1. Introduction
    The Circulation Control Airfoil, Wing or Rudder (CCA, CCW or CCR) is a high lift wing which employs a ' thin tangential jet blowing over the upper surface of a rounded trailing-edge to yield very high lift augmentation [1,21.  According to potential flow theory, the circulation around the section of CCW can have any arbitrary value, that is, potential flow solutions, with finite velocity every where, exist for any value of the circulation.  In viscous flow, on the other hand, separation occurs near the training edge, and the circulation assumes a definite value which depends on the separation positions of the upper and lower surface flow.  However, if a thin tangential jet is injected into the upper surface boundary layer near the trailing edge, the thin jet remains attached to the wing's rounded trailing edge due to the balance between centrifugal force and the pressure difference produced by the jet velocity. This phenomenon is called "Coanda effect".  The jet moves the separation point around the trailing edge toward the lower surface of the wing.  This produces a net increase in the circulation of the wing. Circulation control by jet blowing has been achieved.
    Circulation control can realize very high ratios of lift coefficient to jet momentum coefficient(C_L/C_m), together with good lift-to-drag ratios.  Furthermore, the lift developed by a circulation-controlled section can be easily and rapidly changed simply by varying the supply of blowing water.  The section shape facilitates the attainment of relatively high structural stiffness.  This and other advantages are particularly attractive in the design of lifting rotors, fixed-wing aircraft [3-6], ship's rudder, ship's propeller etc.
    The concept of the Circulation Control Airfoil was developed in England [7,8].  Till now, most research of CCW concentrated in applying it to aircraft.  From early of 80's, some researchers of Wuhan Transportation University engaged in numerical simulation and wind-tunnel test in order to apply CCW to ship [9-11].
    This paper describes the experimental investigation in a circulating water channel to study a circulation control method by using blowing to increase the lift characteristics of a low aspect ratio rudder. This method could generate large-lift force at small or even zero angle of attack while using only small amounts of blowing water. Besides, this paper also presents the results of the effects of two endplates and the longitudinal V grooves.
 
 

2. Description of the model
    A NACA0015 airfoil profile was chosen as the basic model for the investigation of CCR.  For achievement of circulation control, the model's trail was cut and replaced by Coanda surface.  The model had a chord of 247.4mm and a span of 228mm.  The blowing slot consisted of a uniform 0.45mm opening along the span at upper surface of trailing edge.  The blowing water entered the model through a pipe and passed through a duct in the model before exiting through the slot.  The slot exit was directed parallel to the chord and, consequently, made tangent to the Coanda surface.  The experimental investigation have demonstrated that the lift augmentation for a low aspect ratio CCR is much less then that of a 2D CCR.  This is due to the 3D effect that a larger jet momentum loss should be expected than in the 2D CCR.  In order to reduce these effects, endplates were fixed on both ends of CCR's span. Figure 1 shows the size of endplate.

Fig. 1. Configuration of CCR with endplates.

    The results also show the CCR produced high drag augmentation with increased jet blowing.  Howard [12,13] and Quass [14] found that longitudinal grooves reduced the drag of bluff bodies and the effective separation point was forced around the shoulder to the lower surface of the body.  In this test, longitudinal V grooves were used to decrease drag augmentation.  Figure 2 shows the position and the size of the longitudinal V grooves.
 

3. Water-channel Apparatus and Test Procedure
    The CCR was tested in the rectangular open surface circulation water channel in Wuhan Transportation University.  The length of test section is 6m, the width is 1.8m, the depth is 0.9m. Load cell transducers of the strain-gauge-beam type were used to measure the lift and drag force. An IBM computer was used to pick up data of lift and drag.  The water of the jet was provided by a rotary pump and the flow of the jet was measured by a flowmeter.
 

Fig. 2. Construction of longitudinal V grooves.

    In the present study, the characteristics of the CCR at zero degree angle of attack were obtained first.  The blowing rates were varied to provide a jet momentum coefficient range from 0 to 0.63. Then the angle of attack increased from 0 to 24 degree in 6-degree increments and the jet momentum coefficient changes from 0 to about 0.21.
    After that, the endplates were installed.  The angle of attack and the jet momentum coefficient were changed as these of CCR's test.  Finally, the Coanda surface of the CCR was replaced by that with longitudinal V grooves.  The experiments were carried out with the B, indicating the position of grooves on Coanda surface, of 40, 62, 81, 99 degree and a groove depth of 0.5, 0.78, 1.5, 1.9mm.
    The experimental data obtained during the tests were reduced to three section parameters: the jet momentum coefficient C_m , lift coefficient C_L and drag coefficient C_D. With these parameters, the lift-to-drag ratio L/D could be obtained too.
 

