as a hairdryer.
This is quite crazy, I love 3d printing and being “cheapo” but this is the extreme combination of smart and dumb. It reminds me the guy that died making his own rocket to see how the earth was flat.
That was the guy who sat on top of a rocket shaped giant kettle and was grifting the flerfers wasn’t it?
Read this a few days ago, I’m wondering if the part was designed and sent with bad specs to some commercial 3D print service or the service cheaped out on material
Report: “A review of the design of the laminated induction elbow in the Cozy Mk IV plans showed
that it featured a section of thin-walled aluminium tube at the inlet end of the elbow, where
the air filter is attached. The aluminium tube provides a degree of temperature-insensitive
structural support for the inlet end of the elbow. The 3D-printed induction elbow on G-BYLZ
did not include a similar section of aluminium tube at the inlet end.”
So we have a different construction that does not match the idea behind the function of the laminated epoxy part.
Correct! Most cars have a plated steel insert or aluminum insert in the end of the tube, which supports the tube when an air filter is installed or the tube has to make a transition to a silicon boot for ducting and assembly reasons.
A lesson for many in this.
Take great care!
One thing is for sure;
You can tell who does their homework and who doesn’t.
It’s the Pilot’s fault, time to hang it up Dude before you get yourself killed and take someone else with you.
And how are you supposed to non-destructively test a part for your light aircraft in the hanger? If it looks and feels right, which I’d bet it did, and no published specs point to it being flawed…
Really can’t blame the pilot or aircraft maintainer if the part is replaced with a in theory perfectly valid part by specification that happens to be either a dud or intensional cheaping out by the vendor/manufacturer. Its only their fault if they knowingly fit a substandard part, which might be the case, but doesn’t seem to be.
“”The part in question is normally made from laminated fiberglass and epoxy, with a glass transition of 84° C. “”
There is a big difference between the glass reinforced part and one printed out of crap plastic.
All I have to say; you have to be highly inexperienced to not be able to tell the difference.
It is still the pilot’s fault, as it seems he maintained the plain himself.
Just because there is a way that part has often been made doesn’t make a functionally valid substitution and invalid choice.
Which ABS-CF filament construction should be – might well actually be better than the ‘normal’ construction for the job, not like you are replacing that classic car’s asbestos brake parts with fresh asbestos….
Unless of course for regulatory reasons you must use 1:1 certified parts of something.
This is why so many of us in the comment section get irritated about misuse of 3d printing and especially PLA. Of course we’re usually overturning our chicken tendies over some fool ground-based contraption that has minimal safety implications, but we’re imagining “what if our engineering mattered?” when we get so excited about poor choices.
I understand doing it on your own airplane. If you have an experimental certificate. This looks like a Rutan kit plane so i imagine it qualifies. But printing something for resale for someone else’s airplane! That’s ballsy and clearly wrong. The liability is epic. There’s an unambiguous tort claim here for tens of thousands of dollars. I’d be curious to know where the mistake was made but once you’re offering it for resale, you own that mistake.
The Long EZ has a poor safety record even without this sort of thing, RIP John Denver, among others. This was a pretty minor crash, the pilot was lucky. 3d printed parts in an experimental airplane are technically fine, but in the engine compartment? For something as important as this I don’t know man. That was definitely the wrong resin. I’d say unless you do testing on every single roll of filament, maybe even the the seller of the thing didn’t know what it was made out of.
The report has lots of interesting information, worth a read:
This was not a Long EZ but a Cozy Mk IV airplane (a 4-seater version of the 2-seater Long EZ).
The elbow in question was part of an overall modification done to replace the engine’s throttle body fuel injector with a mechanical fuel (meaning, not part of the original design)
The pilot/builder had purchased the part at an airshow, and understood from the vendor that the part was printed in ABS-CF
It’s been a long time since I read John Denver’s accident report but it was not a result of an inherent design fault of the Long EZ, but a combination of multiple issues (I’m quoting from memory here):
That Long EZ was fitted with a non-standard fuel tank switch mechanism (positioned in a different location). I think also the airplane as-built was lacking fuel level sensors and instead relied on some see-through viewports into the fuel tanks.
John Denver was not truly familiar with the airplane (I believe that was his first trip on it), including fuel consumption, emergency procedures, etc
As a result, he wrongly believed he had enough fuel and decided to not load additional gas in his last stop
Eventually, he run out of gas over the ocean, and when tried to switch to the other fuel tank, found that it was also empty
Yeah the holistic approach to safety means we aren’t just looking at “design flaws”, but systemic flaws. The vari-ez / long-ez / variants are numerous enough that pilots are getting ahold of them who aren’t really well-positioned to operate an experimental aircraft. And they are each a little bit different from eachother so there’s little “type familiarity” among the scattered pilot population. These basically social and economic factors are what killed Denver, and are an inherent safety flaw in popular kit-built aircraft even if they aren’t a “design flaw.” Sometimes, problems emerge that just can’t be solved by the designer.
