FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1955
1955 - 1436.PDF
546 NO THRUST ij NET THRUST THE JET FLAP . . . PRESSURELOOP JET TURNEDHORIZONTAL momentum flux line and the flow is everywhere potential, theparadox of d'Alembert must apply, and this gives rise to the problem that in an idealized two-dimensional jet flap systemthe gross thrust is equal to the total jet reaction and independent of the angle of deflection of the jet. ''On account of the jet mixing process this identity cannot holdexactly in practice, there being for this two quite different reasons. Firstly, at large values of the thrust coefficient the mixing processwill involve a momentum loss which must appear as a loss of thrust. Secondly, since the mixing occurs generally in a regionof varying pressure the system may function as an introverted thrust augmentor, so that the total thrust can in theory be eithergreater or less than the jet reaction. "Usually the losses exceed the gains, but experiments have beenmade in which the total thrust was greater than the total reaction. In such circumstances the excess thrust appears, of course, as aform, or pressure, thrust." Work at the N.G.T.E. began in November 1952, but previousresearch along these lines dates back at least to 1938. In this year the German aerodynamicists Hagedorn and Ruden wereworking in Hanover on boundary-layer control, in the course of which the effect of large quantities of ejected air at the trailingedge of the wing became apparent. They published a report, but this did not become available in Britain until after 1945.Similar work is also known to be going on at the O.N.E.R.A. in France, two of the principals in this programme beingM. Jousserandot and his chief M. Poisson-Quinton. During their work on jet deflection for the experimental Meteorthe N.G.T.E. decided to attempt to discover a more efficient method of increasing lift by deflecting the jet. The jet flapreally stems from this. The first model made was a simple aerofoil of brass sheet bentaround a tube and soldered. Air was blown from a downward- facing slit along the trailing edge, and the results proved theincrease in lift but failed to show any induced thrust. Accordingly, a better induced-thrust model was made, capable of showingaccurate lift and pitching moments for comparison with the theory from the first model.In potential-flow (no-loss) theory, forward thrust is not a func- tion of jet angle. With the jet ejected at 90 deg, i.e., vertically, theN.G.T.E. measured about 30 per cent of forward thrust; this was, however, a crude model, and considerably better results arepossible. Later it is expected that it should be possible to recover half the theoretical form thrust (1 - Cos 6), so that at 60 deg tothe horizontal 75 per cent of the true thrust should be possible. A wing equipped with the jet flap has three main derivatives:dC L dCL dCL da ' d0 3na J is the jet reaction or momentum of fluid issuing at the spanwisenozzle (or, alternatively, the total reaction on the engine bearers). The thrust is then approximately J Cos 8 + \ J(l -Cos 6). Thisis an empirical formula, scarcely better than a hunch, but inaccuracies are not thought to be important in practice sincesmaller deflection angles would be employed. At very high lift conditions the jet acts as a physical entity which is very powerfulin comparison with the main stream. It operates at low speeds to entrain a considerable air flow and, even if separation occursabove the wing, re-entrainment continues due to sink effect. Research then continued with a truly elliptical wing sectionof 12^ per cent thickness, this haying a sufficiently "beefy" trailing edge in order to insert a spanwise plenum chamber and ejectionslit. An approximate area ratio of 40:1 was chosen between the plenum chamber flow and the flow through the slit. Maldistribu-tion of gas would naturally give bad spanwise distribution of lift and induced drag. The experiments were nominally two-dimen-sional, but it was possible to achieve induced drag, the circulation falling off in the tunnel boundary layer. On the second model, the chordwise slot measurement was0.018in and deflection angle was originally 90 deg, for the assess- ment of pressure and induced thrust. The maximum inducedthrust was found to be 37 per cent, at which point leading-edge separation occurred.One of the main aerodynamic effects which was experienced FLIGHT (Lett) mechanism by ythich thrust is produced by a jet issuing vertically downwards from the trailing edge. Variation of lift coeffi- cient with angle of inci- dence (angle of attack), i.e. dCh/d*. Jet angle in this case was 58.1 deg. Maximum lift occurs at zero angle of attack, so that an aircraft with such a wing would always fly in a level attitude. -1O -5° O 5° 1O° ANGLE OF INCIDENCE oc on the first series of tests was the sink or entrainment effect due tothe mixing process, this being found to operate even at 0 = 0 deg. Provision for cruising flight will have to be made, and it is clearlyimpracticable to have two jet exits. Accordingly, a third model (a thrust/drag model) was made. With this mbdel, considerablemixing effect and noticeable variations in static-pressure distri- bution were discovered around the wing, but it was demon-strated that these could be removed in cruising flight. A fourth model (a thrust-boost model) was then made to see if it werepossible to obtain increased thrust augmentation. Here the jet was split into multiple fishtails, but the experiment failed owingto induced drag and other effects. An accompanying diagram shows the mechanism by whichthe thrust is produced. The fact that, in forward flight, the jet is turned horizontally implies that there must be a thrust reactionon the wing. Thrust is provided by the strong curvature of the flow forwards along the under-side of the leading-edge and backacross the top of the wing. The main airstream is diverted upwards upstream of the leading edge, and the centre of the totallift is somewhere near 50 per cent chord. With a suitably designed wing the local flow should be unstallableup to a lift coefficient of at least 10. Applied to an actual aircraft some compromise would be essential and a rather different wingprofile would have to be adopted. Nevertheless, the term dCL/d* would not become negative until considerably after the onset ofseparation, owing to the jet impingement effect. Equally, the onset of instability would only occur at very high lift coefficients.Aircraft control would be primarily a function of jet thrust, the tailplane being relegated to the role of a stabilizing member.Thus, the pilot might not be able to put the aircraft into a stalling attitude. (concluded on page 558) ! —- The almost incredible distribution around a MVi per cent elliptical wing with jet at 58.1 deg to the horizontal. Tunnel velocity was 30ft/ sec and Reynolds number about half-a-million. Lift coefficient in the photograph was about 12, or almost double the limit (of 2x) set by the Kutta Joukowski hypothesis. With the jet flap, maximum lift coefficient is infinity. Note: the experiment was actually run the other way up.
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
