The de Havilland Comet changed aviation for good | Hacker News Reader

The de Havilland Comet changed aviation for good

bbc.com

84 pointsnairteashop60 comments

The corner of a square hole in a flat plate has 4x the stress of the surrounding material. It's called a "stress riser". If you look at a concrete slab with such an inside corner, there's usually a crack there.

Boeing solved the problem by using rounded windows, and riveting on a forged doubler around the window frame. The skin was also organized so a crack will not propagate far. The success of the latter has shown itself in various incidents where an airliner has lost a panel in flight, but went on to land safely.

Real Engineering has a nice explainer on this topic:

https://youtu.be/7rXGRPMD-GQ

> Boeing solved the problem by using rounded windows, and riveting on a forged doubler around the window frame.

Boeing was by no means the first to use radiused windows in pressurised airliners; the Avro Tudor was doing so in 1946 and there may well have been earlier examples.

Conversely the Boeing Stratoliner was flying with square windows in the early 1940s and didn't suffer fatigue failures, nor did the later DC-6. Corner stresses were understood at the time.

What really undid the Comet was that the fuselage skin gauge was insufficient to accommodate the increased rate of change of pressure differential that was introduced with jet flight. Whereas the preceding piston-engined airliners were lucky to hit 1000 feet per minute climbing to 20,000 ft, the Comet climbed at three times that rate to 35,000 ft +.

The Comet had skin reinforced to 20 swg along the window belt, whereas the succeeding Comet IV had 18 swg with 16 swg framing.

Fatigue failure was poorly understood at the time. It still is incompletely understood, but much better models have been constructed to predict it. Fatigue damage goes up as the cube of the load, and the Comet was pressurized at 35,000 feet rather than 20,000 for the Stratoliner, meaning much more load.

I assume I am wrong on this, but surely if it pressurized to 35000 feet, then that would be less load? As they pressurized it to that it would be closer to the air at its altitude?

The cabin pressure inside is 10,000 feet, the air pressure outside is 35,000 feet, making for more of a pressure differential than at 20,000 feet altitude.

> Avro Tudor

Curiously, this did not extend to the cockpit windows:

http://smg.photobucket.com/user/Coridano/media/TB01.jpg.html

suggesting that stress cracking was not the reason they did it. (There are other reasons for round windows - such as portholes on ships which are easier to make a seal.)

The 707 shows a more radiused corner for the cockpit windows.

In the past couple of years I've been updating my CAD skills, one of the most fascinating discoveries for me was how fillets and rounds distribute the stresses that affect sharp corners. First-year engineering stuff, sure, but pretty amazing to learn what a difference even the smallest details can make.

The other problem the Comet had was economic. It could only carry 36 passengers. The 707 could carry 140.

The larger number of passengers meant the 707 per-passenger cost made it cheaper to operate. The 707 could do 5 times the work of a piston powered DC-6 at only double the cost. It was a huge money-maker for airlines.

http://generalatomic.com/jetmakers/chapter7.html

When I read the headline, I was really hoping this article was about the de Havilland DH.88 Comet.[1] That really was a beautiful aeroplane.

[1] https://en.wikipedia.org/wiki/De_Havilland_DH.88

Not sure why the downvotes. I thought this as well. The DH.88 absolutely changed aviation in it's own way. The first fully monocoque, aluminum skinned, twin engine, variable pitch prop aircraft ever produced. We are still making aircraft with this same basic design to this day.

Many of the design elements from the DH.88 can also be found in the DH.98 Mosquito,[1] probably the most successful multi-role aeroplane of WWII.

[1] https://en.wikipedia.org/wiki/De_Havilland_Mosquito

Such a wasted opportunity. Around the same time, the UK was throwing away its leadership in computing too. Weirdly, both ended up in Seattle.

It's hard to keep up when you've lost the better part of two generations in a row, and most of the stockpiled capital of a century of imperialism - especially when you also lose the empire that generated all that wealth.

I've often argued that the root cause for loss of empire, was the best and brightest of a generation was left on Flanders fields. The best and brightest of the ruling class at the time was simple used up as cannon fodder, its also the cause for the decline of the aristocracy too.

And this was why postwar Germany lost its scientific and technical advantage and is now a ruined backwater - wait, no. Germany not only lost technicians and an aristocracy in WW1 and 2 but murdered those of its Jewish population that were not able to escape. Survivors defected to the US like Von Braun.

