A Cycle of Misery: The business of building commercial aircraft | Hacker News Reader

A Cycle of Misery: The business of building commercial aircraft

construction-physics.com

68 pointsHR0164 comments

“ In the late 1960s, Lockheed and Mcdonnell-Douglas drove each other to ruin by each developing a new aircraft (the L-1011 and the DC-10, respectively) for what turned out to be a very small market: less than 700 aircraft were sold between them. Lockheed ended up abandoning the commercial aircraft market after losing $2.5 billion on the program (~$9 billion in 2023 dollars), and McDonnell-Douglas never recovered, ultimately selling itself to Boeing in 1997 as its market share declined and it was unable to fund the development of new aircraft.”

I don’t think anyone who worked in the industry would attribute the 1997 merger to the DC-10 some 30-40 years earlier. First off, the DC-10 is, as the name implies, a Douglas aircraft. That fiasco did indeed cause a merger, to McDonnell… in 1967.

Also, if McDonnell-Douglas “never recovered”, how did it afford to fund in the 80s and early 90s, the enormously expensive design and development of the C17, entirely with its own money? An effort so expensive no aerospace company ever tried to do it again?

The cause of the 1997 merger was if anything the loss of the JSF contract in 1996 (which became the F35) , which was made necessary by the preceding loss of the ATF contract (which became the F22) and slow sales of the MD-11; most analyses of the 1997 merger focus on the MDC-Boeing rival in commercial airliners, but MDC would have lived through the MD-11 and the merger would have never happened in 1997 if it hadn’t lost two generational fighter contracts. At least, the merger wouldn’t have happened before 9/11.

Missing in this analysis is the MD-95 (now renamed as Boeing 717). McD did a bunch of expensive reengineering and the gamble didn’t pay off when the launch customer was involved in a crash that eventually sank them.

The MD-95 didn’t even get type certification until 1999. Are you thinking of ValueJet? That was in 1997 and wasn’t a MD-95 but a DC-9, and the cause of the crash was, you know, it was ValueJet.

MDC in 1995 and 1996, when the talks were occurring, was not in dire financial shape. The C17 had a contract finally and was printing obscene amounts of money. But there was nothing new in the pipe, so once they bombed on JSF, it was either merge then at the top or ride the trainwreck for the next 30 years.

And Boeing made out, because they picked up the desperately needed cash from C17 and the military business. Which was the right move, since there have been 2 or 3 times since then where commercial nearly bankrupted the company.

PS I agree the MD95 sucked and was a poor decision. My understanding is that a new design was considered too expensive at the time (the 80s), hence the decision to go with the derivative. It can work, but if it backfires you end up with the MD-95 or the MAX.

The first table shows staggering loss after staggering loss yet both Airbus and Boeing are still around. They are large, diversified companies with loads of defense business and all that implies.

The table also shows cost estimating is nuts in both the private and public sectors. Any honestly costed program will not make it past a Board of Directors or Congress. But something has to go forward and so overoptimism makes some careers in the near-term and kills others in the long-term.

They are also considered critical companies for national security / capability reasons and neither will ever be allowed to fail by their respective governments, unless a local replacement exists.

as much as this place is a temple of the private sector;

most old and big companies exist because of public money

Couldn't agree more. Let's democratize them!

They already have been. Tell me what other industry has as much oversight. The FFA exist with the sole intent of regulating anything to do with air travel. Imagine if the USDA decided that McDonalds had to stop selling Big Mac's because they found 1 that contained listeria, and had to inspect Millions of pounds of meat prior to letting them resume sale. Car recalls exist, but you don't see us stop driving when brakes fail or hackers can hijack your CAN system.

That's not what democratizing means. It doesn't refer to regulation or oversight, but rather to worker ownership and organisation of the group or business in question. Collectively owned / horizontally managed businesses do exist at scale, from farm cooperatives to the Brazilian industrial giant SEMCO.

You mean like when they basically shut down Blue Bell for listeria? That is exactly what the FDA and CDC did.

https://www.cdc.gov/listeria/outbreaks/ice-cream-03-15/index...

What does this mean?

It means that ownership is democratized. I.e. either owned by the workers or by the stakeholders/users.

