Troublesome advanced engines for Boeing and Airbus jets have disrupted airlines | Hacker News Reader

Troublesome advanced engines for Boeing and Airbus jets have disrupted airlines

seattletimes.com

93 pointswallflower81 comments

I remember watching a documentary about the design of the trent 1000 and being amazed by how small the team was. In fact it was held up where I work as a demonstration of how inefficient we are. But now we see the consequence of penny pinching short sighted know nothing accountants running a company. How much has this cost in contrast to an extra 20 engineers on the team?

I’m not sure adding 20 more engineers guarantees these problems don’t happen.

There is very little substitute for real experience running a design for millions of flight hours, as the article points out.

Makes me wonder if they are developing talent for the next generation or are we gonna have a similar issue as farmers in the US (avg age is like 65)? Albeit the job is more attractive .

What fraction of airplane engines are produced in companies that look like family farms?

If you look at aerospace, the craziest rocket company (RocketLab) looks like it has an average age much less than NASA, with SpaceX in the middle. Which is about what you would expect.

It was a pretty loose comparison and only used it to allude to my question. Mainly was noting how OP said they were a really small team and very experienced. Also discussing jet turbines at nearly century old companies not start ups.

I used a related aerospace field because I don't think any "new" companies produce airplane jet turbines -- the newest ones are joint ventures of older companies. However, the examples do produce rocket engines, whose preburners are turbines.

Farming will be AI based and corporate. It’s self contained with most edge cases well known and insurable.

Well there is little, but there is also some. There's a whole field of accelerated testing of components, by degrading various variables, e.g. increasing temperature, vibration, sounds, corrosive environment, materials, etc., and then using curves of exaggerated failure times to predict failure times under nominal conditions. (after making tons of models and analysis)

You can also improve your simulation (here the sky is the limit).

It really depends on the cost of mishaps and design failures, if it's high enough you can almost always improve design with additional expenditure.

Are you suggesting that Rolls Royce didn't know this?

Well, they didn't matched the resonant frequency of the blades with that of the enfine vibration at high thrust. Probably didn't allocated enough people and resources for even such basic stuff.

It was just a comment on how it is possible to accelerate testing. I wasn't suggesting anything, just discussing and exchanging information. Whether they used this, it's probable, but to what extent only an employee could tell.

The Soviets used to run their new airliners on cargo and mail flights for two or three years before introducing them to passenger service. Great for real-World debugging but imagine the anguish if that was proposed by the FAA...

Mechanically jet engines are pretty simple devices, though - basically just a bunch of connected fans. But when covered with wires and pipes and such, they may look quite complicated at first glance.

What line of business are you in that your team was considered inefficient in comparison?

It's worth noticing that China has been unable to figure out to build high-performance jet engines, to the point that most of their military and civilian engines are imported. The metallurgy is apparently quite difficult, commercial product volumes are high, and customer expectations are even higher.

Yes, the materials issue is usually the root of the problem - weight vs. strength vs. durability. (I expect that manufacturing to the needed tolerances is pretty hard, too.) Modern jet engines are so optimized for these factors, and also for fuel efficiency, that the same general designs are now used for power generation on the ground, where things like weight don't really matter.

This is why AICC bought the An-225. They don't just get the planes, they get the tooling and training to build their own, including the engines.

And at this point, even a multiple-generations-old high-bypass turbofan is a big jump for China's domestic industry. They are not "easy" or "simple" to make and understand once you get down to the metallurgy and the fabrication techniques for stuff like the fan and compressor blades.

Designs for the basic jet engine date back to at least the 1920s, and basic operating principles date back much, much further. Jet aircraft were actually flying by the 1940s, and those engines were built using the relatively primitive metallurgical and manufacturing techniques of the day. So, as I said, _mechanically_ jet engines are relatively simple devices. Materials, manufacturing, control systems, fuel efficiency, noise abatement and all that may be a very different story.

I didn't say otherwise.

Jet engines have always been held back not by their mechanical complexity (they're mechanically pretty simple, as you point out), but by the difficulty of obtaining/fabricating materials that can withstand the temperatures and pressures of operation at the desired weight.

Reciprocating engines were far more complex mechanically, but lasted as long as they did before being supplanted by jet (gas turbine) engines precisely because of how much easier it was to build and maintain them.

Funny you should say that. I always thought that jet engines rapidly replaced piston engines because they were easier and cheaper to build.