The loss of empire was after WW2 not WW1, and partly due to the US insisting (Suez) and partly because it became morally and economically impossible to subdue (Mau Mau etc.). The postwar British disease was labour relations, which were handled far better in Germany with worker representatives at board level.

(I might argue far more cynically that the aristocracy was by definition not a meritocracy, and that only its violent destruction by war enabled Europe to have a chance at becoming more egalitarian.)

I agree with your assessment, pjc50. All major European nations lost most of their privileged/gifted young men in WW1 and WW2, not just UK.

>The postwar British disease was labour relations, which were handled far better in Germany with worker representatives at board level.

Interesting, UK went socialist after WWII while Germany went social democratic. I wonder what made the difference. Perhaps part of the answer is that the aristocracy retained more power in the UK? I have read that socialism in the UK was supported by the aristocracy as a way to keep control of the economy, but I am not sure if that is correct. Also, Germany was under US occupation for a number of years, I would suppose that made a difference, since the US is quite anti-socialist.

"Socialist" is almost a useless label these days, it's used to mean too many things that the US doesn't like.

It's true that some of the aristocracy (e.g. Tony Benn) supported Labour and some very left-wing politics. But that was a small part of the whole. What is true is the Fabian Society proposing incrementalist socialism as an antidote to revolutionary socialism - a way for the aristocracy to keep their heads.

The UK nationalised a lot of industries after the war but then failed to run them effectively. After a few decades of this the Thatcherite response was to just privatise everything and give up on trying any form of state management. This leaves us in the bizarre situation where Germany, France, the Netherlands etc have state-owned rail companies running UK railways recieving UK subsidy.

I recommend Len Deighton's well-researched _Blood, Sweat, and Folly_ for a view of the disasterous management of the war effort - that the same people would mis-manage the peace isn't surprising.

Classic case of first mover disadvantage. Their failures helped Boeing learn some tough lessons about metal stress, and Boeing got many aspects of the fuselage of their first jet liner so right they're still used on the latest generation of the 737 which will finish in the top two biggest selling aircraft this year. Of course, us Brits learned tough lessons of a different kind with the Concorde: no matter how brilliant the engineering is, you need a mass market to break even on an aerospace programme.

It's not as if the UK's contribution to Airbus or other parts of the aviation supply chain today is insubstantial though.

Airbus is one of the companies that will be hit hard by brexit. Arguably the single market is what made Airbus possible to begin with.

Airbus is based largely in France and Germany. BAE will likely suffer more as a contractor.

http://www.airbus.com/company/worldwide-presence/airbus-in-u...

ARM is in Seattle?

You might be interested in James Hamilton-Paterson's book, "Empire of the Clouds", which is about the glory days and subsequent decline of Britain's aviation industry.

Amazon link: https://www.amazon.com/Empire-Clouds-Britains-Aircraft-Ruled...

It's been a while since I read it, but one of the factors he talks about is echoed in the article, about the airplane being built "by engineers" -- it's an improvement on being built by marketers, of course, but it's only part of the right answer. Hamilton-Paterson's book talks about British aviation engineers' deafness to important feedback from pilots, and the additional role the British class system played. Engineers were mostly educated people from good schools, and pilots were mostly farm-boys who'd been recruited by a desperate nation during the war and then advanced professionally in their RAF service. When conflicts between them percolated up to managers, who were also educated men from good schools, the engineers tended to carry the day, whether or not they were right.

The portion of the article where the pilots talked about the lack of force-feedback on the new hydraulic controls, leading to several over-rotation accidents on take-off, reminded me forcefully of this -- obviously I don't know about that issue specifically, but it's exactly the kind of thing that might have come up in flight testing where test-pilot concerns might have been ignored.

The other piece of British aviation decline is government priorities -- aviation innovation is expensive and not profitable in the short term, so the UK decided to hedge their bets by becoming junior partners with Europe. British engineering and technology remains excellent, but it's not leading. Instead, the UK focused on financial services.

The contrast with France is particularly striking. Unlike the US, France was similarly ravaged by war, but the French government basically decided that aviation was a national priority, and supported it accordingly. Today, they are senior partners of European aerospace, and even have a large independent jet manufacturer, Dassault.