It's a fascinating read, and makes a lot of sense why it's so hard to develop a commercial aircraft that actually ends up being profitable. One thing that got me wondering, and maybe someone here might know this, is why hasn't Boeing attempted to miniaturize the 787 as opposed to continually re-purposing the 737? From the outsider's (and complete layman's) view it seems like this would be a more sensible way to build a successful modern aircraft in that vein as opposed to trying to repurpose a design which is now well over half a century old.

I fly a fair amount an honestly the 787 is an impressive aircraft. I'm surprised that Boeing hasn't tried to take what they learned from their developments there and apply them in a way that could potentially help them recoup the costs of that program.

I’m not 100% sure but I think there are some fundamental differences between wide-body and narrow-body aircraft (things aren’t just a different size but different shapes) that mean you couldn’t just shrink it down and change the seat pattern, but I expect they definitely could base a new aircraft on the same materials tech, avionics, etc. as the 787.

The article mentions that at the time they were looking at the 737 replacement in the early 2000s they didn't think they could scale up the production to the volume needed for the 737's market. I think this fits the general theme of the article - they probably could if they started work now, but they had to make the decision on replacing the 737 or creating the 737 Max whilst guessing on the future years down the line.

Scaling up and down aircraft is usually done by changing their length since that's both structurally very straight forward, and it can usually be done with relatively benign aerodynamic impacts.

The 787 is significantly larger than the 737. It's not just built to have more passengers, but also to carry way more fuel so it can go more than twice as far. While you can always take off with less fuel to stave off the worst of the penalty when making shorter flights, a penalty remains.

Scaling down a 787 to hit the 737's operating niche likely means:

* Reducing the body diameter/width * Shrinking the wings (the 787 wing has 3x the area of the 737 wing) * Reducing the length

The first two operations are really non-trivial. Certainly not impossible, but challenging enough that calling it a "787 redesign/miniaturization" vs "clean sheet model with 787 heritage" gets really blurry.

The weight and efficiency benefits of current composites vs aluminum lithium (AlLi) frames are not enough to justify their costs.

Also wide-bodies are just different from narrow bodies in length or diameter. Narrow bodies are designed to go through more and frequent pressurization cycles, fit and weigh enough for certain gates and runways, carry different amounts of cargo.

The other variable is the cost to build the plane you describe. The reason Boeing decided to rengine the 737 for the NG instead of the 757 is the 737 costs less to build and operate. The 737 MAX 8-200 and the MAX 10 are very economical to fly on a level a shrunken 787 couldn't reach.

It's also important to remember the 737 Max was kind of a stop gap on the higher capacity variants for the NMA. If Boeing had been willing to give the MAX a slightly different type rating and difference training for the MAX, MCAS would not have been necessary. Then the MAX 9 and 10 could be replaced by the NMA and bought Embraer with stretched E2 jets replacing the MAX 7 and 8 if executives were concerned with more than their annual stock comp.

> If Boeing had been willing to give the MAX a slightly different type rating and difference training for the MAX, MCAS would not have been necessary.

I see this a lot, but MCAS would still be needed to comply with the regulations on control column force curves so I don't think Boeing could have just gotten a different type rating and done that.

The problem is the airlines. They can't be bothered to do any kind of retraining for their pilots. They'll just buy another brand if this is needed. Which is why the whole MCAS thing happened.

Even 787 pilots will have to be retrained if they scale it down to the size of a 737.

Boeing could recycle some of the core 787 technology to design a new single-aisle airliner. But the concern is that by the time it is certified it might already be obsolete. They are in a holding pattern now, waiting for research results on truss-braced wings. If those work then they could significantly reduce fuel costs.

https://www.seattletimes.com/business/boeing-aerospace/boein...

The 737 and 787 are fundamentally different designs. The 737 is made to be small enough to deplane using stairs to the tarmac at smaller airports while the 787 is exclusively for modern airports with walkways. The 737 is about 12 meters tall and the 787 is 17 meters tall as a result.

That was the driving factor behind the 737 Max disaster. The most efficient engines are high bypass turbofans which are getting taller and taller to fit the main rotor so they had to place the engine nacelles in front of the wings rather than below in order to fit them. This changed the flight envelope and necessitated retraining which Boeing avoided by implementing MCAS which caused the crashes.

This is an argument for smaller airplanes. There will never be an A380 again but also the larger twin isles may be be an endangered species.