There’s a lot of room for debate over terminology but there’s no denying that popular kit airplanes in general have a lousey safety record, and the EZ planes (being among the most popular) have a significant number of mishaps.
The Rutan designs have an excellent safety record compared to many other homebuilt designs.
The big failure mode is not following plans, which you are allowed to do. It is experimental aviation, it says so on the airplane!
There is a safety margin in the design because not all builders are perfect.
For my authority check my email address. Yes, I built one of these planes. Every time I decorated from the plans, I asked myself “how might this kill me”.
It looks like PLA. ABS should’ve stood up fine, I print ABS engine parts all the time.
That’s insane that anyone would put PLA near any major heat source. Its predictable that it would soften and fail
This is the issue with hobbyists and uncontrolled modifications. In civil passenger carrying engine applications we are required to certify the components with endurance and cyclic testing. This isnt the case in this type of aviation and a blase attitude to BoM control and conformity means these accidents are just waiting to happen.
Proof print for fit. They probably printed it in a similar color as their production prints and somebody packaged and sold it as a finished part.
Never do that. They should be using a high contrast color for prototype/fit parts, and only use production colors for production parts.
(And it wouldn’t be hard to embed some 3D elements to tag such prints.)
I am sure that, as with counterfit money, after suposedly handling “proper” 3D printed parts, you should be able to feel that this PLA test is different (weight, texture, the way it flexes and sounds, etc).
If its CF filled I doubt you’d be able to tell the difference without real lab tools without a both materials in hand at the same time – printer setting, print geometry etc all make such a different to the texture and density of a part that its only with the cad model and slicer settings to calculate the correct mass or that direct A-B comparison you’d have much hope of feeling the difference between stiff CF filled slightly textured plastics.
If that exact same part from exactly the same supplier is noticeably different from a ‘genuine’ version of the part with experience maybe, but nobody is handling hundreds of ‘identical’ parts for an aircraft like this one…
I mean, the easiest way to check is to heat it up in a controlled environment to 85-90C. If it starts getting melty and collapsing on you, it’s not CF filled ABS.
Has no one mentioned the filament supplier? ABS-CF is $50 to $120 per kg while PLA is $20 to $25 per kg. Very tempting for filament suppliers to mislabel and get a nice premium.
However, shame on both the part printer/provider and the buyer to not verify.
Assuming that this was the cause.
Lots of other possibilities too, including an extremely overheated engine, unidentified damage to the part affecting the CF, etc. pretty sure if you compress ABS-CF in the wrong way, you lose fibre integrity.
Just not enough to know but honestly, buy a part at a show and put it in your dishwasher? Sure. In your car? Maybe. In your plane? Fugedabotit.
It’d be hard for me to put a 3d printed part in the dishwasher either, since i know it gets hot and if it fails it’ll potentially flood my house. Lower stakes than an airplane but i’m still paranoid about it. :)
I’m not sure if you’ve ever 3D printed anything but ABS wound print with PLA settings and vice versa. It’ll quickly be apparent something is wrong a little way though the print.
And where else would you find the the right sort of parts for your odd hobby/device but at a gathering filled with the right experienced white good/classic car/aircraft people – if you bought a part for your aircraft off me at a model railway show you are fool, or know I’m an aero engineer or something and this was just the easy place to make the exchange – better be a darn good reason to trust it. But if you buy that same part from me, likely with my own aircraft with similar/identical part fitted and even flying at a meeting of aviators….
Worried about a new part, especially if its a new part construction method I can get, but at some point you have to use what you can get and trust the vendors to actually sell you what they claim – not like even the biggest names in the industry never make a mistake, or end up without knock off lower grade copies of parts entering the chain somewhere… Can’t trust it just because it says Boeing on the box!
This is a fine attitude for you to have and i don’t mind much that you might fly over my house someday. There is luckily a lot of empty space around me that you’re more likely to crash into than my small house!
But this attitude is frankly wrong and completely incompatible with good airmanship. “Where else would I get this part?” is fundamentally not part of the safe aviation formula. If you have the choice between getting the part at the wrong place, and not getting the part at all, YOU DO NOT GET THE PART AND YOU DO NOT FLY THE AIRPLANE. This is drilled into everyone at flight training and licensing. Every pilot takes an oath to decide not to fly if he cannot fly safely.
Sure, they violate this principle all the time. But it’s the only correct principle according to the FAA. Taking off is optional, landing isn’t.
