Hacker News new | past | comments | ask | show | jobs | submit login
U-2: The veteran spyplane too valuable to replace (bbc.com)
192 points by pseudolus on Dec 13, 2020 | hide | past | favorite | 95 comments



I admire anyone who flies the U-2. As the article says, it is extremely easy to get yourself killed. The flight envelope of the U-2 was so narrow that the difference between flow separation from the wing (called a buffeting) and stalling is around 20 to 30 knots. It is, as far as I know, the place from where the term "Coffin Corner" came.

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

From NASA's book on the U-2,

> The U-2 has always been a difficult airplane to fly, challenging even the best pilots. Lightweight construction and long, thin wings give it terrific soaring ability but also make it hard to land. As the airplane approaches the ground, it floats on a cushion of air—known as ground effect—that causes the U-2 to remain airborne even at idle. The pilot almost has to force it down using spoilers to extinguish any remaining lift. A safe touchdown is further complicated by the unusual landing gear configuration. Near or on the ground, the U-2 appears to be a clumsy beast. The mysterious Dragon Lady seems at ease only when soaring to altitude or cruising through the stratosphere. Even there, danger lurks within the thin margin between Mach buffet and stall speed, the dreaded “coffin corner” of the flight envelope. Technological improvements have eased such problems as high-altitude engine flameouts and balky autopilots, but only a select few can hope to admire Earth’s curvature from more than 12 miles above the surface in what has been dubbed the Near-Space Plane.

- https://www.nasa.gov/sites/default/files/atoms/files/unlimit...

It is an extreme craft made for an extreme time.


my grandfather flew one over southeast asia in the 50s, and later was in charge of a team that flew u-2's and a-12's (cia version of sr-71) over china. i transcribed a letter he wrote about it, mostly regarding the mechanical experience of piloting:

https://hyperstition.al/u2-james-black/

he actually mentions a bit about the difficulty of landing:

> The most difficult part of flying the U-2 was to make a good landing. It was a tail-dragger and in order to make a decent landing, you had to come across the threshold at three to five feet and ten knots above the stall speed. And the threshold speed was based on the amount of fuel on board. The aircraft attitude at the threshold had the tail higher than when parked. The residual (idle) RPM was 65% and during the check-outs, only 15% of flaps were used. The trick was to fly down the runway, holding the main wheel one-to-three feet off the runway, while the airspeed dissipated and the attitude of the aircraft rotated from that slightly-tail-high to the parked attitude, without letting the main gear touch the runway. For if the main gear touched too early, you got a bounce and that was a go-around. The same thing is true in a J-3 (40 HP) cub. In other words, you do not make “wheels” landings – placing the main gear on the ground first. It is said, that you can never recover from a bounce. For you will always be 180 degrees out of synchronization. Years later, as a reconnaissance staff officer at the 15th AF Headquarters, March AFB, Calif., I worked for a lieutenant colonel who had been in the RB-47 program at Farber AFB, Kansas for many years and the RB-47 was also a “tail-dragger” and the same was true about it and the bounce. (But it didn’t have a high-aspect wing, such as used on gliders, and the -47 had six engines, so when all six were in idle it settled down very smoothly.)


you might enjoy this story (many pages): https://www.otherhand.org/home-page/area-51-and-other-strang...

the whole site is fantastic, check out his other adventures.


indeed, this is amazing


That is amazing. Thanks for posting it and getting the stories from your grandfather!


fantastic read thank you


The AF uses muscle cars to chase it down the runway as it lands to give the pilot feedback on their height above the concrete. They have to stall it from 2 feet up (61cm) to have a successful landing. If they do it from 4 feet up (1.2m), they'll probably damage something.

They used to use Pontiac GTOs (aka Holdens), then switched to Camaros. Not sure what they're using now.

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


I've been in one of the chase cars as they talk the plane down. One of the most memorable experiences of my life. This was over 10 years ago, but they were using Subaru Imprezas and Mitsubushi Evos then (this was in Cyprus, btw)


Some airman has the most awesome job ever.


Electric vehicles would be perfect for this job.


They tested Tesla in 2017 and Camaro don’t seem replaced. Perhaps those were too slow at high speeds.