The A321 XLR seems to be a harbinger of the future, smaller airplanes serving more city pairs point to point. Ultimately lower cost airlines will win out because flying is largely an undifferentiated product and connecting flights are just an opportunity to be left stranded somewhere. They add complexity and cost.

But in this scenario there is still the need for longer flights in larger planes and we're seeing that in A350s doing pairs like Perth-London or Singapore-New York. Even those are more economically marginal with the larger planes and making something somewhat smaller with a very long range makes more sense.

> Ultimately lower cost airlines will win out because flying is largely an undifferentiated product and connecting flights are just an opportunity to be left stranded somewhere. They add complexity and cost.

MentourPilot made the exact opposite argument recently. The low-cost airlines are being disproportionately negatively affected by current market conditions. Some of those are due to a fluke, and some due to the low-cost carriers buying the wrong size aircraft, and some of it is due to the big carriers intruding into their market.

https://www.youtube.com/watch?v=6_nKC7HxyOk&t=9m00s

I think aviation as a whole is ripe for massive disruption. I think the end result will be much smaller planes that do point to point but the exact 'how will it happen' is still an unknown. This will force regulatory changes too since a lot more smaller = totally different airspace management and cockpit requirements (please allow autonomous!). My personal bet for the disruptor to do it is is on electric. Look at the graph of battery capacity/kg/year and it is clear that electric will out-pace JET-A in the not distant future. The benefits are massive when that happens since (as the article points out) developing engines is a huge part of the challenge. Electric engines are going to be vastly easier to design and certify and provide massive new options since they can be mounted in new ways making aircraft more efficient and cheaper to build and operate. They will also likely change the altitudes and speeds that aircraft fly at since the same issues that impact turbofan engines don't impact electric (there are other issues obviously though).

Only someone unversed in the aerospace industry could make such a claim. The barrier to entry is vast. It isnt like other industries where you can break things and rapidly innovate. Ive seen first hand how hard it is to translate innovation onto the factory floor. For example there is innovation on AI or more accurately neural nets for the complex software governing landing gears that is functionally impossible to implement. What already exists, is essentially a million line IF statement governing every atmospheric condition and different angle gradient. The trick is getting the new innovation past the regulator in such a way that it is demonstrably more safe.

What currently exists, presumably took decades to write and improve and tweak for every possible scenario and is extremely robust. It is also proprietary vital software out of control of both Airbus and Boeing. Meaning it is both expensive and creates a huge amount of leverage on the part of the supplier as no aircraft can be delivered without a landing gear and there is a very limited amount of companies who can produce it.

While the neural nets perform better in lab scenarios it is probable the innovation will never be permitted to be deployed as it is extremely difficult to demonstrate if it is superior and more crucially: safer, than the tried and tested horreondous milion line statement. It could take 5-10 years of further testing and its only objective may be to just have some leverage in negotations (that both parties will know is very weak ammunition).

The second bottleneck is you need a huge ecosystem of suppliers (much like the electric's ecosystem in Shenzhen) which can rapidly and safely fulfill purchase orders. The manufacturing is about the highest most precise manufacturing there is. Just to train someone to install parts on the assembly line can take 12-18 months. Also some suppliers such as CFMA, Rolls royce, P&W, GE will never entertain a small start up because why would they?

They plan production in terms of 5-10 years and they need deep commitments to satisfy their creditors and their workforce. Can a startup sign a 10-15 year service agreement that guarantees revenue for maintenance? Can a start up guarantee other types of corporate agreements on spare parts or engineering workshops or the other thousand myriad different exchanges of information.

These plants produce engines on maybe a monthly or quarterly basis. The parts inside the engines are protected on a national security basis as they are so difficult to manufacture. The engineering IP is a genuine national security risk as it can be translated into jet fighters.

Aerospace industry is about as open to disruption as big pharma is. Its surrounded by arcane redtape such as Airworthiness directives that are designed to make sure no one dies.

Where you could see innovation is electric or hydrogen for short haul flights on smaller planes. The problem is the batteries and hydrogen especially, can be more dangerous than the current kerosene. And again only shorthaul. For long-haul it is likely kerosene is never replaced bar some unforseen revolution in materials science not seen in 50-100 years. You cannot fudge the physics of weight, density and speed versus power output.