Peter Duffey first flew during WW2 https://www.buckinghamcovers.com/celebrities/view/127-captai...

he also wrote a book about his flying experiences https://www.amazon.co.uk/Comets-Concordes-Peter-Duffey/dp/18...

My biggest "surprise" always is how little we have progressed if you just look at how the planes look and how fast they fly (6h11m NY->London in 1958, that's pretty much what they do today, not?). It seems like the airline industry got a lot of stuff just right the first time and since then just has been doing performance tweaks (where I take "performance" to be very broad to include things like safety, economy, ...).

It's down to the laws of physics. They still fly at around 80% of the speed of sound, as that is close to the so-called "coffin corner".

https://en.wikipedia.org/wiki/Coffin_corner_(aerodynamics)

It's unrelated to the coffin corner.

Planes in the ~80s were flying NY<>London faster than today -- the fuel savings generated by flying slower are what is driving the decision.

Yeah, you are absolutely right :)

Well the Concorde was twice as fast (record: 2 hours and 52 minutes), but it's decommissioned now.

In 1958, the cost of a ticket was about 4x what it is today (adjusted for inflation but not earnings growth), flying was at least 20x less safe in terms of fatal crashes per departure, and consumed at least 4x more fuel per passenger-mile. Those are some impressive performance tweaks.

Those of us with legs suffer at least 4x more - The price has been at least partly compensated for by cramming more people in.

A first class ticket is about double the cost and has the traditional leg room. If you can live with only half the cost of the 1950s then its an option.

A little slower today, for fuel economy.

Going supersonic uses 3x the fuel. That's what killed the Concorde.

This is the real reason. The 757 could cruise at Mach 0.8, but it's a lot less efficient at that speed.

The 757 use to make the SFO-JFK route in a little under 5 hours, these days most flights make it in just under 6 (a 20% increase in time to save fuel).

Air travel is technically capable of big improvements, but economics don't work out.

Nonstop SFO-JFK flights are 5:35 to 5:48. But this isn't all due to flying slower. Airlines have increased their scheduled flight times to improve their on-time arrival statistics. So if there are no delays then they actually arrive early.

> airline industry got a lot of stuff just right the first time

Not quite the first time: by looking at 1958, the start of the jet age, you have deliberately set as your starting-point the exact point at which they got things right, along that particular dimension (speed). Prop planes such as the DC-3 were most certainly slower.

But it's true that commercial airliners have not got much faster over the past 50 years, and that's, as other replies to your comment point out, down to the speed of sound.

Air friction goes up with velocity squared, so a small decrease in speed gives you a relatively large savings in fuel. It's just too expensive to fly fast.

There's a sweet spot where fuel savings, customer satisfaction, and capital costs meet, and it's at a much lower speed than engineers are capable of producing.

This is common in many areas of technology. There is a phase of rapid technological development, then it reaches the fundamental limits of the technology, and from then on it is all refinements.

For instance take a look at liquid-fueled rockets. The first one was built by Goddard in the 1930's, they advanced rapidly with the German V-2 and then the Soviet and American military and space programs, reaching a peak with the Saturn V that has not been exceeded to this day.

It's indeed a beautiful aeroplane.

But what's striking, and indeed rather UFO-like, are the small engine intakes. Anyone know what's up with that?

Compact turbojet engines rather than high-bypass turbofans like modern airliners. Blended into the wing for less drag compared to underslung wing pods like the contemporary rival 707.

They're turbojet engines, not the turbofans that you see today.

Turbojets have a smaller intake and the entire thing produces thrust. Unlike today's much more efficient turbofans, where a small turbojet turns the larger fans to produce most of the thrust.

OK, I get it. Thanks :)

They are very cool looking.

FWIW, other turbojet airliners had larger intakes. E.g Tu-104 (which, coincidentally, was propped by Comet's fall, becoming the only jet airliner in service for a couple of years):

http://i.imgur.com/L4KLas2.jpg

The Nimrod maritime patrol aircraft was a modified Comet airframe:

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

In the early 2000s, BAe Systems had a project in train to upgrade them to the MRA4 spec, essentially remanufacturing the existing MRA2/MRA3 Nimrods around a new wing structure that could accommodate RR BR700 turbofans, doubling the aircraft's range:

https://en.wikipedia.org/wiki/BAE_Systems_Nimrod_MRA4

(Go look at the photos on the respective wikipedia pages and examine the engine inlets at the wing roots: the MRA4's air intakes are much larger.)