4. Results and Discussion
4.1. Characteristics of the CCR
    The characteristics associated with trailing-edge blowing of the CCR at zero degree angle of attack are shown in Figure 3. It is evident that as the jet blowing increased, the lift increased significantly and the lift curve slope decreased.  A jet momentum value was reached, however, where the lift augmentation ratio was zero and then the lift decreased abruptly with further increasing in jet momentum.  This phenomenon is called C_m-stall [15].  The lift-to-drag ratio of the CCR is shown to increase rapidly with blowing, reach a maximum, and then decreasing with further increase in blowing rate.  As C_m stall happened, the drag curve and the lift-to-drag ratio curve had a marked drop.

Fig. 3. Characteristics of CCR at zero                Fig. 4. Lift as a function of angle of attack. Momentum for CCR.

Fig. 5. Drag as a function of jet momentuni for CCR.   Fig. 6. Lift-to-drug ratio as a function ofter momentum for CCR.

    The effect of blowing ratio at some angle of attack on lift, drag and lift-to-drag ratio are shown in Figures 3. 4, 5 and 6, where the lift of the CCR generally increased with increasing jet momentum, whereas at zero degree angle of attack, the drag curve went down first and then went up with increasing values of jet momentum, but at relatively large angle of attack, the drag decrease was not clear.  As the angle of attack increase, the slope of lift-to-drag ratio curve decreased.  The maximum of lift-to-drag ratio occurred with zero degree angle of attack and the jet momentum of about 0.055.
    Figures. 7 and 8 show the effects of the angle of attack on lift and drag.  The lift curve is approximately linear with the angle of attack and by the introduction of jet blowing, the angle of attack stall (or alpha stall) occurred at an angle of attack at which the lift of the CCR without jet blowing went up continuously.  The drag curve changes from a nonlinear function of angle of attack at zero jet momentum to a roughly linear function at relatively large jet momentum.

Fig. 7 Lift as a function of angle of attack for CCR.        Fig. 8 Drag as a function of angle of attack for CCR.

4.2. Effect of Endplates
    Figures. 9-11 shows the results of the CCR with endplates.  The endplates clearly limited the flow in spanwise direction, so the lift obtained in this case without blowing water was greater than these of the CCR in the same condition, but the drag was less. At relatively small angle of attack, the model's lift became worse with endplates, and at large angle of attack, the result was reverse. Except of small angle of attack and jet momentum, endplates resulted in lower drag and drag curve slope. Figure 11 shows that large values of the lift-to-drag ratio could be obtained by CCR with endplate.

Fig. 9. Lift as a function of jet  momentum for             Fig. 10.  Drag as afunction of jet momentum for
CCR with endplates.                                                  CCR with endplates.

Figure 11.  Lift-to-drag ratio as afunction ofjet momentumfor CCR with endplates.

    The data of the CCR with endplates show approximately linear curves of lift coefficient versus angle of attack (see Figure 12).  For the CCR with endplates, the lift curve slope was larger than that of the CCR and the alpha stall did not occur in this test. Figure 13 indicates that a higher jet momentum produced a higher drag augmentation with growing angle of attack.

    The effects of endplates are following:
1) The endplates obstructed the flow in spanwise direction, while the strength of the end vortices decreased.  So the effective angle of attack increased.
2) Endplates obstructed the flow divergence of jet, more blowing water was attached to the Coanda surface and the Coanda effect became more efficient.
3) Some blowing water nearby the endplates did not adhere the Coanda surface. Figure 14 shows the shape of jet wake in still air.

Fig. 12.Lift as a function of angle of attack for              Fig. 13.Drag as a function of angle of attack for
 CCR with endplates.                                                   CCR with endplates.

Figure 14.  Shape of the jet wake in still air.

4.3- Effect of Longitudinal V Grooves
    The effects of the groove’s position and depth on lift, drag and lift-to-drag ratio at zero angle of attack are shown in Figure l5-20.  It can be seen in these figures that:
1) Longitudinal V grooves reduced the drag of the CCR evidently, and the larger jet momentum was obtained the more significant the effect of grooves were produced.
2) C_m-stall of the CCR without grooves appeared at a jet momentum coefficient of 0.52. By using grooves, Cu-stall did not occur as the jet momentum coefficient went up to 0.62.
3) The rudder's lift and drag decreased as grooves were used.  But as jet momentum coefficient was greater than 0.14, the effect of grooves on drag was more significant than that on lift, so the lift-to-drag ratio of the CCR with grooves was greater than that of the CCR.  The maximum of lift-to-drag ratio was reached at B of 99 degree and jet momentum coefficient of 0.17.
4) For the CCR with grooves, a higher lift then that of the CCR could be achieved, but more jet momentum was needed.