The MTOW (Max take of weight) for batteries is vastly higher than kerosene and the energy is used up far faster. The aerospace manufacturers are researching green jet fuel which could be a good way to reduce emissions.

The regulators will never slacken on the rigidity of the regulation. It is so easy to make a minor mistake in manufacturing that can result in a mass casuality event.

In fact for every new engineer that joins Airbus, they are first shown a musuem dedicated to all victims of previous disasters to demonstrate what is at stake and how important engineering excellence is.

Very interesting insights!

I think often disruptive innovation in mature industries must come from vertically integrated companies doing relatively small products. You can't get innovation by using all the same components as everyone else and all the same manufacturing tooling. Though you of course still utilize many, just not everything. For example in the car industry you start with some hand built sports cars before scaling. Your key innovation might be in-house or use a non traditional industry supplier. In aircraft industry you could start with UAV:s and then maybe light aviation or business aircraft.

Right now one area which could be a seed for huge future change could be E-VTOL companies. The technologies, processes and culture they have could change the industry a lot in the long term.

The problem with aviation is that if you get your innovation wrong, people die.

With E-VTOL, which is presumably aiming for local-to-local flights, you have the added problem that if you get your innovation wrong, potentially bystanders on the ground die when your flying car falls on them.

It's absolutely not impossible to do it right (i.e. with nobody dying). But if you do that you need to hit existing aerospace engineering safety standards for stuff like multiply redundant flight control systems, traffic routing that avoids extensive flight over densely populated areas, and so on.

And then you discover you're competing with an unexpected combination like, say, robo-taxis feeding a high speed rail station with 220mph trains running every 15 minutes and your business model turns out to be the new Zeppelin, not the DC-3.

For example some E-VTOL systems have six quite independent systems, each with a battery, motor and propeller. I think it helps the design to be more fault tolerant. Helicopters are allowed many things even when they are noisier and don't have similar redundancy.

There was a scheduled helicopter between Helsinki and Tallinn. It was somewhat expensive but not unreasonable. Much faster than a ship of course. It crashed, everyone died. They continued later but stopped eventually. They are complex mechanical solutions with very high maintenance requirements.

The industry is very entrenched in regulation that is not going to change anytime soon. I just find it difficult to imagine any room for disruption with an agency like the FAA having full control over who and what gets to pilot an airplane, but also hates to certify experimental aircraft. Autonomous aircraft are going to be a century away at least due in part to red tape because there is almost no on-ramp to enter this industry.

I am also not hopeful for battery technology as, unlike cars, where we are not weight sensitive, commercial aircraft rely on being able to land lighter than they took off due to fuel burn. In small aircraft, battery storage can work for short runs, but we’re going to need to see some sort of fuel cell system for anything commercially viable.

Innovation can happend in places where FAA has no say. There are a lot of countries using short distance flights where electric airplanes would drive down costs a lot.

E.g. Scotland, some intercity flights in Europe and between a lot of small islands such as Sint Maarten.

When this happends, the FAA will have to follow.

Short inter-city distances over land run into competition from high speed rail. Which I know is not a thing in the Americas, but is all over Europe and can compete with turboprop airliners on point-to-point time. (Remember, railway stations are almost all in the centre of the departure/arrival city and there's no security theatre to delay your boarding: also, service frequency on some high speed inter-city routes is as high as one departure every fifteen minutes, utterly unlike Amtrak.)

Scotland isn't really a market for short distance electric flights unless you're thinking of the Highlands and Islands, where less than 10% of the population is scattered across crinkle-cut fjords.

France/Netherlands to Sint Maarten/St. Martin is, IIRC, served by wide-body airliners because it's both legally part of France itself (at least the St. Martin half of the island) and it's a significant tourist resort.

And if you think the EU and UK have laxer safety/regulatory standards than the FAA, you might want to re-think your position ...

There is an argument for large planes on long-haul flights: the money comes from people who are willing to pay for comfort.

I fly a few times a year across the Atlantic. On those flights, economy class is typically only 30-40% of the cabin. Then there is a small premium economy section, and the remaining half of the plane is dedicated for business class and first class.

Single aisle planes have the same business class seats, adapted for single aisle, but much the same.

I guess this is airline and route specific.

