Circular Runways

I thought of that, but the video seemed to show planes approaching the ring (and not yet over it) to be quite low. Seems to me that the more time you spend in the airspace above the ring, the fewer takeoffs and landings you could have in a given amount of time (while still keeping adequate separation between planes).

It’s an interesting idea, and there seem to be some serious people looking into it, so I don’t want to just dismiss it out of hand. But I can see an awful lot of issues that need to be sorted out before this gets used on any sort of real scale.

How does it handle planes with a range of landing speeds (and a range of turning radii)?
How tricky will the transition from straight approach to curving touchdown be, and at what altitude?
What do you do when your airport needs more capacity than one ring can provide, build another? How do planes taxi from one ring to the other?

Landing is hard enough without having to cope with a constantly changing wind direction that a circular strip would provide. And while theoretically several planes could be taking off and landing simultaneously, there could be serious conflicts if one plane needed more room or ran into trouble. Which would be worse, colliding with another plane or falling off the edge as happens now?

OTOH, I didn’t like auto roundabouts until I had to use them daily. Now they don’t seem so bad. Maybe circular is where it’s at.

On a taxiway, silly.

The Navy successfully tested circular runways in 1965, as seen in this Popular Science article from June 1966.

That article falsely states that no one had thought of the idea before. It’s come up multiple times throughout the 20th century. Time discussed the idea in its March 21, 1955 issue. Flying Magazine did a full article on it in April 1951.

Some plans in the 1920s even had them on top of buildings. Of course, planes needed a lot less landing room in those days.

All in all the guy’s idea is IMO profoundly ignorant. But it raises some fun opportunities to talk about. It’s better to treat it like the *eating a blue whale *thread or an xkcd what-if: silly, not serious.

Which takes us full circle (heh :)).

The reason airfields are called airfields today is that back in the 1920s they started out as entire farmer’s fields 1/4 or 1/2 mile on a side. Why? So the pilots could land or take off in whichever direction the wind is blowing.

Eventually they discovered the airplanes cut ruts in the field in just a couple of directions based on the prevailing winds. Somebody started mowing and flat-rolling those well-worn paths in the grass and the modern airport runway layout was born.

Marker beacons? Marker beacons!?? We don’ need no steenkin’ marker beacons!

At least in the US they’ve mostly been decommissioned and the existing ones are left to run until they fail and are then removed. Compass locator beacons are circling the same drain.

The latter. ILS are stunningly expensive and finicky. GPS-based straight or curved approaches are the way of the future. For conventional runways, curved runways, and the new and even more spectacular rotini-shaped runways.

After all, if curved runways let us have multiple approaches at once, think how much more airport capacity we’ll have if we can stack multiple layers of arriving and departing aircraft each landing on their own turn of a giant spiral screwing upwards into the sky. :smiley:

Yup.

The AC-130 does a similar precision circling maneuver versus the wind all day long but from higher altitude. Although the folks in the central terminal might not like the test outcome much.

You read about the US Navy tests that Expano Mapcase links to?

I just did. The magazine breathlessly reports the Navy tried it under benign conditions and nobody was killed. It said the Navy was enthused and would promptly carry the idea forward. Looking around today 50 years later we see the enthusiasm apparently waned pretty quickly.

As a matter of physics and raw airplane-vs-sky I don’t see why it couldn’t work, given a large enough radius of curvature vs the speed of the aircraft type(s) using it.

It’s the interface to the rest of the aviation system that presents impossible obstacles. We’d need different airport design standards, different airspace and airport operations standards, instrumentation to support the novel features, training for everybody from designers to builders to ATC to finally pilots.

Plus of course the political problems with noise pointed out by JoeyP. Concentrating noise on the politically powerless works politically. Spreading noise across everybody doesn’t work politically.

It’d be the work of decades to bring the first one online.

And for what? It mostly solves problems we don’t have. There certainly are airports plagued with nasty crosswinds. The reason those airports don’t have appropriate runways already is they don’t have land to build on. Or there are nearby hills, mountains, etc.

My overall take is IMO the required curve is a lot more gradual than he’s positing. You’re not building a ring 1 mile in diameter; more like 2 or 3 miles in diameter. If you have a square flat area 3 miles across available with no hills in any direction you already have the ability to build conventional runways pointing in whatever direction the winds demand.

Some thoughts.

  1. A large airport setup with 2 or 3 parallel runways can land multiple aircraft at the same time, into wind. A circular runway can’t do this.

  2. The wind at many airports only comes from a few directions, so a circular runway in these places would have large unused sections that experienced almost constant crosswinds.