For example, the Wikipedia article on the BMW003 says: "The BMW 003 proved cheaper in materials than the company's own 801 radial, RM12,000 to RM40,000, and cheaper than the Junkers Jumo 213 inverted V12 piston engine at RM35,000, but slightly more costly than the competing Junkers Jumo 004's RM10000.[10] Moreover, the 004 needed only 375 hours to complete (including manufacture, assembly, and shipping), compared to 1,400 for the 801." https://en.wikipedia.org/wiki/BMW_003

i.e. the first two german production jet engines were a third of the cost and took less than a third of the man-hours to build compared to piston engines of the same era.

Again, doesn't contradict anything I said.

An iPhone is cheaper to build today than a typical computer of the 1960s. Does that mean it would have been cheap or easy to build a current-model iPhone in the 1960s?

Because that's the argument you're making here, essentially. Once you have the necessary tech under your belt, yes, you get a bunch of benefits from switching. But until you have that tech, you're probably not going to say "oh well, won't build any planes for the next few decades, because the engines we'd use today are way more complex and less useful than the ones we'll have eventually".

Jet engines are simpler than reciprocating engines. I'll happily write that on the blackboard 1000 times, if you'll write "But until you have the fundamental technology for it, you can't build any jet engines" 1000 times.

Also, read your own link:

Completed engines earned a reputation for unreliability; the time between major overhauls (not technically a TBO) was about 50 hours.

Even those early engines, despite being cheaper, and quicker to assemble, still weren't up to the point of being usefully reliable. Jet engines are hard.

> Again, doesn't contradict anything I said.

Yes it does. You said: “Reciprocating engines were far more complex mechanically, but lasted as long as they did before being supplanted by jet (gas turbine) engines precisely because of how much easier it was to build and maintain them.”

I pointed out that they were not easier to build. Jet engines were.

> An iPhone is cheaper to build today than a typical computer of the 1960s.

That’s just being disingenuous. The first production jet engines were cheaper to build than piston engines of the same era.

> you're probably not going to say "oh well, won't build any planes for the next few decades, because the engines we'd use today are way more complex and less useful than the ones we'll have eventually"

It wasn’t a few decades. The first production jet engines obsoleted piston engines in fighter aircraft almost immediately. And in airliners in the 50s, by which time they were already more reliable.

> Jet engines are hard.

Maybe, but we've been building them successfully for 70 years. To get back to the grandparent point, What ‘fundamental technology’ is China missing that makes them unable to build advanced jet engines? Are they really so far behind metallurgy that they can’t?

I pointed out that they were not easier to build. Jet engines were.

What I am saying is there was a point in time, let's call it T, when people first acquired both the knowledge and practical techniques to build jet engines.

What I have said is that, prior to time T, building jet engines was not easier than building reciprocating engines. In fact, prior to time T, building jet engines was generally impossible, because people did not yet possess the knowledge and practical techniques to build jet engines.

Therefore, prior to time T, building reciprocating engines was easier. They could be built using light adaptation of well-understood existing technology, and "can build today" is more practical and easier than "might be able to build in a couple decades", if what you want is airplanes right now. Even though later advances would produce jet engines, and those engines would be mechanically simpler, easier, every adjective you'd like to throw at them, those advances had not yet happened.

You seem to be arguing that:

* After time T, jet engines are easier, therefore

* Prior to time T, jet engines also were easier, therefore

* There was no time at which jet engines were harder

And you seem to be doing this for no other reason than to try to pedantically show someone up on HN. So I'm going to stop engaging with you now.

> Jet aircraft were actually flying by the 1940s, and those engines were built using the relatively primitive metallurgical and manufacturing techniques of the day

Yes, and the Junkers Jumo 004 jet engine used in the ME 262 "only had a service life of some 10–25 hours" (!) after which time it had to be replaced [0]. Admittedly German industry lacked raw materials to do better metallurgy, but in the 1940s there was a long way to go in terms of designing jet engines

[0] https://en.wikipedia.org/wiki/Junkers_Jumo_004

Jet engines today are also routinely torn down, inspected, and rebuilt if necessary - much more often than piston engines are, I gather. But that's because of the overall materials stress and the necessary safety factors involved, not because of any general mechanical complexity. I mean, I guess you can count having all of those individual blades as being "complex", but in reality these are just the parts that make up mechanically simple fans and compressors and such.

