Slashdot Mirror
Dutch Scientist Proposes Circular Runways For Airport Efficiency (curbed.com)
An anonymous reader quotes a report from Fast Company: While airport terminal architecture has a solid history of style and innovation, rarely is a proposal put forth to utterly redesign the runway. But that's precisely the aim of Henk Hesselink, a Dutch scientist working with the Netherlands Aerospace Center. Dubbed the "endless runway," Hesselink's brainchild is a 360-degree landing strip measuring more than two miles in diameter. Since airplanes would be able to approach and take off from any direction around the proposed circle, they wouldn't have to fight against crosswinds. And three planes would be able to take off or land at the same time. Hesselink's team uses flight simulators and computerized calculations to test the unconventional design, and have determined that round airports would be more efficient than existing layouts. With a central terminal, the airport would only use about a third of the land of the typical airport with the same airplane capacity. And there's an added benefit to those living near airports: Flight paths could be more distributed, and thereby making plane noise more tolerable. BBC produced a video detailing Hesselink's circular runway concept. The concept is fascinating but there are many questions the video does not answer. Phil Derner Jr. from NYC Aviation writes via Business Insider about some of those unanswered questions in his article titled "Why the circular runway concept wouldn't work." The fundamental issues discussed in his report include banked runway issues, curved runway issues, navigation issues, and airspace issues. What do you think of Hesselink's concept? Do you think it is preposterous or shows promise?
A computer doesn't give a shit if the runway is straight or curved, because it can handle a little more left (or whatever) while it's managing dozens of other things. But a human can't do that. You want to make pilots have to account for bank and curvature in addition to everything else? That's obviously a shit idea.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
One of the major issues seems to be what happens when a plane comes in too fast. Straight runways handle that well. It is hard to handle that with circular runways. There are a lot of other safety advantages of the standard setup.
I love how he places his drawning in the middle of nowhere with no roads or train tracks, or even neighby restraints on the layout of the land. Sure you can pull all that underground, but he seems to just ignore it.
My first thought was: how the heck are you going to keep this runway clear of snow? You've gone from a single (or dual) short strip to a (pi*2mi)= 6.28 mile loop. That's a lot of runway to plow.
Then there's the long taxi time from the outside to the terminal in the center. That's a 1 mile radius taxi. Lots of wasted time.
Then there's the poor saps living around the airport. Instead of a well-defined small number of houses with noise pollution, you've spread it all over a huge area. Lots more people to complain. I doubt people want to build houses *inside* that 2 mile loop of land, so the footprint of this beast will be impractical for an airport near anything existing at all.
And if there's a consistent level of wind (from any direction), that "3 at the some time" argument goes away, and you're back to a small strip of usable runway, at least until the wind dies down.
We Dutchies have the best marijuana in the world. Assuming he wasn't actually sober when he came up with the idea.
This is great and all during the day and in clear weather, but constantly changing approach vectors means you can't have any approach lights unless you have a ring of lights around the whole ring that extends at least an additional 1000 ft. So your "3km" footprint for the whole airport is now about 5km. Either that or all landings with less than 1 mile visibility would require autoland. You'd also need to have a system that dynamically turns the lights on/off as approach vectors change. You'd also have to completely redesign the airport charts and the approach vector would already have to be known 30-60 minutes out so that the crew can do a landing briefing which might negate the benefits of using a round runway as the winds could shift in that timeframe so you still have crosswind. You'd have to cross the runway for access to hangars, maintenance facilities, cargo warehouses, etc which most likely wouldn't fit inside the ring. Which means you are losing a quarter of your landing space pretty much constantly to allow aircraft to be towed across, cargo to be delivered to/from flights, etc.
The only thing necessary for evil to triumph is for it to be pitted against a slightly greater evil
It wouldn't cost much to model this in a simulator, and let a couple dozen pilots try it out. We'd find out pretty quick how easy or hard it really is. It would be an interesting semester project for some students at Embry-Riddle.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
For the one benefit of your initial landing possibly being inline with the prevailing wind, you're adding dozens of safety issues, inefficiencies, and implementation issues.
This idea was studied in the 1940's, and was rejected then, it's still a bad idea.
