Ground Effect; an essay
Introduction
The Concorde Stick and Rudder Book came out in 1990. It is a collection of virtual lecture notes for airline students about the peculiarities of Concorde behaviour when compared to that of a conventional jet airliner. Landing technique featured strongly, and was introduced by a couple of pages about ground effect, but I was not happy with these. I have not completely abandoned the idea of a revised S&R book, despite the fact that no one will ever fly a Concorde again. Now that comprehensive fly-by-wire disguises the individual characters of aircraft designs, and deliberately places a wall of unknowing between pilot activity and its result, it may be that some readers may become even more interested in historic ‘you get what you ask for’ handling, especially relating to a public transport aircraft of such radical shape and fame. This Ground Effect essay would replace the original paltry effort, and the present tense is generally maintained because understanding and knowledge should survive history. As a post WW2 technological masterpiece the Concorde still lives in our imaginations, as should its fellow air-squeezer, the Caspian Sea Monster.
Subject matter
The objective of the Concorde Stick and Rudder book is to present pilots with credible imagery that will help them understand and accept the special (and therefore different) behaviours of a Concorde, and very near the ground is no exception, in fact it’s an important and special case. The shape of the Concorde is designed for Mach 2 cruise, not landing at 160 kts. The double delta compromise - narrow at the front, wider at the back - is a modification of a simple, triangular slender delta shape, and the wider, camber-and-twist-with-drooping-wingtips part at the back is included because it reduces the rearward movement of the centre of pressure with increasing Mach number. In fact it gently reverses this movement as the Mach number rises above M1.3, resulting in the Concorde’s familiar 59% C of G with ½° down elevon cruise situation at M2. The American Valkyrie’s wingtip sections actually hinged downwards for the same reason. So what is special about this design? It works.
What has this got to do with landing technique? Nothing at all. It just so happens that the slender delta separated airflow can provide enough lift for takeoff and landing - to the surprise of not a few. So what about the ground effect, Why is it such a big deal with a Concorde?
While landing, most winged aircraft experience ground effect, which starts to take effect at approximately a wingspan above a solid surface (usually the ground), but the effect varies depending on most of the variables you might think of. Very low aspect ratio (high span loading) and proximity to the ground are two of them.
Conventional aircraft with wings and tail do not have to raise their noses by all of 3° in order to remove the airliner’s 3° descent path towards the ground; something less usually does the trick, and the pilot can retain his instinctive feeling that he is doing the levelling off by raising the nose. The subsequent floating behaviour also allows some fine adjustments for a nice touchdown. This does not work with a Concorde - at all! The ground effect is strong enough to completely arrest the approach descent in, say, the last 30 feet of ground closure - at constant attitude. It is a particularly exponential function (barely perceptible at 10 metres, and becoming very strong by 1 metre), but it is a self-adjusting phenomenon provided that the pilot coordinates with a slow but continuous up elevon action, accepts that this does not raise the nose, and does not interrupt this consistent technique because the descent appears to have stopped. Exponentially rising pressure behind the centre of gravity does the work if these rules are followed. This is the crucial feature of Concorde landing technique.
Ground Effect Research. How much is there?
When an aircraft prototype makes its first return to Earth it’s possible that landing behaviour has not been a major design priority: more a case of ‘suck it and see’, fingers crossed - good luck, tell us what you think, etc..
Before the Concorde flew, there was concern at its expected behaviour when pitch control is applied near the ground. In changing pitch attitude the elevons push the mass of the Concorde the wrong way. ‘How are we going to land it?’ Darrol Stinton describes this quandary in ‘The Anatomy of the Aeroplane’, and Alan Smith who taught us aerodynamics at Hamble mentioned it in 1962.
He had previously worked on Concorde subjects before becoming our lecturer, but the answer - no change of attitude - emerged later following real life aircraft testing on both sides of the Iron Curtain. Concorde and Tupolev research made level autopilot runs over runways at various (low) radio altimeter heights for autoland research), to find out what was going on. Power/drag, elevon position, attitude and so on all can be used to build an aerodynamic picture, but other serious and quite different work in the mid ‘70s was done in Russia, specifically to make a ground effect military transport:
The Caspian Sea Monster proto getting up steam
The operational model firing a missile in the cruise.
This prototype plan drawing explains why first satellite photo interpreters expected wings. The two tail engines were not needed. Eight engines, just outside the cockpit windows, sufficed.
Performance
The numbers will amaze you. Max weight 540 tons [includes tanks (cavalry), jeeps, field guns, ammo, troops, and six anti-ship (nuclear?) guided missiles on top (good for a 90 km clearing shot ahead)]. Wing loading 820kgs per sq.m. Best cruise: 230kts, Altitude: 4 to 14 metres clear of the water. Biggest problems: manoeuvrability - very large turning circle (like a Concorde Round-the-Bay?) - and a chop/wave height limit of 1.4m (four and a half feet).
