1.Introduction

Manoeuvrability of ships has been received a great deal of attention both concerning navigation safety and the prediction of ship manoeuvring characteristics, especially at the preliminary design field.

Since 1970, large size and high-speed ships have appeared in the world, which led to poor the manoeuvring capability, consequently leading to many ship casualties. The IMO has recommended that ship designers from the initial design should evaluate manoeuvring. In the meantime, high-lift devices could be applied to design of rudders in order to improve control stability of ships.

It is necessary to evaluate the correct performance of rudders for improvement of ship’s manoeuvrability. Rudder open water test can not be studied in this study, but it is considered PIV analysis of two special rudders; a flapped rudder and a water-blowing rudder. It is also for flow analysis around special rudders.

So it carried out flow visualization and investigation of flow field around the flapped rudder and the water-blowing rudder using PIV technique. The flapped rudder is simply a portion of the trailing-edge section of airfoil, that is hinged, which can be deflected upward or downward. And the water-blowing rudder indicated “Coanda effect” which increase lifting force better than the traditional single rudder. the Coanda effect has the tendency of a jet of fluid emerging from an orifice to follow an adjacent flat of curved surface and to entrain fluid from the surroundings so that a region of lower pressure develops.

Daichin et al. (2007) investigated the PIV measurements of the near-wake flow of an airfoil above a free surface in a wind/wave tunnel at Re≒3.5×10³ using PIV system. The lift force and pressure drag acting on the airfoil increase when the airfoil is getting closer to the free surface.

Wong et al. (2006) studied on a supercritical airfoil with a Coanda trailing edge at subsonic speeds. The forces measurements of a Coanda airfoil with steady and unsteady blowing were compared with the same dimensions supercritical airfoil with hinged flap. The lift to drag ratio obtained on the Coanda airfoil with blowing is significantly higher than the conventional airfoil. And the application of steady and unsteady blowing on subsonic airfoils can be effective means of circulation control, reducing drag and increasing lift.

Wong et al. (2006) studied on a supercritical airfoil with a Coanda trailing edge at subsonic speeds. The forces measurements of a Coanda airfoil with steady and unsteady blowing were compared with the same dimensions supercritical airfoil with hinged flap. The lift to drag ratio obtained on the Coanda airfoil with blowing is significantly higher than the conventional airfoil. And the application of steady and unsteady blowing on subsonic airfoils can be effective means of circulation control, reducing drag and increasing lift.

Traub et al. (2004) carried out an experimental investigation to evaluate the effectiveness of a synthetic jet actuator for flow control on a pitching airfoil. It is found that the variation of the synthetic jet actuator driving frequency showed the ability to control the flow from massively separated to fully attached at 25 deg instantaneous angle of attack.

In order to measure both the total lift and induced drag in three planar regions downstream of an airfoil having a NACA 0015 profile, Grant et al. (2006) compared with between a NACA 0015 wing and the wing with a 3.33 % Gurney flap attached to the trailing edge. The results compare favourably with past research using Pitot pressure techniques and CFD. This research proves that PTV is a viable tool in obtaining quantitative flow field characteristics rapidly. A natural progression of this research would be to obtained PIV and time-averaged data to allow a more exact comparison with CFD and traditional data collection methods. The advantage of PIV method is to provide the instantaneous information of wake behaviour, not available by other experimental methods or CFD in general.

Weier and Gerbeth (2004) investigated that the application of time periodic Lorentz forces to the control of the suction side flow on a NACA 0015 hydrofoil for Reynolds number range 5.2×10⁴ ≤ Re ≤ 1.5×10⁵. A specific lift increase with respect to the value for the separated flow can be oscillatory with fractions of the momentum input necessary for steady Lorentz forces. In contrast, an equal increase of the maximum lift gain requires a similar expenditure of momentum for both control methods. The Lorentz force allows for a great flexibility in providing the time dependency of the forcing.

Deng et al. (2007) developed a DNS code with high-order accuracy and high-resolution and investigated separation and transition processes on a NACA 0012 airfoil with and without jet blowing on the surface. The transition of turbulence on a 2D airfoil will reattach the separated boundary layer. The separation and stagnation points are closely related to the pressure distribution on the airfoil surface. The shear layer becomes turbulent and reattaches to the surface. This process is then further sustained by the global feedback mechanisms. In the large mass blowing case, they did not observe the unstable mode pick up mechanism, because transition takes place in a very short distance and the large mass flow of the blowing jets causes the bypass transition. In the pulsed jet with large mass blowing, the separation zone and the drag are reduced, but lift also decreased. With 30deg pitch and 90deg skew angles, the separation zone is reduced and drag is decreased while the lift maintains approximately the level is in the baseline case, which shows the pitch and skew angles are important to the efficiency of separation control by blowing jets.

Jang et al. (1998) investigated the two-dimensional numerical simulation to determine the effect of a Gurney flap on a NACA 4412 airfoil. The Gurney flap is a flap plate on the order of 1-3 % of the airfoil chord in length. They also compared between the computational results and the available experimental results. The numerical results show that some Gurney flaps increase the airfoil lift coefficient with only a slight increase in drag coefficient.