The video is pretty funny, they say you can always take off and land into the wind, but then state you can take off or land 3 planes at once at different places along the circle, but that would require one of those planes to land in a 60 degree crosswind, and the other to land in a 60 degree quartering tailwind, unless you had the planes on crossing approach paths (a really, really bad idea). A rejected landing would either put the plane right into another's approach/departure path or put wake turbulence right into short final approach of another aircraft. Wake turbulence from a heavy plane can flip a smaller plane, an A380 flipped a business jet 3 times flying 1000 feet below it last week*, the business jet landed, but the airframe is totalled. Another one also nearly flipped a 737 about the same distance away**. Ignoring it is idiotic, wake turbulence is at it's worst with heavy planes at low speeds, and drifts downwards and outwards. 3 simultaneous take offs still has the problem that if one plane takes off directly into the wind, the other 2 have to take off with a 60 degree crosswind & 30 degree tailwind unless you cross departure paths, at the other extreme it's a 30 degree crosswind into the wind, 90 degree crosswind, and a 30 degree crosswind/60 degree tailwind. You also put wake turbulence into the approach path of the next runway. You really can only use 2 runways on the circle if you account for the wind. Maybe I'm missing something, but I've made a bunch of diagrams and haven't found a solution that actually works for using 3 runways and accounting for the wind with non-crossing flight paths.
All the other potential benefits mentioned ignore the wind as a factor. It also later ups the capacity to doing the work 4 runways without explaining why. But let's ignore all the flaws in the concept itself and think about building one anyway.
11,480 foot runways would fit inside of the space of the circle, and you could build 3 at the same width for less concrete that the circle would use. But the circular runway looks a bit wider then needed, so you probably can get 4 standard runways out of the same area of concrete, probably a lot more since you don't need 50 taxiways. Even if that's not the case, 4 normal runways would use less land, and would not require building an enormous banking underneath the concrete, or the reinforced tunnels under the runways for passengers, cargo and equipment. The banking doesn't sound expensive until you realize that there's about 7 miles of it. The tunnels would need to be able to withstand a fully loaded A380 doing repeated hard touchdowns on it (remember, they can land anywhere on the circle) plus a safety factor, and be large enough for everything the airport needs. There would need to be enough land to build it near a large enough city to require an airport this size, which usually don't have large areas of inexpensive land available. And where are the cars parking?
6 runways - 3 pairs in a triangular format, with terminals and parking, uses far less land than the circular format, and is much easier to expand. The 3.5km runway idea mentioned earlier would use 5.3 km^2 or area for the triangular runways, where as a circular 3.5km diameter runway would require 9.6 km^2 of land, and in the picture much more than half of that area is for the runway and taxiways. You can fit a lot of terminals and parking in the 4.3km^2 you have left over with the triangular runways (and that does not include the area in the center of the triangle). And why would you want the terminal in the middle? It makes ingress/egress more difficult, for little to no benefit. On days with any wind (which is the vast majority of them), you'd only be using one pair of runways/2 upwind paths on the circle.
A complete, well thought out presentation of a bad idea is still a bad idea. They use the wind to justify one "benefit", but then pretend the wind doesn't exist as it is a massive problem to nearly
Oh boys, where to begin.
The obvious one is to take off and land while turning, but this could be corrected with a few straight runway connected around the circle (Kinda like an "angled" sun shape). I'm actually surprised they didn't propose that instead. Of course, that also mean you'll need a lot more space.
Then there's the air traffic management, it's going to be a pain since all plane that either lift-off or land will use the "same" runway at the same time.
Furthermore, if you want a 3.5 km diameter circle, that mean that you'll need to put a lot of infrastructure "undergound". Highway, parking, car renting etc. That's a lot of digging and a lot of concrete.
Also, I have serious doubt it'll raise the traffic. a 747 need over 2 kilometer for landing and take-off. A 3.5 km diameter mean 11 km circumference. So if you're really efficient, you'll have to shut down like ¼ of the runway. In other word, you'll only be able to run 2 corridor of landing and 2 corridor for lift-off at all time at ~90 angle. And that mean you bring back the problem you have been trying to solve in the first place.
So yeah, a lot of new problems only to solve one that isn't that bad to begin with.
Elok
...they wouldn't have to fight against crosswinds. And three planes would be able to take off or land at the same time...
If three are landing at the same time, I'd say that at least one is fighting cross-winds.
Instead of a loop, we should make the runway a Möbius strip! That way, planes can taxi along both sides of the tarmac, allowing it to last twice as long!!
http://archives.chicagotribune...