One of the problem with Tesla’s is they cannot have long periods of sustained acceleration due to the way battery cooling works. They could probably rig up something fancy, but at least none of the models right now do this easily.


The Porsche Taycan has been specifically engineered to do that sort of thing over and over.The issue there is that I really don't see a reason to get a taycan over some american muscle, a 500hp mustang is like 1/3rd of the price of a base spec Taycan and will do the exact same job. It's not like army bases are short on fuel either.


This is largely a myth and was only true of early Tesla vehicles.

Also you're not going "long periods of sustained acceleration" here.

The real issue would be Tesla's slow acceleration at high speeds. It has a very very good 0-60 mph, but it's 0-100 mph is more average.


Both the cooling and how electric motor works. They lose out power after 60mph or 1/4-mile distance whichever comes first.


Difficult to get it back on the ground without killing yourself too - to the point that the backup pilot drives behind at ~140mph, calling out helpful information [1].

Saw a U-2 at an airshow and somehow I expected more. It had really badly damaged paint and just looked like some neglected museum exhibit that had been left outside too long.

1: https://jalopnik.com/the-140-mph-chase-cars-of-the-u-s-air-f...


I saw a U-2 at an airshow (California Capitol Airshow at Mather) and it also looked beat up. I asked about it and they said that it's a working airplane, not some show piece. Basically it just had a lot of hours on it since it was last repainted.


The trainer versions go in and out of Mather fairly often. I assume they're based at Beale AFB and just use Mather for short training hops.


The U2 I saw at Mather was based a Beale (as were the Air Force folks exhibiting it). They said the U2s get rotated in and out of service/training and that this unit had been all over the world.


Mather AFB has been closed since 1993.


At which point it became Sacramento Mather Airport, where the trainers fly in and out of from time to time.


I've seen a U-2 up-close in-person (it was actually one of NASA's versions). It looked pretty decent. It's pretty basic and bare-bones though: every pound saved off of it equals one foot higher that it can climb.


They should get a lighter pilot.


Coffin corner isn't just a thing at 70,000ft in a U2, it applies to regular airliners at much lower altitudes as well. When heavy the maximum altitude for a normal passenger plane can be several thousand feet below it's certified maximum service ceiling.

What helps is that due to procedures nobody flies manually at such altitudes. So it's up to the autopilot to maintain a specific airspeed.

A similar effect exists at very heavy weight on takeoff, the gap between the stall speed and the maximum speed for the flaps can be as little as 20kt in a fully loaded 777. That requires good concentration and constant monitoring in the climb to not get a stall warning or a flap overspeed. This part is often flown manually.


From the article the maximum speed seems to be limited by structural constraints. But surely a loaded 777 at lift off isn't on the verge of disintegrating bc it's going to fast. Maybe it's a different kind of speed limitation?


Question on the stall issue: if a pilot starts to stall, wouldn’t they simply maneuver to get the nose down until they get enough speed to break out of the stall?


Coffin corner is the speed between stall speed and overspeed, some number close to mach 1 that the airframe cannot handle. A stall-induced dive at that altitude risks hitting a speed that will damage the wings. The U2 is essentially a fighter jet with glider wings: two things that dont want to play well together. It is a frankenstien's monster of an aircraft.


It might be impossible to recover from the resulting dive without exceeding maximum speed in this particular case.


Isn’t Keeping the correct speed a good fit for an autopilot? Plus I would imagine there would be less turbulences and surprises at this altitude.


Modern autopilots yes, old autopilots weren't nearly as good.


This is another entry in my mental file of the many modern day technologies that have merely been iterations of things discovered or created in the mid-50’s to early 60’s. A vague, indefensible notion to be sure, but that era is pretty incredible.


There's this concept of "mature" technologies that have reached more or less the optimum characteristics and are only going to see slight tweaks until requirements or materials change.

eg the assault rifle has seen only one major change since the '40s (the switch to small caliber high velocity in the '60s), transonic passenger and cargo airplane airframe shapes have settled on a single general configuration, &c

The '60s in particular just happened to be a point when lots of aerospace stuff happened to reach maturity, along with our understanding of the common-case aerodynamics.


