I will be very honest, up to what i know indy cars work more like tailplanes, stable aircraft need a negative lift vector on their tail to keep the nose up indy cards do the same.
The mechanics of flight are very complex, however my only point was to tell you the negative deflection does not mean necesarily pitch down of the nose, but it is a way to achieve higher lift in some situation read this and later see the video and check the X-31`s canards deflections
"2. Brief Description of the Related Art
Until about 1978, the region beyond stall was considered an unacceptable flight regime frequently characterized by uncontrollable flight in spins and by undesirable deep stalls. Any deep stall condition is characterized by a stable trimmedflight but at a high angle of attack from which return to normal flight may be difficult or impossible. A deep stall may be defined as an out-of-control condition at an angle of attack greater than the angle of attack for maximum lift with nosignificant motion other than a high rate of descent. Conventional airplanes usually stall and lose control effectiveness at angles of attack in the range of 18.degree. to 20.degree..
However, according to U.S. Pat. Nos. 4,261,533 and 4,099,687, it is now possible, through the use of a rotatable horizontal tail on aft-tail configurations or through the use of tiltable engines on the wings, to provide stable and controllableflight at extremely high airplane angles of attack.
Because movement other than a high rate of descent can be controlled by varying thrust levels and all moveable control surfaces with large deflections, the safety and usefulness of flight at extremely high angles of attack are being re-examinedand redefined.
The essence of the longitudinal control concept, as set forth in U.S. Pat. Nos. 4,261,533 and 4,099,687, is to rotate the tail or to deflect large chord elevons to magnitudes of approximately the same order, but of opposite direction, as theairplane angle of attack, so that the effective tail aerodynamic angle of attack is below the tail stall angle and is thus capable of providing both stability and control for the entire aircraft.
Although rotatable canard arrangements are known from U.S. Pat. No. 4,569,493, U.S. Pat. No. 4,281,810, U.S. Pat. No. 4,010,920, and West German Offenlegungsschrift 2421524, such arrangements deal strictly with the stability and control of aircraft and models in level and unstalled low angle of attack regions of flight and do not address the problems of stability and control of aircraft in the high angle of attack regions of flight.
In most cases, the upper limit of normal flight is associated with conditions for maximum lift, beyond which the wing is completely stalled. For some aircraft configurations, however, for example, those employing wings with high leading edgesweep angles or incorporating strakes, i.e. a continuous band of plates on the fuselage, partial flow separation of the wing or control surfaces may induce stability problems below the attack angles for maximum lift and impose lower limits on the normalflight regions.
Solutions of these problems will allow flight above these lower limits. Flight above the normal limits is considered of a supernormal nature."
later see what the say
In piloted supernormal flight of the aircraft of the present invention, the wing of an aircraft, such as a superagile tactical fighter, is either partially or completely stalled, while the longitudinal control surfaces, such as in a rotatablecanard arrangement, are deflected to approximately the same magnitude, but of opposite sign, as the angle of attack of the aircraft, so that the canard arrangement remains effective to control the aircraft through large ranges of angles of attack, pitch,and flight path. Such angles may vary from descending flight to deep stall, i.e. -45.degree., to ascending flight in vertical climb, i.e. +90.degree.
The thrust T, lift L, drag D, velocity V, and various angular attitudes associated with and acting upon the aircraft in piloted supernormal flight are best illustrated in FIG. 2C. In such supernormal flight, the aircraft operates at an attackangle .alpha. much greater than the angle of attack for maximum lift so that the fixed wing 15 is either completely or partially stalled while the canard surfaces 19 are deflected in a negative sense through the deflection angle-.delta..sub.c.
The absolute deflection magnitude of the canard surfaces 19 is approximately the same as the attack angle .alpha. for the entire aircraft so that such canard surfaces 19 are nearly aligned with the local air flow and are, therefore, unstalled. Thus , the canard surfaces 19 remain effective as lift surfaces in providing the required forces and moments for controlling the entire aircraft.
[video=youtube;YA8QXenJzgA]http://www.youtube.com/watch?v=YA8QXenJzgA[/video]
Check the J-20 is not flying with thrust vectoring, niether is close to a stalled wing situation the jet is basicly, executing a shallow turn What song`s paper is trying to address is something similar but he metions TVC nozzles, on that video J-20 does not have thrust vectoring niether flying at high AoA, the jet is just simply deflecting the canards like a heavy loaded Rafale would do in order to increase lift.
In my opinion today`s J-20 still is far from being the ideal jet Song`s paper describe because it lacks TVC nozzles and without it the contradictions of stealth and aerodynamics represent a higher performance penalty