History
Prototypes were ‘flown’ by two government test pilots, not company employees, but designer Alexeyev insisting on sitting between to work the engines. During its development, Kruschev-era traditional heavyweight ground force tactics were to be replaced by modern Brezhnev hi-tec, missile-based politics, and Alexeyev’s work was quietly sidelined. The engines were large and had been fitted with controllable vectored thrust nozzles, to blow some air down and under the wings to help get the monster lifting. One can imagine that managing the critical nature of getting to 230 kts may have required some crew skill. The large tailplane suggests significant pitch requirements, and, though Alexeyev was politically retired, development continued. But an amateur was posted to the engineer’s seat, and felt nervous about applying enough thrust and tweaking the eight nozzle positions at the same time. The pilots might also have regretted Alexeyev’s departure. A subsequent takeoff, accel. and climb attempt went wrong and, though everyone got out, damage was serious and the device sank in deep water.
Pictures of it charging across the water started something of a worldwide craze for the ekranoplan (the Russian word for ground effect flying machine). Experimenters worldwide did their own ground effect research - small versions, of course - but a consensus emerged about ground effect science. Its simplicity sends an impressive message to all teachers:
Ground effect increases pressure under the wing - extra lift force.
It significantly reduces induced drag by changing (washing outwards?) wingtip effects and flattening downwash. The result is equivalent to increasing wing aspect ratio i.e. the virtual wing gains in span as the ground is approached.
Pitch trim alters - to an extent depending on overall design. E.g: for conventional transports, tailplane position makes a difference - VC10, high tail, not much pitch effect: B707, low tail affected by wing downwash and its own circulation near the ground, therefore tail downforce reduces, more up elevator required to compensate (as well as that required for thrust reduction pitch effect).
Concorde: a special aerodynamic combination of complexities
The Concorde takes off and lands with nose down elevon trim: takeoff trim setting is actually as down as it gets for the whole of a typical flight! Between 250 and 400 kts (top of the subsonic speed range) the elevon position stays much the same: and, of course, Mach 2 is a mere ½ down. This is all unconventional (even considering the C of G movement) - so why Is this?
At low speeds - both at lift off and approach at the familiar 13½⁰α - much of the wingtip area does not provide lift: the angle of attack is too great for the less swept wingtip area, and beyond that the long, virtually 90⁰ swept wingtips join in a farrago of confused, non-lifting airflow: in fact, early drawings show the separated flow actually leaving the wing en masse prior to reaching the trailing edge. Of course this does not mean that pitch control does not work, but it now operates in its own environment.
Let bygones be bygones?
It is worth saying something about this separated flow drawing. Dr Dietrich Kuchemann drew it or dug it from his notes while working at Farnborough. He was a lonely proponent of the outrageous slender delta principle that was finally accepted as the idea that would work for a Concorde. How could he have been so sure? Until the allies swept into Germany in 1945 from both directions, snapping up scientists and engineers as they went, he had worked on German fighter performance improvement at Braunschweig. We are not talking about a souped-up Me109 here. Did the Germans know what a Mach 2 aircraft might look like in 1945? It certainly looks like it. So the Concorde is German? Just a thought, but patriotism apart, I would be prepared to consider it.
The tip stalling problem as a way of life
It has long been known that swept wing aircraft lose lift from their wingtips as incidence is increased. The wingtips are behind the centre of lift, and less lift at the back can cause pitch up (especially in a fighter’s steep turn. This macho manoeuvring problem could also apply to straight wings - for different reasons). The Concorde is an extreme case of sweepback (regarding pitch behaviour), but the result is much more benign and progressive, because of the detached flow principle. In fact it has no traditional wings - just two very large wingtips - and the movement of lift within the speed range, between 1150 and 160 kts, or incidence 2⁰ - 13½⁰, is surprisingly progressive, to the extent that minimum extra control gadgetry is needed. Marketing terminology such as ‘natural stability’ might have helped certification, but some alternative truths can be justified by the masterpiece of the radical design that this aircraft represents.
The Aerospatiale Concorde aerodynamics man, Dudley Collard of Stepney and Toulouse (he thinks in French), recently (2018) told me that the shaking above 7⁰𝞪 was caused by upward flow around the forward fuselage (lift) breaking away from the fuselage sides at this incidence (and, perhaps as more shed vortices, getting involved with the forward wing vortex system). So the fuselage also has its own flying contribution and provides lift at the front at low speed. Makes sense. Is this the whole story?
What happens when the Concorde lands?
Pictures of airflow patterns, diagrams of streamlines and so on give you the impression that it’s the air that is on the move, around a stationary flying machine. Of course this applies to wind tunnel and classroom life, but real life is different. When the stationary air above a stationary runway is approached by a landing Concorde it gets squashed - squeezed like Polyfilla under an expensive filling knife. Polyfilla is a little on the soft side for our imagery - putty is better. It can be squeezed, but resists, and does not react well to hasty and uncertain putty knife work - the desired smooth result requires just the right smooth, consistent technique.