True and jet engines didn't really become viable until the discovery/invention of the inconel alloy that could operate at very high temperatures.

At anything more than a superficial glance,they are extremely complicated. You don't get the raw power, power-to weight ratio, fuel efficiency and reliability - and yes, they are extremely reliable in general, despite these recent troubles - by just connecting a bunch of fans. In fact, you would not get anything at all from trying it, except maybe set your garage on fire, if your 'just a bunch of connected fans' comment is an indicator of the depth of your understanding.

You seem to be confusing materials and design issues with mechanical issues. I expect that I have more understanding of the situation than you do - decades' worth, in fact. I've even laid hands on - and spun, as best I could - military jet engines that had been partially dismantled. (I've also seen open civilian jet engines up close but have never laid hands on one.) Once you stripped away all of the external paraphernalia their relative mechanical simplicity was obvious, as was the reason why those particular engines had been grounded.

They're simple conceptually but are highly intricate, complex implementations in reality.

Also, these are not jet engines: they're turbofans and are more than "a series of connected fans".

"a series of fans within a nesting of tubes"

A "high-bypass" design, where a large fan is driven by a relatively small jet engine. Still pretty simple mechanically, though.

Simple concept, yes.

But to say that it's simple to make a turbofan engine would be like saying it's simple to make your own crypto.

After all, it's just multiplying some prime numbers together.

Actually crypto _is_ pretty simple; all you really need is an XOR function. It's in things like key management and such where it gets complicated.

BTW, if I ever _really_ wanted to hide something, I would in fact roll my own crypto, using techniques that no sane crypto expert would probably even consider. Then I would wrap that in at least one layer of "official" crypto, knowing full well that this layer was possibly fundamentally flawed (even crypto experts often don't get things right), or that it might even contain a backdoor.

https://en.wikipedia.org/wiki/Snake_oil_(cryptography)

Specifically that your post is essentially security through obscurity. If your implementation is found then the fact that you use some bogus scheme will not help you.

The "snake oil" here is that security by obscurity isn't a real thing. But it is, because otherwise the concepts of camouflage and out-of-sight, out-of-mind and secrecy in general simply wouldn't work. Only they do to a great extent!

One of the funniest things that I've read involved professional cryptologists examining files where steganography had supposedly been used. At the end of it all they confidently declared that there was simply nothing to be found there - no hidden, encrypted information. And I thought to myself "So you're saying that steganography really works, then!" :)

Like I said, I would choose non-standard techniques that no sane, trained cryptologist would probably even consider. (Ideally they would never even realize that there was anything there to potentially decrypt.) And then, in order to play it safe (and because I'm not an idiot), I would wrap my payload in an officially sanctioned encryption technique, maybe even more than one. Perhaps you missed those parts?

BTW, I would never "pitch" my system to anyone else, because then I would expose its existence. Nobody but me ever needs to know that it exists, much less that there is anything valuable being hidden by it.

Do you have any idea how complicated a single turbine blade is? It's a monocrystal of a superalloy that retains strength at over 1.5k degrees F, and it has channels for coolant because it operates beyond that temperature.

You're like saying a computer is a bunch of wire and some melted sand between some copper.

Huh, interesting. I thought monocrystal blades were only for military engines and civilian engines made do with cheaper stuff which required a slightly lower maximum engine temperature. Then again, I did get my information from a friend in the pub (he’s a Cambridge PhD, but still).

Once again, you're confusing materials and manufacturing complexity for actual mechanical complexity. I remember when I first got a close look at a jet turbine fan blade, and being puzzled by some aspects of it, like the channels that ran all the way through it. (I remember thinking that those channels must have been pretty tricky to make.) But then I realized that these were probably for cooling, like you said. I was most impressed by its relatively complex shape, and by no doubt the exotic material that it was made of and the tight tolerances it held. But in the end it's nothing more than a simple fan blade.

BTW, I had some similar impressions about the close-up look I once got of a large rocket engine nozzle. Then I realized that it was really nothing more than a big coiled cooling tube surrounded by an external shell - pretty simple stuff, actually. And the "de Laval" part of that nozzle, which is what makes it work to begin with, is as simple as can be and is a design which dates back to over a century ago.

I know that it's a big container holding a bunch of fan blades on the same rod that push air around at different speeds.

