Twisting thrust for supermanoeuvrability: the majority of aircraft engines produce forward or reverse thrust roughly parallel to the fuselage axis. Given variable-geometry exhaust systems, combat aircraft powerplants can also generate substantial jet lift and pitching and yawing movements.

Thrust vectoring, implying changes in both jet orientation and throttle setting, is a problematic term coined by the late Professor Theodore von Karman. The words became reality in the 1960s as the result of:

* the predicted Nato need for dispersible stovl aircraft to avoid NBC strikes and

* the known drawbacks of tail-sitting Vtol projects (namely, difficult piloting and lack of overload/stol capability).

For those trying to conceive a 'flat-riser' stovl ground attack aircraft in the 1950s, the challenge was to install three times as much thrust as normal while retaining a useful amount of fuel and warload, and using an internal arrangement that would 'balance' about the required centre of gravity (COG).

In a conventional jet fighter the power-plant in the rear roughly balances the crew, cockpit and equipment in the front. The problem with stovl was to invent a jet lift powerplant that did not need to be placed so far forward that a normal cockpit position was impossible.

Turbofan engines were then still new, and for simplicity had initially been developed by adding a turbine to power an aftmounted fan. In his BE.52 project Gordon Lewis of Bristol Engines took an Orpheus turbojet and to it added a front-mounted fan, which was linked to a low-pressure turbine via a co-axial shaft. Ducting the fan efflux to two rotatable nozzles incorporating 90[degrees] bends provided jet lift under the aircraft CoG, with the engine sufficiently far aft to allow a front cockpit.

Working with Ralph Hooper of Hawker Aircraft, this three-nozzle concept was developed into the BE.53, with four rotating nozzles and contra-rotating shafts. This formed the basis for today's Rolls-Royce Pegasus, as used in the BAE Systems Harrier and Boeing AV-8B Harrier II series.

The BE.53 was basically a pastiche of Orpheus and Olympus components, and this minimal-cost approach was to characterise the entire British Harrier programme. The Harrier was technologically a poor substitute for several more advanced stovl designs (R1154, HS.1170, R1216) proposed by the Hawker/HSA/ BAe project office.

Digressing, it may be noted that in later single-engined vectored-thrust projects Hawker designers returned to the three-nozzle arrangement, using what Rolls-Royce now terms a Three-Bearing Swivel Module for the centreline exhaust. This 3BSM forms part of the Rolls-Royce LiftSystem for the Lockheed Martin F-35B.

Although it represented only a low-cost, low-performance approach to stovl, the Harrier convincingly demonstrated the flexibility provided by thrust vectoring in both land- and sea-based operations. If major coalition airfields had been attacked with ballistic NBC missiles in 1991 or 2003, stovl might now be taken far more seriously.

VIFF 

Early on, Harrier pilots grew curious about using thrust vectoring in forward flight (viff) and one RAF officer was court-martialled for 'endangering his aircraft'. Nasa performed some tests with a Kestrel (Harrier-precursor) and concluded that viff was worth 0.5G of additional load factor, coupled with strong deceleration and a nose-up pitch, giving a 'square' turn.

The US Marine Corps decided to conduct its own viff tests with an AV-SA and assigned the task to Lt Col (subsequently Lt Gen, and more recently Lockheed Martin deputy director for the F-35 programme) 'Harry' Blot.

Blot boldly decided to begin viff tests at an airspeed of 500 kt (925 km/hr). His report stated that, << the airplane started decelerating at an alarming rate, the magnitude of which I could not determine because my nose was pressed up against the gunsight. I was now straddling the stick, with my right hand extended backwards between my legs, trying to hold on for dear life >>.

The US Marine Corps concluded that viff (which for the Harrier works like a large airbrake that is invisible to the opponent) is a useful last-ditch manoeuvre for dislodging an enemy from the six o'clock, but that the pilot must be tightly strapped in.

An important technical point (which was missed in the first Hawker stol performance predictions) is that the thrust vectored is gross thrust, equal to net thrust (direct from the manufacturer's brochure) plus intake momentum drag, the product of engine mass flow and aircraft velocity. At high speeds, gross thrust is far greater than net thrust.

Rear-End Motion

If there is no requirement for vtol, the thrust vector does not need to pass through the aircraft CoG, and a conventional rear engine location can be used. In this case, turning the jet through relatively small angles (20[degrees] or less) can produce large pitching and yawing moments, giving powerful attitude control. This is especially useful at low airspeeds, where aerodynamic controls are ineffective.

This type of thrust vectoring was first investigated in the US as a means of stall/spin prevention and spin recovery, since fighters traditionally lose energy in close combat and may diverge from controlled flight. Only later were simulator studies carried out to assess its use in close combat manoeuvres at post-stall angle of attack.

Aside from increasing in-flight manoeuvrability, powerful pitch control can give a stol capability, with take-off rotation at subnormal speeds and approach at far higher angles of attack than can be employed at touchdown.

Such thrust vectoring, the turbine-engine offspring of Germany's wartime V-2/A-4 and Wasserfall rocket jet deflecting, required no great invention. It was instead a matter of whether the weight penalty was justified, and whether flight at large sideslip and an angle of attack far beyond the stall was really practical. Among other considerations, the pilot could no longer see where he was going. The concept became more attractive with the development of digital controls for both the aircraft and the engine (to avoid surges due to inlet flow distortion).

America conducted a whole series of vectored-thrust flight trials, using dedicated research aircraft and modified fighters.