On a serious note, you can approximate this by simply paving an entire square mile. I once heard that Lawrence Livermore National Laboratory was such an airport in WWII, allowing damaged aircraft to land without having to line up for a runway. I don't know if that's true or not (it was a naval air base according to Wikipedia, so it might be), but the idea may be valid.
If you have a location with high winds that approach from many directions, and you have a wide open area, then something along these lines would work. Of course, there aren't going to be many locations where you have those wind conditions and the space, so you could approximate it with several different runways aligned based on the predominant wind conditions. That would work great! And that's pretty much exactly what they do at all major airports.
My local gliding club is a great big grass area. They have six winch points, but can shift the launch point to suit the wind. It also makes emergency landings safer as you have a huge area to aim for. The question isn't "will it work", it's "is paving the area worth it", and I very much doubt it. Take Dallas/Fort Worth for example - it has 7 runways. The total paved area of those runways is about 147 hectares (0.6 sq miles), assuming the average width is 60m.
To fit just a single 4,085m runway in a circle the paved area would have to be 1310 hectares (5.1 sq miles) - an increase of almost 800% on the current paved area. It would be even greater as the circle would need to be bigger to have parallel 4,085m runways. Granted, I'm not including taxiways, but even so I can't see it being economic.
The Navy tried that already in 1964. Popular Science ran an article about it: https://books.google.fr/books?...
Nobox: Only simple products.
Centripital force is
F = mv^2/r
or a = v^2/r
At a typical takeoff speed of 150 knots, the lateral acceleration needed to keep the plane centered on a round runway with a 1.5km radius is 3.97 m/s^2, or 0.40g. On a freeway you'd just tilt the roadway based on the expected transit speed (about 24 degrees for 0.40g). But with a circular runway, planes are going to be traversing every part of it at all speeds from 0 to 150 knots, so there's no single tilt which will eliminate the problem. Likewise, during the takeoff roll the required lateral force will increase with velocity. So you can't just tilt the wheel/joystick at a certain angle and hold it there while taking off. You have to constantly adjust it as your velocity increases.
If a plane has to make a no-flaps emergency landing at 200 knots (which also happened to be about the regular takeoff speed of Concorde), now you're talking a lateral force of 7.06 m/s^2, or 0.72g. Which brings us to why runways are straight in the first place. It's not because it's easier to design and build. It's because it's a stable travel path. If for whatever reason during takeoff or landing the plane's controls stop working, the plane will want to go straight. Making the runway straight means the plane naturally (and with a little luck) will stay on the runway. Making the runway round means if you lose that lateral force being applied by your control surfaces for whatever reason, the plane is guaranteed to depart the runway at speed.
Aircraft landing gear are designed to land in a straight line, they are not designed to handle side loads.
It doesn't take much side load for that gear to fold up under the airplane.
Not to mention instrument approaches... I'd love to see a precision instrument approach chart to a circular runway.
Most large airports have several parallel runways. LAX for example has 4 parallel runways, normally all in simultaneous use - 2 are used for departures, and 2 for arrivals. This gets a lot of airplanes in and out at the same time.
If the runway was a circle, you might be able use on side for arrivals and the other for departures at the same time, but that's it.
Circular runways remains a stupid idea.
Disclaimer: I'm also a pilot.
WTF do PhDs have to do with science? Next thing you'll try and tell me sociologists (spit) with PhDs are scientists too.
Scientists do science (hypothesis, theory, test, publish repeat). Engineers apply science and business. Occasionally doing a little art in between the two sets of constraints. Much more challenging, scientists don't get it, only see the 'applied science' part and think they're experts.
The old school version of a circular runway is a triangle of 3 crossing runways. Somewhat common at military bases.
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
I have a feeling you aren't either...
While sims are mostly faithful to their airplane type in the air, on the ground and landing, it isn't the same. That is the one thing I tell my trainees: "I don't care about the landings in the sim beyond you not crashing." There are small differences between the sim and the real thing that throw off landings. Coupled with the lack of visual cues and environmental (seat of the pants) cues, it makes sim flying not quite the same.
All of us airline pilots have our normal landing technique in the plane, and our "sim landing" technique. At least, that's my experience among three types of airliners, 7000+ hours logged, and another 400 hours just in simulators.
Scientists do science (hypothesis, theory, test, publish repeat). Engineers apply science and business.
That's the old fashioned way.
The new way is:
1) Do a study with lots of measurements
2) Crunch the numbers looking for an interesting trend
3) Create a plausible explanation for the trend
4) Publish!