But from what I found, the Trent 1000 (Rolls Royce engine for Boeing Dreamliner/787) cost $8B to develop and has 30,000 components in it. I just don't agree with you that the elegance of it's mechanism of action means that it's not insanely complicated in its implementation.

And that component count no doubt includes very basic things like rivets and nuts and bolts and wires and connectors and so on. Maybe even the individual balls in the ball bearings, if they still even use such things. Probably also the individual electrical components in the control circuits and such. And how many individual fan blades are in them these days - hundreds, thousands maybe?

Note the attached link; strip away all of the external pipes and wires and such (all part of the component count, no doubt), and what you're left with is mechanically relatively simple. A challenge to build and maintain, maybe, with tons of fan blades, but still pretty simple.

https://en.wikipedia.org/wiki/Rolls-Royce_Trent_1000#/media/...

"In theory there is no difference between theory and practice. In practice there is."

But when covered with wires and pipes and such, they may look quite complicated at first glance.

This is like saying programming is just a bunch of 1’s and 0’s and it only look complicated because of the random curly brackets and semicolons.

Early computers were indeed programmed using toggle switches and such to input a bunch of 1’s and 0’s, which is all the computer hardware itself understands. Additional layers were added later so that we humans could work with something that we understand, but ultimately all of that still gets translated down to individual bits. So in the end we're still programming with 1’s and 0’s, only we usually just don't realize it.

If you think "penny pinching short sighted know nothing accountants" are the cause of a small team, I think you perhaps don't understand the domain involved.

To borrow a popular phrase, modern jet engines are literally made of black-magic fuckery. The advanced metallurgy and fabrication techniques that go into their design and manufacture are such that even if you wanted to double or triple the size of one of those teams, there may simply not be enough people in the world today who have the requisite background. And it's also not a matter of "oh, just go read these papers and get up to speed"; the engine manufacturers derive most of their advantage from the fact that they can do things nobody else can do, and they don't publish all the deep stuff they know.

Mechanical engineering PhD student here. I'm skeptical that there's a shortage of the engineers needed for this.

I know someone who interviewed at a major jet engine manufacturer. I don't know if he would have got the job or not, but he had no interest in working there after the interview. From what he described the place seemed to be a hybrid of the worst of academia and corporations. Publish or perish, poor job security, bad location, etc. I don't recall what the pay was, but I don't think it was spectacular. I won't be applying there when I graduate. If they really needed more engineers, I think they would be making the place more attractive to work at.

Some representatives of the company visited our research group once after the person I mentioned graduated. I didn't get the impression that the company was short of people in their jet engines division aside from one area they mentioned (and there's no shortage of qualified people in that area at my university). Plus, one of the people who was doing the interview actually said they (partially) transfered out of the jet engines division. I got the impression the jet engines division was downsizing overall.

Also, assuming that required knowledge is not public, not documenting important internal developments is a bad practice. I doubt jet engine manufacturers make this mistake. The company representatives, as I recall, seemed to want someone working specifically in the area they were looking to hire, but they didn't seem to need someone who was familiar with the exact methods or materials they used. Learning on the job is probably part of the job description.

I think I misremembered some details now. I'm not certain that their PhDs have "publish or perish", but they definitely have poor job security. The guy I know interviewed at the company's gas turbine division, not the jet engine division. And I think the company representatives were from the gas turbine division as well. The two are obviously related technologies and I would not be surprised if there was overlap between the divisions. Overall I still don't have reason to believe there's a shortage of qualifies engineers in this general area.

So, you go to big universities, you recruit the output of key Ph.D. programs and you sit them down with the existing team for 3 years. Or 5 years.

If these programs aren't producing the folks you need you set up ones that do.

Unless you are a penny pinching short sighted know nothing accountant.

Rolled Royce have over 50,000 employees directly involved in designing, manufacturing and testing engines. The team to design an engine is well over 500 people and takes more than a decade from concept to first production run. It's by no means a small team or a small undertaking.

Virgin Atlantic were/are still so disrupted that they ended up buying 4 A330-200 aircraft to fill the void in fleet schedules this summer as all of their B787 use affected engines. (For reference they have a fleet of <40 aircraft in the first place)

It’s such a difference in scale that engineering oversights in the design of engine blades leads down a road that brings you to buying whole planes to get around the problem

> safety agencies now limit Dreamliners powered by one of the Trent 1000 engine models under scrutiny to flying no more than 2.3 hours from the nearest airport — down from 5.5 hours previously.