The purpose of the present paper is to investigate the influence of an flap and a water-blowing device on the flow around a foil at Re≒3.0×10⁴. Two-dimensional flow characteristics around NACA 0012 with the special devices were investigated. The experiment was carried out with 10 and 20 angles of attack to obtain the flow information around the foil using two-frame grey-level cross-correlation PIV method.

2.Experimental Arrangement

2.1.Coordinate system

The rudder model used for this study consists of NACA 0012 attached a flap. Experiments were conducted in a circulating free surface water channel. The coordinate system which is adapted through the whole experiments in order to measure two dimensional flow characteristics of flapped and blowing rudders is shown in Fig. 1. Angles of attack(α) and flap's deflection angles(δ) are varied 0 to 20 degrees respectively. And the controller has ±0.2 deg error to adjust the angles of attack and flap deflection angles.

The water-blowing rudder consists of the NACA 0012 and physical shape and installed a blower at 30 % of the chord length from the leading edge as shown in Fig 1. But there is no flap. The test models are made of acrylic acid resin, so that laser light propagates easily from the model.

2.2.Circulating Water Channel

The experiments were carried out in a small circulating water channel. The channel has a section of 1000 mm(L) × 300 mm(W) × 300 mm(H) which is made from 15mm thick transparency plexiglass sheet. The water as a working fluid used in this experiment was kept at a constant value of 20±1°C, fresh water. The water pump is driven by an electric motor with a variable speed controller. The flow is driven by the effect of gravity under atmospheric conditions in room temperature. The model is installed in the middle of the channel which has a distance h = 30 mm between the free surface and bottom. The experimental configuration used in this study consists of flow imposed around a rudder at a Reynolds number (Re≒3.0×10⁴) based on the chord length, 150 mm.

The overall of the circulating water channel with model installed is shown in Fig. 2. The distribution of u-components is more confusion near the bottom than the free surface of the channel. That is the reason why the Reynolds number should be Re≒3.0×10⁴ in this study. It is tried to find out the minimum interaction between the bottom and the free surface of the channel while retaining high flow. The test section, the flapped rudder consists of NACA 0012, 150mm in length attached a flap at two third of the chord length near the trailing edge.

2.3.Particle Image Velocimetry (PIV)

PIV is a powerful optical method used to measure velocities and related properties in fluids. PIV has been applied to a wide range of flow problems, varying from the flow over an object in a wind tunnel to vertex formation in heat valves. The algorithms based on two-frame grey-level cross-correlation (CACTUS 3.1) can be implemented in a matter of hours. In this study, total of 500 frames divided into 125 frames per second are calculated to measure the velocity flow field. The fluid is seeded with particles which, for the purposes of PIV, are generally assumed to faithfully follow the flow dynamics. It is the motion of these seeding particles that is used to calculate velocity information. Measurements were conducted using PIV system that consists of an optical LED laser (0.5 watts), CCD camera and a grabber. A CCD digital camera with the resolution of 1280 × 1024 pixels recorded the movements of the particles. An optical arrangement to convert the laser output light to a light sheet using a cylindrical lens. A fiber optic cable connects the laser to the cylindrical lens setup. The laser acts as a photographic flash for the digital camera. And the particles in the fluid scatter the light. It is this scattered light that is detected by the camera. The average ratio of error vector during the calculation procedure is less than 4 % in the result report base on CACTUS algorithms. This is converted to a velocity using the time between image exposures. The schematic arrangement of PIV system is shown in Fig. 3.

3.Experiments

Fig. 4(a), (b) are the u, v velocity distribution curves at the end of chord length apart from 10mm. The components are extracted 50 components through the vertical direction of x-axis. Fig. 4(a) shows u-velocity components through the vertical direction each angle of attack. It is found that the evaluation of u-components for 0deg and 10deg is stable situation. The values of 20 deg, 30 deg and 40 deg are not similar to the other in between 20 mm to 100 mm of y axis. The values quickly decrease below zero m/s. Fig. 4(b) shows v-velocity components also shown in the areas in Fig. 4(a). V-velocity components are almost not changeable through the whole cases.

NACA 0012 has its own flow separation at around 15 degrees of angles of attack. As the results of Fig. 4(a), we can consider the generation of flow separation in between 10 to 20 degrees. So the boundary layer will be thicker as the angle of attack are increased. It means that the lift effect on the NACA 0012 are decreased as the angle of attack are increased.

3.1.Visualization of Flapped rudder

Flow visualization was performed for a two-dimensional flapped rudder at condition which is Reynolds number, based on chord length, of Re=3.0×10⁴ using a high resolution digital camera. When the 20 deg of the angle of attack is fixed, Fig. 5(a)-(b) show that the particle traces increased downstream behind the flapped rudder, while changing flap’s deflection angles. In addition, the flow around the rudder at negative 40 deg of a flap’s deflection angle is shown in Fig. 5(d).