The Rockwell/Boeing X-31A Vector was jointly funded with Germany, reflecting the latter's interest in the post-stall reversal. Known as the (Wolfgang) Herbst Turn or J-Turn, the aim is to achieve a 180-degree change of direction, without losing energy or visual contact with an opponent passing on a reciprocal heading.

The X-31A has a compound-delta plan-form, with a canard and a General Electric F404-GE-400 engine. It has three carbon-carbon 'paddles', stressed to turn the full afterburning thrust through 15 degrees. The first (1990-1995) Nasa-Dryden tests were flown under the Enhanced Fighter Maneuverability programme, involving operation to 70[degrees] angle of attack.

In the second (2001-2003) series, 'extreme stol' (estol) was tested with funding assistance from the US Navy. Extreme landing depends on an automatically controlled 'derotation' from a high angle just prior to touchdown. A video camera mounted under the nose gave the pilot a view of the runway. Increasing approach angle of attack from a conventional 12 to 24 degrees reduced speed from 324 to 224 km/hr and the corresponding ground roll from 2440 to 520 metres.

Once the value of rear-end thrust vectoring was established the question was whether it should be retrofitted to existing fighters.

In a joint Nasa/US Navy programme a prototype Boeing F/A-18A was fitted with external vane jet-deflectors for its General Electric F404s. The deflection system added 1000 kg, while ballast and other changes took the total penalty to 1680 kg. Flight tests with this F-18 Harv (High Alpha Research Vehicle) from 1987 to 1996 included Herbst turns and the use of an offensive descending spiral inside the flight path of a tightly turning conventional opponent, tracking the target for up to ten seconds and a 60[degrees] angle of attack.

The add-on jet-deflection systems of the X-31A and F-18 Harv were crude and heavy. For its Fll0-GE-100 General Electric developed an Aven (Axisymmetric Vectoring Exhaust Nozzle) system that achieves 17[degrees] vectoring in any direction simply by moving the divergent flaps of the standard con-di nozzle. The Aven was fitted to a Lockheed Martin NF-16D and tested in 1993-94 under the Matv (Multi-Axis Thrust-Vectoring) programme.

The NF-16D was later tested with a Pratt & Whitney F100-PW-229 and that company's PYBBN (pitch/yaw balanced beam nozzle), giving 20[degrees] of deflection in any direction.

Two of these PYBBN engines were later applied to the US Air Force Flight Dynamics Laboratory's Boeing NF-15B stol/mtd (maneuvering technology demonstrator), which in 1996 used thrust vectoring up to Mach 1.95. The NF-15B 'Agile Eagle' had previously (1989-91) been tested with two-dimensional thrust-vectoring/reversing nozzles developed by Pratt & Whitney, demonstrating take-off rotations as slow as 66 km/hr, and landing ground rolls as short as 415 metres.

A two-dimensional thrust-vectoring nozzle is heavy and can (usually) provide vectoring only in pitch, but is mandatory in stealth aircraft to reduce the rear-aspect infrared signature. This was consequently the approach adopted for the Pratt & Whitney F119 engines of the Lockheed Martin F-22 Raptor, which entered service in late 2005.

Russian Revolution

Vectored-thrust developments in Russia broadly mirrored those in the US. Ground tests with external vane deflectors began in 1986 and a two-dimensional nozzle pitch-control system was flight tested on one Su-27 engine in 1989. Following tests in 1996 of a Sukhoi Su-27 with two axisymmetric nozzles with 15[degrees] up/down deflection, in November that year India signed the initial contract for Irkut-built Su-30MKIs with Saturn/Lyulka AL-31FP thrust-vectoring engines.

The AL-31FP with AL-100 nozzles providing 15[degrees] up/down movement was launched into low-rate production at Ufa (Umpo) in 2000. The nozzles are installed in the Su-30MKI with the plane of rotation at 32[degrees] to the vertical, allowing yawing moments (coupled with small rolling moments) to be generated by asymmetric pitch applications.

The Su-30MKI entered service in September 2002, giving India a three-year lead over the US in thrust vectoring. In Exercise Cope India 2004, the IAF Su-30MKI was reported to have won eight out of ten engagements with US Air Force F-15Cs. An exercise with Singaporean F-16s is said to have been a total wipe-out.

The AL-31FPs have also been flying in the prototype Su-27KUB naval trainer, suggesting that these engines may well be retrofitted to the Russian Navy's Su-33s, which formally entered service in 1998. They probably feature on the Su-33 being marketed to the Chinese Navy.

In 1998 Klimov unveiled a mock-up of the MiG-29's RD-33 engine with a thrust-vectoring Klivt nozzle, providing 15[degrees] off-axis deflection in any direction at a rate of 60 deg/sec. In August 2003 this first flew on the MiG-29 OVT, which made its public debut at Maks 2005. Its sensational manoeuvres made it the star of Farnborough in 2006 and the production MiG-35 version is now being offered to China and India. The Klivt nozzle is also being promoted as a retrofit for existing MiG-29s, presumably with an independent fly-by-wire control system.

A nozzle based on Klivt technology is available for the Salyut-built AL-31FN-M1, which is being promoted for China's J-10 aircraft.

The Russians have thus taken the lead in terms of service introduction, exports and air show demonstrations of rear-end thrust vectoring for supersonic fighters. It undoubtedly provides a major advantage in ski-jump take-offs and close combat, but it may be argued that the Russians are exaggerating the significance of the dogfight, in order to boost sales of basically old designs.

Like the US Air Force, the Russian Air Force shows no serious inclination to retrofit its older fighters with thrust vectoring, perhaps because a new integrated digital flight control system is needed for full benefit, and because they expect to make most kills beyond visual range.