Fig 15 Effect of groove's position on lift.                       Fig 16 Effect of groove's position on drag.

Fig 17 Effect of groove's position on lift-to-drag ratio.     Fig 18 Effect of groove's depth on lift.

Fig 19 Effect of groove's depth on drag.                         Fig 20 Effect of groove's depth on lift-to-drag ratio.

5. Conclusions
1) A high lift coefficient was produced by the CCR with trailing-edge jet blowing at zero degree angle of attack, and the lift coefficient of the CCR at a fixed angle of attack increased notably with jet momentum.  The maximum of lift-to-drag ratio occurred at zero degree angle of attack.  Based on this, it could be used in ships to improve the maneuvering performance, especially when the ship sails at low speed.
2) Lift of the CCR did not always increasing with jet momentum.  When reaching its maximum, the lift decreased abruptly with further increasing jet momentum, which is called C_m-stall.  The lift curve slope decreased as jet momentum increased before C_m-stall. At zero degree angle of attack, the drag decreased first and then increased and at large angle of attack, and the drag generally increased as jet momentum value went up.  The lift-to-drag ratio increased at first and reached a maximum, then decreased with increasing jet momentum.  When C_m-stall appeared, the drag and lift-to-drag decreased abruptly.
3) With endplates installed, the drag and lift-to-drag ratio characteristics were improved, and the lift at relatively large angle of attack increased.  The alpha stall was postponed and the lift curve slope became large when using endplates.
4) By introduction of longitudinal V grooves, the lift and drag was less then that of the CCR, but except of the cases of relatively small jet momentum, the lift-to-drag ratio was greater.

References
[1] Abramson, J., and Rogers, E. O., High-speed of Circulation-Controlled Aerofoils, ALKA Paper 83-0265, 1883.
[2] Nielsen, J. N. and Bigger, J. C., Recent Progress in Circulation Control Aerodynamics, A-IAA Paper 87-0001, 1987.
[3] Eular, R. J., Circulation Control for High Lift and Drag Generation on STOL Airfoil, Journal of Aircraft, Vol. 12, No, 6, 1975, PP. 457-463.
[4] Eular, R. J., Trobaugh, L. A. and Hemmerly, R. A., STOL Potential of the Circulation Control Wing for High Performance Airfoil, Journal of Aircraft, Vol. 15, No. 3, 1978.
[5] Nichols, J. H. Jr., Eular, R. J., Harris, M. J. and Hason, G. G. , Experimental Development of an Advanced Circulation Control Wing System for Navy STOL Airfoil, ALKA Paper, 81-8151, 1981.
[6] Loth, J. L., Circulation Control STOL Airfoil Design Aspects, NASA CP-2432, 1986.
[7]    Kind, R. J. and Maull, D. J., An Experiment Investigation of a Low-speed
Circulation-Controlled Aerofoil, The Aeronautical Quarterly, Vol. 19, May 1968, pp. 170-182.
[8] Cheeseman, I. C. and Seed, A. R., The Application of Circulation Control by Blowing to Helicopter Rotors, Journal of the Royal Aeronautical Society, Vol. 71, July 1966.
[9] Wang Xianfu and Xiong Xinmin, A Numerical Method for Solving the Circulation Control Airfoil with Wall Jet, Acta Aerodynamics Science, Vol. 9, No. 3, 1991.
[10] Wang Xianfu, Huang Jinwen and Pan Weimin, Numerical and Experimental Study of Circulation Control Rudder, Journal of Hydrodynamics, Ser.  A, Vol. 7, No. 1, 1992.
[11] Pan Weimin and Wang Xianfu, A Study of Application of Circulation Control Wing on Sail Assisted Ship, 8th International Conference on Wind Engineering, Canada, 1991.7.
[12] Howard, F. G., Quass, B. F.,Weinstoin, L. M. and Bushnell, D. M. Longitudinal Afterbody Grooves and Shoulder Radiusing for Low-Speed Bluff Body Drag Reduction. ASME Paper 81 -WA/FE-5,NOV. 198 1.
[13] Howard, F. G. and Goodman, W. L., Axis-symmetric Bluff-Body Drag Reduction Through Geometrical Modification.  Journal of Aircraft.  Vol. 22, No. 6, June 1985.
[14] Quass, B, Howard, F. G., Wrinstoin, L. M. and Bushnell, D. M. , Longitudinal Grooves for Bluff-Body Drag Reduction.  AIAA Journal, Vol. 19, April 1981
[15]  Wood, N. J. and Nielsen J. N., Circulation Control Airfoil--Past, Present, Future, AIAA Paper 85-0204, Jan. 1985.