I see a lot of back and forth if it’s the supplier or the pilot’s fault. Here’s the biggest question in my mind: Is there a better way besides “heat until failure & note temperature” to distinguish between plastic types? Because walking up to vendors with a battery powered heat gun (if such a thing even exists) doesn’t seem like a great plan. Because maybe my purchase won’t make my plane fall out the air if it’s wrong, but it’ll ruin my day in another less dramatic way.
FWIW, when it was printing, they could smell the difference between ABS and PLA.
Can always use a battery powered soldering iron like a pinecil or ts100 to do a spot test.
A part that crucial should have been made from cast aluminum or machine made, 3d printed plastic does not belong near a hot engine
My inner engineer screams, “if the original part was made from fiberglass, why abandon reinforced materials for the replacement part when composite 3D printers already exist?” I know they are more expensive, but “amateur hour” has its costs too. I hesitate to suggest more regulation. The experimental aircraft community is made up of serious people who can take care of themselves, but the advent of 3D printing does cry for 1) new self-imposed standards, 2) as-build documentation (e.g. start by writing down specs from the filament reel), and 3) better product testing by “cottage manufacturers” of aircraft parts. This incident might have been prevented by “an extra set of eyeballs” (e.g. peer review by a materials engineer). I am certainly grateful to my own friends who keep me from making stupid mistakes. Fortunately my hobbies would only threaten to burn down my house if I’m not careful enough.
I’ve been prototyping 3D prints since around 2012, it’s become rare for a major brand to sell you straight ABS as they really don’t want returns from failed prints because of how hard it is to print on consumer machines. Much like PETG to PET, seemingly very different but much of the PLA and ABS come with modifiers that can have significant mechanical differences. I have to retest the same filament brand’s ABS every year to make sure its still the same stuff. At this point I can smell the difference between ABS, PC, nylon, etc and when they don’t smell the way they should
I mean, not a critical part but I’ve had 2 GM vehicles (a 1988 Buick Century and my 2012 Cruze) where ridiculously the glue used on the hood ornament is not high temperature enough, they both had the glue melt and the logo insert fall off in hot weather. The 1988 I’m like ‘meh late malaise era’ but I would have thought they used better glue in the intervening 24 years.
I’m just saying apparently even a big manufacturer will use inappropriate materials occasionally.
I often print test parts in PLA because it’s faster and usually less headache. Once I am happy with the fit, I print the final part in ABS. I could see how finals parts may have gotten mixed up with a test part. Very unfortunate situation. Glad nobody was hurt.
You need to print in different colors if you make test parts for critical components. Also devise a theoretically non-destructive way to test, if it does not pass it does not sell.
Well call me old fashioned, If you cannot buy the original OEM part, and have a 3d sketch, bring it to a machinist and ask them to make one from aluminum or stainless, will last forever, you flight on that thing damnit…
I pretty much agree. This would cost a couple hundred if made from curved tube and a metal plate. I don’t know how they certify the welds, but even casting it from aluminum is a great option, you can use a 3d printed plastic to make a sand mold.
I think the terminology of the report and this article is a little confused.
Glass transition temperature in polymers is mostly an academic property and doesn’t mean anything with respect to determining how it will handle heat. Nylon 66, for instance, is more heat resistant than PLA but has a lower glass transition temperature than PLA. What you need to look at is instead Heat Deflection Temperature at some reference pressure. For instance, Nylon 66 has a heat deflection temperature of ~180C at 0.46 MPa but a glass transition temperature of only ~50C. Meanwhile PLA has a higher glass transition of ~60C but a heat deflection at 0.46MPa of only 53C.
This may also be a case of wrong application for 3d printing. From the looks of that part an off the shelf Aluminum or Stainless elbow welded to a laser or plasma cut plate would have worked wonderfully. Aside from FAA certified welds being expensive it would also probably be cheaper.
It’s not that I don’t trust 3D printed parts, it’s just I do not trust them bolted to combustion engine in an automobile, let alone a plane.
Good news, any more of these and the FAA will jump in with certification requirements and we can all go back to using metal. Bare minimum I would expect the material source to be certified and actual heat tests performed on the part. At least tracked by a number so one failure can trigger a recall.
My pilot training was in gliders exclusively, although I’ve flown SEL a bunch with friends.
So every landing I do is a no go-arounds possible landing, and for me (and the whole community) it is perfectly safe. In casual inspection of SEL pilot training, “dead stick” or engine failure training definitely is a thing. Soooo regardless of why some garbage part failed, why did this pilot wreck the aircraft so badly? Can an SEL shed some light on that? Maybe cannot feather prop and huge drag causing controllability issues?