Extra pressure underneath: This squashing of the downwashed air between wing and solid ground raises its pressure, and increases lift, naturally enough as a function of proximity to the ground. The Concorde trailing edge gets closest to the ground, so to prevent the tail rising in ground effect the elevons must be progressively raised (consistent with every centimetre nearer the ground). Bernoulli is all very well, but this is not quite the same thing.
Increasing assistance from the wingtips
Reducing incidence = more lifting flow restored to the wingtips. There are more than two squashing-the-putty features which help stop the Concorde going down.
Progressive nose down trim change as the trailing edge approaches the ground. This is why the Concorde’s elevons must be progressively moved upwards; the reason being to reduce too much lift at the trailing edge. There are two causes, squeezing of the air, like the Monster, and modification of the wingtip vortex behaviour. The elevons move from down to up, reaching several degrees up at touchdown. Could this also allow a simpler exit flow from the trailing edge, cleaning and reattaching much useful wingtip flow? ‘Whoosh’ is what you hear if you get it right, the shaking stops and the Concorde feels as if it has left the cobbles and stepped on to the ice as drag reduces noticeably! This only happens in perfect GE conditions; no turbulence or crosswind, and when the wheels are within one metre of the ground. These symptoms indicate landing perfection. Instruction and advice about technique follows in the next chapter.
Ways of looking at this
Before going on to describe the correct piloting technique (and its problems) it may be useful to consider what the tailless Concorde wing is doing at landing speed.
The C of G position in steady flight implies that lift is equally divided fore and aft of it while flying in free air - above ground effect. On an approach (or after takeoff) this position is close above the mainwheels, maybe returning to a metre in front of them on the ground when the nosewheel lands. The area of wing behind this line looks large, but much of the wingtip area does not provide lift at takeoff and landing incidence. Outboard of the kink in the leading edge where the sweepback initially becomes less (from 75⁰ to ‘only’ 55⁰) the leading edge then curves around again in ever increasing sweepback angle. The slender delta shape at the front continues to lift strongly as speed is reduced. A stabilising vortex starts at this kink in the leading edge (7⁰𝞪?), and streams aft over the wing. The giant ice cream cone vortex on top of each wing - low pressure - is hereby stabilised in position. But outboard of this line (just outside the engines) the wingtips give rise to a variety of vortices and disturbances, but little lift (the wingtips provide increasing lift at supersonic speed). This is why the elevons are at their most down position at takeoff, to keep the Concorde balanced (in trim). Instead of lifting wingtips the elevons (like flaps) have to push the air down at the back. No further special pseudo science is necessary, except to leave the reader with these thoughts:
If we draw a 53% chord line across the wing, the centres of pressure and gravity position, we have two wings, each providing an equal lift force during a steady approach before ground effect is encountered. There’s a 75⁰ slender delta of low span flying high above the ground at the front, and a considerably larger and higher aspect ratio wing at the back, but flying with very stalled wingtips, and this rear wing gets very close to the ground at landing. They usually work as a team, but which one will be more affected by ground effect? The answer is simple (the low-flying wing at the back, behind the C of G).
Without a change of attitude the correct ground effect manoeuvre curves the flightpath (asymptotically) towards level flight as the ground is reached. This means that the general angle of attack reduces from 13½⁰ to 10½⁰, allowing more of the wingtip area to fly.
Ground effect Polyfilla or putty has been mentioned, but there’s another image for the pilot: ground effect is like encountering a sausage balloon under the trailing edge. It needs to be squeezed by the pilot with continuous up elevon until the wheels touch the ground. If this elevon is released at any point the balloon will escape. (At the same time the rising tail means less incidence for the strongly lifting slender delta at the front, and less essential energy goes into the spinning ice cream cones on top of the wings. From here, the only way is down! And unless the runway can be lowered 50 feet for another attempt the result will be disappointing.
Why this special essay?
These ideas are deliberately informal, unmathematical, and maybe not entirely scientific, but experience proves that pilots who visualise new techniques in a way that is credible and matches the evidence adapt to a new challenge quickly. Training money and the cost of damage can be saved. The vertical accelerations and therefore the structural loads of a bad Concorde landing tend to be attenuated by the fuselage, and not fully appreciated at the cockpit, but the central fuselage and wing structures definitely take the strain. You might be surprised at how much the mass of the engines can bend the wings downwards; it’s disconcerting to see!
The real point of featuring the Caspian Sea Monster is to indicate how much lift can be supplied by a small wingspan when very close to the ground.
No apology is made for the detail and apparent repetition in the Stick and Rudder pages which would follow. For those who read and digest it as background understanding it will be worth it, both in maintenance costs, base training time, and pilot peace of mind.
In the book, here starts the next chapter about landing technique, dos and don’ts and analysis of the problems of abandoning traditional instincts.
Now we’re all friends there are many Caspian Sea Monster pictures that can be Googled, and comments from Russian enthusiasts: “Why don’t we make machines like this any more?” They have a point. What a toy, is the monster!