Fig. 5(a) depicts that the flow characteristics of a traditional NACA 0012 rudder without any appendage because the angle of attack is 20 deg and the flap’s deflection angle is 0deg. It is highly important to understand on the separation condition due to the pressure difference to find the flow profile around a rudder. It is well known that the separation starts at approximately 16 deg in the flow visualization. The separation point and velocities around the 2-dimentional rudder affect the lift and drag forces. But the variation of lift and drag forces due to the separation shifting do not take into account in this paper.

Fig. 5(b) shows the flow visualization at 10 deg of flap’s reflection angle. The lift force takes place the direction of y-axis due to the velocity gradient between the upper and lower surfaces.

Fig. 5(c) shows the flow visualization at 20 deg of angle of attack and 40 deg of flap’s reflection angle. The increase of boundary layer and the separation point appear at the front part of the rudder, leading edge. And it is also observed the flow separation below the rudder due to a steep flap’s deflection angle. In this case, the flow separation acting the angle of attack and the flap’s deflection angle have the wide vortex separation flow above the rudder and disturb the free stream the upper side of the flow. In comparison with Fig. 5(b), the wide region of the irregular stream area and the formation of vortex owing to the flap’s deflection angle can be seen in Fig. 5(c).

Fig. 5(d) shows the opposite side of the flap’s deflection angle of Fig. 5(c). The interaction between the main body and flap has a tendency to decrease the each boundary area. It is a good example to reduce drag force as control flap’s deflection angles.

3.2.Velocity analsys around Flapped rudder

Fig. 6(a)-(c) shows the mean velocity profile of the 2-dimensional flapped rudder using PIV. The whole cases are fixed 10deg angle of attack, then the flap’ deflection angles change step by step as δ = 0 deg, 10 deg, 20 deg and 30 deg respectively to look into the flow variation according to the changes of flap’s deflection angles. The effect of the flap’s deflection angles in the separation zone, mean velocity profile is different from the alteration of flap’s deflection angles.

The Flow profile and velocity distribution at α = 10 deg, δ = 0 deg is shown in Fig. 6(a). It means that the laminar flow of the pure NACA 0012 having a thin boundary layer. The flow in this case, there is no turbulence, flow separation and vortex. And so the small velocity gradient generates between the upper and lower surfaces.

In Fig. 6(b)-(d), According as increasing the flap’s deflection angles, wide and thick turbulent boundary layer generated and the flow separation point move to the leading edge. As the flap’s deflection angle increase, the boundary layer rises in proportion to a rise in the flap’s deflection angle. The boundary layer becomes fully turbulent after re-attachment.

Fig. 6(b) shows that the minor downstream appear at α = 10 deg, δ = 10 deg. A vortex creates on the upper side at the middle of the flapped rudder. Subsequently, boundary layer is bigger than Fig. 6(a) and the separation point moves to the leading edge.

Fig. 6(c) shows the wake behind the flapped rudder apart from more than one times of the rudder. A big vortex appears at the end of the trailing edge. The downstream owing to the flap’s deflection angle and the boundary layer are much bigger gradually.

Fig. 6(d) shows that three vortices appeared after the rudder. The flow pattern has become more complicated and also has wide range of boundary layer before and after the rudder and upper and lower surfaces.

As the flap deflection angles increased, the boundary layer and the flow was much more complicated from the leading edge. it is considered that the flap deflection angles affect the whole flow pattern despite the angles of the main foil because of the influence of turbulence flow around the training edge.

3.3.Water blowing rudder

A Water-blowing control rudder experimental device consists of NACA 0012 attached with a blower that 30 % from the leading edge at a Reynolds number Re≒3.0×10⁴ based on the chord length, 150mm. The reason why it has the blower at 30% from the leading edge is the best location of the blowing effect. The jet velocity classifies 1.5 m/s (low), 1.75 m/s (medium) and 2.0 m/s (high) into three types based on Reynolds number. Fig. 7(a) shows the v-components velocity contour at α=20 without blowing. The flow separation occurs at the leading edge. At this point, the u value at upper surface is lower than other region. In Fig. 7(b), when the jet velocity of 1.6 m/s projected, the flow boundary layer past the rudder gradually decreased. And also it is found that the surface connected with vortex region observed the backward flow. Fig. 7(c) shows that the boundary layer decreased over 50 % than the case of without blowing. The effect of the water-blowing is recoveried the flowstream at the end of trailing edge.

Fig. 8 indicates the u-velocity distribution curves at the point of x/C=1.5. The u-velocity components are extracted 50 components through the vertical direction of x-axis. It is found that the difference of u-components for blowing and non-blowing cases is quite sensitive near the trailing edge.

As the jet velocity increased, the separation point above the leading edge, comparison with the non-blowing cases, decreased relatively. It is found that the effect of water-blowing reduced the vortex generation. And also the effect of the water-blowing Coanda effect affects the flow at the end of the trailing edge.

4.Conclusions

I considered here the flow profile of uniform stream for two special rudders to evaluate the control ability.

The flow around the flapped rudder in a variety of flap’s deflection angles and the water-blowing rudder at 30 % of the chord length from the leading edge were visualized and studied by using the 2-frame cross correlation PIV method. The following results are summarised based on our observations and data analysis.