Toronto to Paris has planes that look like this on air Canada:

https://www.seatguru.com/airlines/Air_Canada/Air_Canada_Boei...

(Supposedly seatguru is out of date, just still!)

Yeah I just flew Abu Dhabi to Melbourne with Etihad and that was more or less the seat layout on the flight too.

> Business travelers make up 12% of airline passengers, but they pay higher rates than other customers and are typically twice as lucrative, accounting for as much as 75% of profits.

* https://www.investopedia.com/ask/answers/041315/how-much-rev...

I know that self driving cars are eyerolling to hear about, but highway self-driving with convergent infrastructure (not the drive-me-thru-Taco-Bell kind) is going to come to trucking, for the basic reason that you can drive the trucks overnight at a minimum.

Smaller planes seem mostly designed for shorter flights, but if I can get a better-than-sober-me stats on highways, and I can sleep in a car while it drives me to another city, why would I fly?

- take a lot more stuff with you

- for families it will be insanely cheaper

- if an EV, it will be lower carbon

- I will have a car for last mile/city driving when I get there

- much better seats

I honestly think the 20 year prognosis for <800 mile flights is going to be really bad. 800 miles flying is 2 hours + taxiing + waiting in line + transport to airport + giving yourself enough buffer + land/get luggage + ground transport. That is probably 5-6 hours in many situations.

So a 10-11 hour drive where you don't have to pay attention and can sleep overnight would be my primary choice if I could do it now. Even if there was, say, a 1500 mile trip to florida, if there is someone to visit along the way, I'll drive.

Anyway, the gist of it is that I think large planes or trips with water barriers are what the industry will end up with, I don't anticipate some massive groundswell of air travel demand for shorter flights in the long run.

You've also made the case for railway travel, except the last mile car. But that last mile car is an impediment in a proper (dense) city with proper transit.

Railway travel has schedules you have to adhere, and all the boarding/unboarding overhead, and luggage limitations.

With the high expense of rail infrastructure (e.g. transporation in cities) I always wondered if you simply designated a dedicated paved lane for busses if that would simply be cheaper.

Buying tix online makes rail travel a lot simpler, ektually. But there's a place for buses too, and a crossover point where increased traffic makes rail better long-term.

But long-dx on buses is no fun. Trains rule. Get some shuteye, or watch the world go by.

You used to be able to travel to Florida in your car without having to drive, in a "motorail" that simply carried the whole automobile on the train:

https://en.wikipedia.org/wiki/Auto_Train

That seems to have the humans ride in passenger cars.

The Channel Tunnel vehicle trains generally leave the humans in the cars during the journey: https://en.wikipedia.org/wiki/LeShuttle

That's because the trip is only 30 minutes. The whole hassle of relocating isn't worth it.

Very long range is de facto inefficient because you have to bunker so much fuel.

Intermediate stop operations would be more efficient (stopping midway just to refuel, if not scramble passengers around to/from different destinations)

But then you have more tear and wear on the airframe due to having twice the number of landings and take off.

Indeed! Carry burn about 4% per hour a pilot friend mentioned last week. Flying is more like space flight than driving in this respect.

I think there will be an A380 again, unless you think air traffic will go down, and I don't see any trend of that. Number of runways is a much tougher constraint to expand than plane size, so the segment of point to point occupied by A330/350 and B787/777 will ultimately be also saturated and require bigger planes.

Those runways still serve a ton of 737/A320 flights that have plenty of room to be upgauged to widebodies, and most of those runways could increase capacity further still if FAA/ICAO decide newer tech can safely reduce IFR separation.

Even with runway congestion, a big problem with an A380 successor is that any new commercial jet needs the fuel economy of two engines to have a chance of selling enough to be worth developing. Which in turn means that an A380-sized twin engine plane needs engines with ~50% more thrust than the current world record from a GE9X, which is already approaching material limits...

A new 747-like plane with a partial double deck might eventually make business sense to develop within our lifetimes... But I think a blended wing (lol) is more likely than ever seeing another full double decker like the A380.

A380 also ended up having too much dead weight to be useful for freight.

We will probably see a double decker if only because the Middle Eastern megahub airlines need it for their business models to work.

But intuitively efficiency per passenger should only be the binding metric if you are not capacity constrained. ie it may be economical to fly four-engine planes with nearly twice the number of passengers, than not meeting the demand on a route because it is saturated.