  3. Due to point 2 above; Landing in a super strong crosswind, as seen in scary YouTube videos, and cited as this guy’s inspiration, is not all that common.

  4. I think the biggest advantage to a circular runway is that if you had to abort, you don’t have to worry about running out of runway. For this to be of any use though, the runway would have to be unoccupied by other aircraft.

  5. I’m not sure how the runway banking would work. Any angle of banking would only be good for one speed, any slower and the aircraft will want to veer toward the centre of the circle, any faster and it will want to veer out of the circle. How do you design it so it is safe for all types of aircraft to use from an A380, to an F16, to a B737, to a light turboprop or even a C172? An interesting exercise would be to calculate how much banking the runway would need for a speed of, say, 140 knots, on a circular runway with a diameter that fits within a typical modern airport boundary.

I guess one solution would be a very wide circular runway with banking that got progressively steeper toward the outer edge. As you got faster the aircraft would tend to move toward the outer edges of the runway where the banking is steeper and matched for the aircraft’s speed.

Do US ILS installations without marker beacons all have DME for a glide-slope check or are they moving toward GPS distance only?

What is your landing airplane doing in the last several seconds just before touch-down?

Is it flying in a straight line tangent to the expected touch-down spot? Or is it flying in a curved path following the runway?

Remember, airplanes typically flare (pull into a nose-up attitude) into a near-stall in order to slow way down and just sort of float down to a landing. (Well, that’s how we were taught to land, 40-some years ago. I’ve heard that technique has largely gone out of fashion in recent years. Anybody?)

And when planes are flying in a curved path (that is, banked), their stall speed increases.

So does this imply that planes landing on a circular runway must do so at higher airspeeds in the moments just before touching down? And thus, must land at higher speed? And also, with greater danger of stalling into a crunchy landing? Would this affect small GA airplanes more (where the pilot is actually flying the plane), as opposed to a large jetliner (where a landing would more likely be all computerized)?

@Richard Pearse: That proved to be a more thought-provoking (and research provoking) question than I expected.

I’m spoilering the answer since it’s a major pilot geek-out that’s a total hijack to the actual thread the others are having.[spoiler]My carrier services many small town fields across the US that don’t have 24/7 control towers or radar approach control. Having said that, between my being now on bigger airplanes and preferring to be snug in bed well before the wee hours, I don’t visit those places and situations much any more. They’re still in my iPad though.

Anyhoo …

Many ILS, and certainly the vast vast majority at major airports*, are equipped with DME. Absent a marker beacon, the next common legal alternative to identify the final approach fix (FAF) is a cross radial from a nearby (10-15 miles max) VOR.

That seems to be their pro-forma choice when the reality is the airport is served by 24/7 radar approach control. Where approach radar is available you can consider the FAF radar identified, but as a practical matter controller workload isn’t going to permit him/her to call it out for you unless you’re the only jet in the sky.

I vaguely recall one middle-sized city we used to serve in the 727 that needed the cross radial to be identified onboard to find the FAF late at night when local ATC had gone home. Absent FMS I’m betting that still goes on in the wee hours all over our fair land.

There are a scant few that still use an LOM co-located with a marker beacon with no nearby VOR assistance. I was surprised to still find one where the LOM was a designated IAF. So in a steam gauge airplane one could transition from the enroute structure by homing to the NDB then maneuvering via a holding pattern entry or even complete holding pattern to align with the localizer and descend to FAF altitude before fixing the FAF via marker or LOM passage and starting descent to the runway all with no radar hand-holding. Shades of my Dad in the 1950s; hard to believe it’s expected even as a tertiary backup plan today.

I found one fun situation where there’s a VOR situated almost but not exactly along the final approach course about halfway between the FAF and the runway. http://flightaware.com/resources/airport/CID/IAP/ILS+OR+LOC+RWY+09/pdf . The ILS has no LOM or markers; you fly laterally and vertically off the ILS while using the VOR’s DME for identifying the FAF and a later LOC-only stepdown fix. And, unless you like using a stopwatch, the LOC-only MAP if the GS is unusable.

Talk about a setup to screw up if somebody ends up flying the VOR course guidance by mistake. Besides being 1/4 mile off course laterally you’d be at about 1000 AGL descending on what you thought was the localizer when suddenly you get cone of confusion and station passage. Oops! To make it really easy to goof, the ILS has the same base identifier as the VOR: ICID vs CID. Even if you’re doing the right thing it’s kinda weird to have the DME count down to zero then start counting back up while you’re still correctly on your way towards the runway.
The real bottom line nowadays is this:
There’s a general overlay rule that says a properly functioning FMS system using IRS with DME/DME updating or a properly functioning GPS with or without IRS can be used to identify any fix anywhere any time.