The MRA4 upgrade was at least as radical as the proposal to re-engine USAF B-52 bombers with four GE high-bypass turbofans (as seen in the photograph on the web page below, showing a B-52 flying with the new engine in the inboard starboard pod):

http://aviationweek.com/defense/ge-rolls-pratt-vie-b-52-engi...

... Which also got cancelled; in the case of the B-52 the USAF had so many old engines in mothballs that re-engining the B-52 would cost more than just swapping out old units, and in the case of the Nimrod MRA4 the program ran into nightmarish cost-overruns when it became apparent that the aircraft they were trying to remanufacture were all different — the RAF's Nimrod fleet had been hand-built then patched for half a century and simply couldn't be consistently upgrades.

(Moral of story: upgrading a 50+ year old design is harder than it looks.)

Doh, I never really noticed how small those B-52 engines are. It makes the plane so elegant and insectile.

Following the cancellation, the Defence Secretary Liam Fox used the Nimrod MRA4 procurement as an example of the worst of MOD procurement performance

Ah those innocent days, before the QE carriers and the F35 made that look like mere pocket change.

I thought the consensus was that the construction of the carriers had gone really well.

They were supposed to have finished their sea trials by now.

Also they were supposed to be convertible to cats & traps at the drop of a hat, which would have allowed the UK to procure a non-BAe fighter to fly off them, couldn't have that...

There's a course on edX, Introduction to Aeronautical Engineering, that explains quite nicely why engines have these huge intakes these days:

https://www.youtube.com/watch?v=a2PBOdTk3vE

https://courses.edx.org/courses/course-v1:DelftX+AE1110x+3T2...

There is a "square windows" Comet 1A in the http://www.dehavillandmuseum.co.uk/ , next to the northern stretch of the M25. Well worth a day out to this quiet little museum (home of the former De Havilland factory) if you live in or around London.

It's surprising the designers made that mistake. Square inside corners were a known problem in shipbuilding.[1] Welding Journal, 1945: "Another contributing failure to welded ship failures is faulty design and in this category the inclusion of square corners is by far the most common and most detrimental. Occasionally some form of square corner will slip by the designer and and draftsman ... Supervisors should call it to the attention to the drawing room and request it be changed. ... would go a long way to solving our welded ship problems."

Early models of US Liberty ships had that problem, big-time.[2] One broke in half before it got out of the shipyard. About 20% of the fleet experienced a square-corner fracture. Those were rush-job ships, built quickly with square corners to save construction time. It was well known in naval architecture to avoid that. Portholes are usually round.

[1] https://books.google.com/books?id=P1IfAQAAMAAJ&lpg=PA473&ots... [2] https://metallurgyandmaterials.wordpress.com/2015/12/25/libe...

As I understand it, the problematic corners of the Liberty ship doors were literally square, actual right angles often close to a metal seam. The Comet's were rounded off, just not enough to cope with the stress of the speed and pressure differentials it was subjected to. I'm under the impression (but not certain) the rationale behind portholes being circular is/was more to do with concerns about the fragility of the glass than the surrounding metal.

I like the De Havilland story in how it made the managers look like fools: 2 crashes in bad weather, +1 in good weather, a worldwide suspension of operations, resumed a fortnight later with reassuring words from the management, +1 another crash, then a more serious stress test (which should have been done since the beginning), with negative results, then a bankruptcy... Fortunately, that's what happens to the wrong companies.

Another aircraft story I love is the Tu-144, the copy of the Concorde: "Critical alarms were so routine that pilots had to borrow pillows from passengers to stuff the cockpit siren". The full Wikipedia article is worth reading (or the Mayday episode worth watching), it's a good laugh.

Each time I fly, I internally laugh at how crazy humans are, throwing a can of the most inflammable liquid we can find, at 900km/h. "Play with fire, mum says, ..." and play with fire we do.

Could be worse, at least jet fuel isn't hydrazine.

It is possible to extinguish a small fire with jet fuel. Don't try this at home.

F16's carry hydrazine as emergency fuel to maintain control in case of loss of power.

The British Nimrod, a submarine hunter, as well as other roles, was based on the Comet and was in service until 2012.

Modified Comets apparently flew until 1997.

Does the BBC not even get the history of Britain correct? The aircraft never exploded. It just broke up in mid flight.

Reply on news.ycombinator.com