It's worthwhile pointing out that there have been significant changes to aviation since, most notably the move to twin-engine aircraft. But ultimately that was made possible through small incremental improvements in engine design (and material sciences and manufacturing). There haven't been similar leaps forward since.

The Boeing 747-8 is 29% more fuel efficient than the 747-100 in absolute terms before you even start considering the extra space it has as a larger aircraft. Some of that is aerodynamic refinement (and they're limited by starting with the older airframe!) and some of that is newer engines. And the 747-8 is relatively poor compared with many newer airframe designs.


Deinitely lots of refinements!! But didn’t twin engine aircraft emerge with the DC-2/3 era? Unless I’m completely misunderstanding the term.


Yeah. That’s a helpful concept. I noticed that there are just so many technologies whose essential characteristics were discovered/demonstrated in the late 50’s/early 60’s. And by no means just in aviation.

It’s almost easier to name foundational technologies that weren’t discovered by then than the reverse.


http://www.cadhistory.net/13%20IBM,%20Lockheed%20and%20Dassa...

another likely inflexion point was the introduction of CADAM around the period


We can see this with smartphones as well


My very-low-confidence prediction is that we'll see a burst of experimentation once foldable screens become viable.


And cheap? Because many versions of them are on their second public generation, and not too many people are excited about them.


I was including "economically viable" under "viable". I didn't even realize there were any available on the market yet, and looking right now it looks like they're still at the $2k price point and still have mechanical issues.


Maybe that's because retiring them would make it clear that the US has a viable replacement, and any hypothetical replacement is a secret.


We live in an era where our great minds get to ask, how do I get funding for this technology..? Because the great minds of that generation spent their time asking, is it possible..?


>We live in an era where our great minds get to ask, how do I get funding for this technology..?

It was a spy plane...no asking needed, if you could fly over the USSR and bring that intel back in some hours your sold.

But yeah i miss Skunkwork too especially the A12/SR-71:

https://www.cia.gov/news-information/featured-story-archive/...


Spy technology with military applications still has extremely lavish funding.


It would be great if it was documented on how to get access to that funding or even the problems to be addressed in that space for a potential startup.



And one more that addresses broad research funding priorities rather than specific opportunities. Get used reading these sorts of dry documents if you get involved in this space:

https://www.arl.army.mil/wp-content/uploads/2019/12/arl-baa-...



You are right... WW2 and then the Cold War were major catalyst to a lot of modern innovations...

There have been 4 leaps during the industrial revolution. At the beginning the the new tech is just a toy, then it becomes useful and starts taking over previous/old tech, then it reaches a maturity point where improvements are very small/incremental:

1) Steam - 1800-1880s

2) Electricity and the Combustion Engine 1890 - 1940

3) Jet, rockets and Nuclear era 1941 - 1970

4) Digital Information 1970 - 2012-s

Now, AI/Green Tech/ or just stagnation? Who knows...


I think we may look back at this period in history as a time of innovation in social information and interaction. Surely the telegraph and telephone had some impact but the scale with which people are able to communicate has changed radically in the last 20 years. It's impossible for us to know now but I think we will look back at Facebook or Twitter and wonder "how did that work?" in much the same way we look at a party line today.


If you're interested in planes like this, I found "Skunk Works: A Personal Memoir of My Years at Lockheed" a good read.


Also "Remembering the Dragon Lady", a collection of stories from pilots.


Thanks I hadn't seen that one.


When I was a teenager I was in the Air Training Corps, and I was lucky enough to be selected for a "camp" at RAF Akrotiri.

Several U2 spyplanes are stationed there (along with the Red Arrows during off-season), and we were treated to watching them come and go to operations presumably in the middle-east. Watching them land was incredible as they appear to move incredibly slowly and touch down very gently. They're also amongst the noisiest aircraft I've ever been around.

One of the commanding officers organised a camp photo infront of one of the U2s - we were shuttled outside a hangar with this enormous plane behind us. We were told we were strictly not allowed on the other side of the aircraft that they had "sanitised" for us as not to reveal any state secrets, somehow.