> The 737 was only saved because it was finally being sold for less than its production costs, but it was not expected to pay back its development costs.

Presumably, "more" was intended here rather than "less".

Something that is only touched on in passing is the 737NG delay. It wasn’t “chaos” that caused those delays: Flight testing exposed structural problems with the tail section (cracking) and it had to be rapidly redesigned and rebuilt.

The internal rumor mill laid the feet at poor finite engineering work: The current practitioners looked at all the mass in the tail structure and said “We can do better, those old pencil and paper guys were so dumb”

When we think of centrally planned economies, we think of things like communism. But monopolies and duopolies are also a form of central planning. Without competition, decisions are all made by a select elite (and not the market). That inevitably leads to disaster. For far too long we've let corporations grow uncontrolled like a cancer. It is time to break them up. There are many more disasters like Boeing waiting to happen.

As long as regulation drives the market towards oligopoly, breaking up companies does nothing. What is needed is for conditions to change such that competition becomes viable again. In some cases this will be impossible due to financial constraints.

> Without competition, decisions are all made by a select elite (and not the market).

I'm not disputing that this is true, but in this context the statement is a little weird. The article mentions multiple times that Boing or Airbus misjudged where the market - their airline customers - would go, and took huge losses because of it.

If it was up to a select elite, today, the A380 would allow for massively increased passenger numbers flying out of the big hub airports. Instead, the plane is dead, and Airbus begrudgingly builds what the market wants instead.

I'm not sure how you extracted that from this article.

Breaking those two up into more would just make more struggling companies that would inevitably die after losing so much of their talent and advantages of scale.

To me, the most remarkable thing about the modern commercial aircraft industry is how similar the end results are. Only the keen planespotter can tell apart a A350 from a B787, from the outside or the inside, and the Chinese (COMAC) and Russians (Sukhoi) are not even trying to disrupt, just copy. Where are all the delta wings and other exotic designs used regularly in military aircraft?!

IMHO there are two things with the potential to revolutionize passenger aviation brewing right now. One is the possible return of supersonic flight, with Boom racking up some 150 orders already, although they still have a long way to go until first passenger flight in 2029 (not holding my breath). The other is switching from aviation fuel to electric engines, which is inevitable, although the sheer weight of batteries is a much bigger problem for flight than for cars and it's likely going to take a decade plus until they're competitive for longer distances.

The other designs were ground away by cost, law, scaling, or practicality.

These issues are less of a concern when someone is shooting at you.

It is also common meme that all new cars look the same

https://medium.com/swlh/the-zombie-mobile-b03932ac971d

This is just a signal that the field is mature and the design is near local optimum.

tbh I don't see that much exotic designs in military aircraft either. Something like Rafale, Eurofighter, Gripen, J-10 etc are all notably similar in design (pure delta with canards). Similarly, the new gen fighters also are quite similar to each other visually. X-32 was maybe the only outlier on top of my head, and it never left the prototype stage.

You can sense the first tremors of a shakeup in cars though: the interior of a Tesla Model 3 is very far from conventional, and there's a lot more room for play now on the exterior as well since you no longer need to devote the first third to the internal combination engine.

Front and rear lights seem to have diverged in design recently. They don’t look similar at all when illuminated. It gives some sort of character to the otherwise sameness. To me, half of them look like Angry Birds. Presumably replacing them costs a limb.

Cars look the same because of regulations (crash test, passenger safety and fuel economy) and common supplier in a global supply chain, unlike missiles where aerodynamics really dictates design.

Imagine a battery fire in flight - it would be devastating. However, since any sort of fire in current aircraft is still a very serious event, I'm not sure that would be a retrogression

The big thing about electric I think is you can run fans off a common electrical bus powered by the engines. If you add a battery you have the advantages of faster throttle response and the ability to fly with your engines out or throttled back.

If you hit wind shear while landing you can increase power immediately not in a few seconds. You can take off and land under partial power to reduce noise. If you lose your engines you can probably fly 200 miles just off the battery and glide slope. If you lose an engine on take off you don't have loss of thrust or thrust imbalance. Rarely you could dump all your fuel and land off battery power.

The above are all nice to have.

Reply on news.ycombinator.com