This seems like what these agencies should be doing anyway. Why allow a plane to be 5.5 hours from an emergency landing at all? That's just an insane amount of time if there's a major problem. I absolutely hate it when I'm flying over the middle of pacific. It has always seemed insane to me.

I don't see how an engineer would ever recommend being even 2.3 hours away if it's possible to fly a safer route. There should be regulations requiring airlines to fly the safest reasonable route. If it adds even 25% to the flight, so be it. An abundance of caution seems logical.

Despite the ambiguous text, Dreamliners have two engines. I don’t know the rules, but imagine one engine failing means “go straight to nearest airport on remaining engine”. If you can’t trust that the remaining engine will work for at least as long as the scheduled flight, even a money-maximising-sociopath would recognise bigger problems.

I agree with you 100% why would you be 5 hours from an emergency landing?! That's crazy! I would rather sit on a plane for a few more hours, should anything happen you can land some place in a reasonable amount of time.

You have to know these guys have earned their right to stay 5.5h out of range with a safety record. A longer flight also means more expensive. There is a calculation showing once you increase the price too much people start taking other less-safe transportation which results in more deaths and more GDP lost in terms of business. While this doesn't directly correlate here with pacific flights, I am sure they've done the same calculation for this case too.

This is 'disruption' in the operational sense?

The airlines are having to pull massively CapEx equipment (the price tag on a 787-9 is well over a quarter billion dollars) out of their fleets to have the IPC turbine blades on their engines replaced — using a temporary fix, which will require them to be pulled from service yet again, once the permanent fix is vetted.

Of course they mean "disruption" in the operational sense.

I see no basis on which to infer the Silly Valley sense of the term here.

> Silly Valley

I hope that typo was intentional, because it’s accurate.

It is. I don't remember where I picked it up, but I've been using it for years.

Love it, thanks for sharing.

Sillycon Valley is how I have long referred to it.

Shamefully my default assumption was this would be about something good. The SV speak had almost made me forget the original meaning of the word isn’t always positive.

The SV speak had almost made me forget the original meaning of the word isn’t always positive

It isn’t positive in that sense either, given that it really means evading taxes and regulations.

Interesting article but wow what a power pig that page is.

This stupid thing American journalists do with avoiding the word "and" in headlines really makes them difficult to read sometimes.

s/journalists/headline writers

wow indeed

Certainly, I read the headline in a completely different meaning --- Boeing's engines are having trouble, and airlines are switching to Airbus instead.

The Seattle Times headline had 'and' in it originally, look at the url!

In this case, the way you interpreted the headline is ungrammatical; a comma is not a semi-colon.

Alright, we've expanded the headline.

Here's a light-hearted comparison between the airline and PC world:

Motherboard/integration - Boeing/Airbus - a bunch of Taiwanese conglomerates, but the core design comes from Intel

OS - Boeing/Airbus - Microsoft/Apple

CPU/Engine - Rolls-Royce/GE - Intel

RAM/fuel - Saudi Aramco/etc - Samsung/SK Hynix/Micron

I don't know if it's useful in any way, but I found it interesting to do this exercise.

Can't read with an ad-blocker enabled.

You need a better ad-blocker. uMatrix/uBlock Origin works perfectly on this page.

Friendlier article: https://www.fliegerfaust.com/airbus-boeing-engines-problems-...

... but truncated, and missing the part about the Philadelphia accident.

Was this article written by a bot? WTF sentences.

Indeed. It's difficult to read this article.

It's interesting how the term "disrupt" has been overloaded these days because the headline gave me a different impression than the article.

Thousands of passengers' flights / holidays have been affected:

https://www.nzherald.co.nz/business/news/article.cfm?c_id=3&...

And that's just one airline - how is that not disruption?

Because it doesn't "revolutionize" an industry by introducing venture capital-funded competitors with a new type of smartphone app.

Clearly its not that advanced if it doesn't work properly. Perhaps they shouldn't try to rush out prototypes that are incomplete before the kinks are worked out.

Rushing out prototypes without working out the kinks sounds like the behaviour of every javascript project ever.

So is this the problem that led to the recent engine failure, and passenger death?

https://www.independent.co.uk/news/world/americas/southwest-...

https://news.ycombinator.com/item?id=16860902

They talk about this in TFA. That was a really old engine.

Reply on news.ycombinator.com