I still have that photo, it's a great one, but it's also technically OFFICIAL-SENSITIVE and we're not allowed to show it to anyone. Our "official" camp photo was infront of the Red Arrows, which would have been cool enough - but our secret photo was cooler.


NASA has a free ebook on the U-2's development: https://www.nasa.gov/connect/ebooks/unlimited_horizons_detai...

It's good reading.


"The only other U-2 operator than CIA, USAF and United Kingdom was the Republic of China (Taiwan), which flew missions mostly over the People's Republic of China (PRC)." from Wikipedia.

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


> while seven U-2s were lost during training missions, killing six pilots

..out of 19. Is this an exceptionally difficult aircraft to learn how to fly?


Yes. Other comments in this thread have pointed out that the airplane has a very narrow window of acceptable speed that's not too fast (overspeed, damages the plane) and too slow (stall, causes the plane to drop out of the sky).


Which makes me think - what was the motivation to be a pilot of one of them, if the risk is so absurdly high? Prestige of being one of the very few certified to fly it? Did it come with better compensation? Or could you be ordered to fly it if you were good enough, no choice given? Or a combination of all of these?


Mostly it is military, so you pass flight school and then get your orders. Sometimes you can ask for a specific assignment, if it isn't in demand and isn't below you, you can get it, but part of asking is knowing who to ask and how to ask for it. (My Uncle was in the Navy and he got a lot of choice assignments because he asked for them and most of his peers just went where they were ordered)


Might also add true believer (if someone doesn't do this the commies will end my way of life). And, from what I've read more likely, "my buddy will die for sure in that thing, at least I have a chance."


Too fast also causes the plane to drop out of the sky, as you start getting supersonic airflow over the wing causing it to lose lift AFAIK.


I've seen one of the crashed U2 in the PLA museum in Beijing.

It's a very interesting museum, well worth a visit.


s/crashed/shot-down


https://m.youtube.com/watch?v=jtsZaDbxCgM&feature=youtu.be

James May rides in a U2. Pretty interesting.


The original reason for BYEMAN, the U2 has outlived every other BYEMAN era program, and even the compartment itself. https://nsarchive2.gwu.edu/NSAEBB/NSAEBB225/doc06.pdf


They fly these out of Beale AFB and occasionally crash them in the surrounding area. In 1996 a civilian on the ground was killed: https://www.latimes.com/archives/la-xpm-1996-08-08-mn-32366-...

Another U2 crash not far from there happened in 2016.


If you are into podcasts there is an episode of Omega Tau with a U-2 Pilot on http://omegataupodcast.net/109-flying-the-u-2-dragon-lady/


It’s still a bit mind-blowing that they got rid of the SR-71 but kept the U-2.


The reasons are straightforward enough, it's about cost:

Now the 31 operational U-2s in the USAF fleet are about to undergo a $50m (£37.8m) update and acquire a new mission which could see them fly on for another 30 more years.

$50 million would just get you a few flights of an SR-71 and its special fleet of dedicated tankers.


The SR-71 was apparently a maintenance and repair nightmare, the per hour fully calculated operating cost was quite ridiculous. The NRO obviously doesn't publish any useful unclassified info for radar or visual, infrared spectrum observation satellites newer than about 1965. But if I had to guess, a lot of its mission became obsolete when post-1980s era NRO satellites went into service.


> But if I had to guess, a lot of its mission became obsolete when post-1980s era NRO satellites when into service.

That is exactly what happened. Also it is a very politically charged event to blast a Satellite out of orbit (resulting shards). You can blast airplanes out of "your" airspace all day long and tell everyone to pound sand.


This website provides a detailed comparison: https://www.blackbirds.net/u2/c_bennett/bbird-08.html

(Also, just gotta admire the "web 1.0" site that is still online after 23 years).


The SR-71 was crazy expensive to fly and maintain and the U-2 apparently is not.


The U-2 doesn’t fly nearly as high or as fast if I recall. But maybe we don’t need that any more.


Also, these planes fly over the US Southwest pretty regularly. If you use a a site like OpenADSB.com you can see them.


I always felt US has to have a replacement that is currently secret. Lately I am wondering if it is truly the case, sure US does have some secret programs but with the mass proliferation of drones and with their spaceplane it's possible they don't need SR-71 replacement.