So as a practical matter for most jets most of the time nowadays the answer is we’re there when the FMS-driven moving map shows we’re there. Since everybody needs to have ADS-B installed by IIRC Jan 2019 for commercial and Jan 2020 for GA, it won’t be long before FAA, ATC, and the rest of the alphabet soup can assume 100% GPS & FMS or FMS-lite equipage. At which point damn near all ground-based infrastructure becomes optional.

=====

  • At my home base we have 8 runway ends. Probably 80% of my arrivals are on 2 of them and 95% on just 3 of them. One end that I’ve very rarely used is a ILS without DME whose final approach fix is over the water. So no marker beacon or LOM. I had never noticed until I was looking just now that (absent FMS) the final fix is radar only.[/spoiler]

The flare to a near stall is still taught and used in light airplanes.

Working hard to absolutely minimize touchdown speed was the highest priority back in the days of bumpy dirt runways and nose-over prone taildraggers. It’s carried forward to today along with 61 knot stall speeds as a way to improve the survivability of forced landings and the usability of short airstrips by aircraft without antiskid wheelbrakes.

As you get to bigger aircraft the priorities change; it’s more important to be fully in control and touch down sooner along the runway than it is to wring out the last few knots of touchdown speed. So in a light twin or bigger the approved flare and touchdown technique is more like reduce the descent rate to about 1/3rd of the approach descent rate and let it impact that way. With a bit of finesse in the last few inches, but not at the expense of getting slow enough to get sloppy controls.

In a jet you also have the concern that the nose attitude to stall is so high that the tail would drag the runway before the landing gear could reach it. The cure is to touch down at a higher percentage over stall speed. That way the attitude is flatter and the gear sticks down farther than does the tail.

For small bank angles the difference is minimal, a few percent at most. But yes, if folks really expected to land in a 10 degree bank there’d be a need to add a couple more knots or mph.

As a separate matter, about 98% of jetliner landings industry-wide are hand-flown. We only autoland a couple times per year for practice and maybe once or twice for real in serious fog or snow.

One final add-on: A sizeable fraction of the 737 and RJ fleet aren’t equipped for autoland at all. As well, a sizeable fraction of all US airports aren’t equipped for it either. So those jets and those airports see 100% manual landings.

Would circular runways be more useful in military airfields, as they won’t have many of the other considerations that **LSLGuy **mentions upthread.

I envision a giant wall of fans that blow air so fast that the plane can land vertically.

Here’s a tidbit from an older thread on silly games in flight simulators you might enjoy:

Think of the places that use banked tracks today: velodromes, indoor running tracks (Armory in NYC, home of the Millrose Games; not the joggers at your local YMCA), car race tracks (Datona, Talledaga, Indy, etc.) They all have a good bit of bank to them, but then what uses them is fairly narrow. The bank to this runway couldn’t be that much for fear that a large plane could scrape the outer wing. Does that mean that the runway needs to have an even larger diameter?
On the positive side, the pending commercial pilot shortage would be remedied…with recruits from NASCAR. :smiley:

Something I’ve never understood is why a redneck sport like NASCAR makes all their turns to the left. Thet way’s fer Commie faggots.

At least in the OP’s silly video the airplanes are turning right while landing. As God intended all rednecky NASCAR types to always do. So they’ll be naturals at it. YEEE - HAAA!! :smiley:

At first I was thinking that you don’t fly a curved path, you just fly a tangential one one that results in no crosswind. But that does make your landing target much smaller.

This would seem to conflict with the stated goal of eliminating crosswinds as a landing problem. If landing on a straight runway with a crosswind is hard, staying above a circular one as you slowly descend with a constantly changing crosswind seems like it’d be much harder. Is my intuition right about that?

I finally found the article Air & Space Magazine did on the same experiment - it even has the same photo of the T-28. It appeared in the April / May 2004 issue, but I don’t have a way to scan and post it here.

The diagrams seem to indicate tangential paths, but yes, that means landing on a dime.
But I’m not sure how you fly a the last bit of a straight tangential path whilst also banking so that your wheels all hit the sloped runway at the same time. - so it seems that at some point, just before you touch down, you’re going from a straight path to a curved one - maybe the curvature is gentle enough to make little difference to the relative crosswind direction.

But… dunno. I think it’s great to explore ideas like this, but it seems like the mission statement of this thing is “Let’s deal with crosswinds and airport noise by making landing more technically difficult and more dangerous”