The SR-71 used a different jet fuel than normal, special engines, and likely required a lot of maintenance.

The SR-71 program was cancelled because satellites were good enough, and the above maintenance plus over a dozen dedicated tanker airplanes (for the special fuel) made them a headache to operate.

One of the most impressive things to me about the program was that after a certain date, all flights were done in radio silence, including multiple refuelings. A TLA told them that the radio calls were being tracked by Russian trawlers around the world.


>The SR-71 used a different jet fuel than normal, special engines, and likely required a lot of maintenance.

The bigger issue is refueling, which required an entire fleet of separate, dedicated tankers, so that fuel wasn't accidently mixed with the regular jet fuel.

The special fuel was some sort of coal slurry, and would immediately do pretty bad things to normal jets. Additionally, the tankers needed special, segregated tanks that didn't to their own fuel supply.

There was also a radioactive additive that could be used to reduce the radar cross section of the exhaust plume, although it's not clear if it was ever used. This supposedly would have been much more of a maintenance issue for engine wear and maintenance.


The special fuel was specially distilled/blended oil. The problem with using it in normal jets is that it is so hard to light that you can't get the engines started without special ignition fluid. This poses the obvious problem for the tankers that you can't get their engines started without special engines, or segregated tanks.


I read the book about Lockheed’s Skunkworks recently, and it touched on this a bit. The big difference between a satellite and a spy plane like the U-2 and the SR-71 was that you can launch a spy plane on short notice and get the intel you need. Satellites aren’t nearly as agile. But things may have changed, perhaps we have enough satellites in enough places that we don’t have as big a problem as we once did.


> The changes to the plane will include better avionics, a touchscreen cockpit (that you can use with a pressure suit)

I hope they thought about this long and hard. The US Navy is reducing touch screen interfaces again after ship crashes in 2017: https://www.bbc.com/news/technology-49319450


Well, it's a great plane. Also an unfortunate, terrible plane - judging by some important characteristics. It's a pity that this is still the state of the art, with those shortcomings still there. Sometimes the progress is surprisingly hard to get.


They needed a tool, they created the tool which did what they wanted, but at high cost in terms of highly skilled and experienced manpower.

Don't we do this all the time in IT thought? Big issues in SW, which we plug with people's knowledge, time and experience? Isn't that what they were/are doing here?

I only have the utmost respect for these pioneers, literally taking their lives in their own hands...


We don't know if this is state of the art. The U-2 is still fine for surveillance over uncontested territory.

For flights over denied territory, nine years ago, Iran captured an RQ-170. There may be other aircraft that have not been captured. I'm sure we'll find out in good time.


I wonder how many such programs never get revealed at all even after completion. I also wonder how many of those would be spectacular in one way or another.


Hard to fight physics.

Especially when you have spy satellites that can digitally send you data, why design a new one.


Materials science has improved tremendously since the 50’s. Surely a carbon fiber structure would get better performance and/or a larger safety envelope?


I reckon Scaled Composites could kick out a new aircraft for less than the cost of running these for a few more years. Probably one of the only companies with the knowledge to do so.


The pilot is just for flying the plane, isn't it? Because keeping the plane airborne is hard enough, I don't think they have time for any other complex operation at all. Why didn't they try to automate the U2? Keeping it in the air maybe be hard for humans but certainly not for machines and algorithms, especially in the times of AI and ML. The ground crew is then freed to do other more important tasks for a spy-plane.


But it's not intended for any other complex operation. It has cameras that take pictures that are later processed. It's basically a spy plane, not a combat plane, where the pilot is expected to evade enemy aircraft or otherwise engage them.


The article kind glosses over why the Global Hawk isn't an adequate replacement. Seems it would check all the boxes?


Myth Busters did a really interesting episode about it. I didn’t love the season this episode was in but it was definitely a stand out in quality and content. It’s the only plane you fly in where you need to wear a damn space suit.


>It’s the only plane you fly in where you need to wear a damn space suit.

Isn't the SR-71 the same?


I suspect the MiG-31 would also require a spacesuit? It would certainly be advisable.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: