On Aircraft Lift and Drag Reduction Using V Shaped Riblets

Reducing drag during take-off and nominal (cruise) conditions is a problem of fundamental importance in aeronautical engineering. Existing studies have demonstrated that v-shaped symmetrical riblets can effectively be used for turbulence control, with those with dimensionless depth h = 15 and dimensionless width s = 15 having the best drag reduction effect. In the present study, experimental tests have been conducted considering two models of the same size, one with smooth surface, the other with v-shaped riblets of the h = 15 and s = 15 type. The results show that for an angle of attack in the 8°~20° range (take-off stage), the maximum lift coefficient can be increased by 22%. For angle of attack between 8° and 14°, a drag reduction effect can be produced using riblets, which increases with the Reynolds number, leading to a decrease in the drag coefficient maximum of 36%. Flow visualization experiments have been carried out by means of Laser Induced fluorescence.


Introduction
Since the 1970s, NASA Langely research centers discovered that tiny grooves in the flow direction of surfaces could effectively reduce wall friction, and researchers have carried out a lot of continuous research on drag reduction of riblets [1][2][3]. At present, according to the literature review, there is a conclusion: for the physical model of flat plate, the riblet of symmetrical V shape has the best drag reduction effect. When dimensionless sizes of the riblet are: the height h + ≤ 25 and spacing s ≤ 30, drag reduction effect will occur. Moreover, when the range of the height of riblet, h + , is 8 to 15, the drag reduction effect was the best, which could reduce the turbulence frictional resistance by 3%~8% [4,5]. And there were many experiments that illustrated the phenomenon. For the Airbus plane with riblet surface, effect of 1%~2% fuel saving is achieved. Second, in the 1980s, German transport manufacturers applied an riblet surface to the plane which leading to 8% decrease in fuel cose [6][7][8]. Recently, Abdulbari et al. [9] studied the wall pressure fluctuations over the riblets using experiments. they found that spaced triangular riblets worked better than the triangular riblets.
Ho et al. investigated the stability of the laminar flow over streamwise riblets and oblique riblets experimentally [10]. Xie et al. [11] measured the drag acted on the van which was mounted with the riblet film, and drag-reducing effect can be observed.
In the actual flight of the transport, the take-off and climb stege focuses on lift augmentation, while the cruise phase concentrates on the drag-reducing effect. In recent years, the control of lift augmentation and drag reduction of riblet structure are also the research hotspots [12][13][14]. To quickly and effectively verify the lift augmentation and drag reduction effect of V-shaped flow riblet surface, the project team processed two 1:40 models in the systematic research on the aerodynamic characteristics of the conveyor model. In this work, the two models are tested. One is the transport with V-shaped symmetrical riblet on the surface including the body and wings. And the other is smooth surface. Under the same experimental conditions, the force measurement experiment was carried out to verify the lifting and drag reduction effect of the Vshaped symmetrical riblet surface, besides, the flow visualization experiment was carried out to demonstrated flow control effect of the riblets visually. And in the following parts, the force measurement experiments and the flow structure display experiments will be illustrated.

Experimental System
The experimental system consists of wind tunnel, water tunnel, transport model, force balance, data acquisition system, data analysis software, continuous laser and fluorescent agent, etc. The experimental test and data processing were carried out according to national industry standards. The force measurement experiments were carried out in the same wind tunnel whose size is 1.8 m (Width) Â 1.4 m (Height). And the flow field visualization experiments were implemented in a 1 m Â 1 m water tunnel. For a plane, the wing contributes to the lift mostly and the contribution of the part of body to the drag component is the most among all parts.
In this work, sizes and the configuration of the riblet are designed carefully, and under the experimental condition, the dimensionless height h + is 15, besides which the dimensionless spacings of riblet is 15. In actual machining operation, there is a smooth transition between the riblet and the surface of the planes.

Wind Tunnel
The current force measuring experiment was conducted in a 1.8 m Â 1.4 m wind tunnel. The wind tunnel is a low-speed wind tunnel with open/close experimental section of backflow mode. The flow field indexes of the wind tunnel are shown in Tab. 1.

Model and Support System
The model used in this experiment is the standard transport model with high-aspect ratio. The model was designed by the research team and processed by the First Transport Design and Research Institute of China Aviation Industry Corporation I. The scale of the model was 1:40 with whole metal structure. The profile drawing was shown in Fig. 1.
The model was supported by a belly strut in the 1.8 m Â 1.4 m wind tunnel.

Force Balance
TG0201 bar structure force balance was used for force balance (as shown in Fig. 2), and the balance was installed in the model, as shown in Fig. 3.

Measurement and Control System
The 1.8 m Â 1.4 m wind tunnel measurement and control system is based on computer network, mainly including data acquisition system, control system of angle of attack, velocity pressure control system, data processing system, experimental scheduling system and experimental analysis system. The data acquisition system was based on PXI data bus, which could be used for high-precision static data acquisition and high-precision and high-speed dynamic data acquisition. The wind tunnel had a velocity pressure control accuracy of less than 0.3% and an angle control accuracy of 0.05°.

Water Tunnel
The 1 m Â 1 m low-speed water tunnel was a low-speed backflow water tunnel of which the test section has a completely open upper wall, as shown in Fig. 4. The water tunnel was horizontally arranged with the central axis of the loop of 18.71 m Â 3.2 m and the contraction ratio is 9.3. The water tunnel consisted of an experimental section, an outlet sinking section, an axial flow pump power group, a reflux section (pipeline), a burst expansion energy dissipation section, a pre-contraction second-corner section, a steady flow and whole flow section, and a contraction section. A trailer with adjustable moving speed was arranged on the upper side of the experimental section. The whole loop of the tunnel is made of 3.4# stainless steel. And the lower wall and side wall of the experiment section were tempered glass of which the light transmittance is better than 0.9.
The range of the flow speed of water tunnel in the experiment section is 0.1 to 1 m/s. And the common flow velocity is 0.1 to 0.5 m /s, besides, the turbulivity in the model area was ε ≤ 0.1% (0.1~0.3 m/s), ε ≤ 0.2% (0.3~0.5 m/s) and ε < 0.5% (0.5~1 m/s).

Model Inspection and Installation
During the preparation of the experiment, the model should be assembled and tested on the platform, including the measurement of the installation position of each component of the model and the measurement of the installation position of the balance, etc. The measurement results will be recorded for subsequent verification.
Before experiment, the balance and the relative instruments need to be calibrated. If conditions permit, the balance should be installed on the ground for adaptability test. After the balance/model system is installed into the test section, necessary adaptability tests should also be conducted to ensure the system is installed in the proper position and correct the installation error.
The model is installed in the center of the 1.8 m Â 1.4 m wind tunnel section with a belly strut system.
And the initial attitude angle error should be less than 3'. After the model is fixed and connected with the strut, the initial roll angle may be generated. If the roll angle was large, the test results should be corrected accordingly. The initial elastic angle of the support system in the direction of attack and roll should also be measured after the model is installed.

Data Collection
The experimental data were collected by PXI system. The test was completed by step method. The sampling method was tentatively set as: sampling delay of 5 s, sampling time of 8 s, sampling frequency of 500 Hz in each channel. The noise reduction method based on the wavelet analysis theory was used to process the original signal.

Axis Definition
Experimental results shafting was defined as follows: Model body shafting: the origin was at the model moment reference center, the X axis was parallel to the model axis, pointing in front direction. And the Z axis was inside the model symmetry plane, perpendicular to the X axis, pointing down, besides, the Y axis was perpendicular to the ZOX plane, pointing to the right.
Model wind shafting: the origin was the same as the model body shafting, the X-axis was opposite to the incoming flow direction, the Z-axis was in the symmetry plane of the model, perpendicular to the X-axis and pointing down, and the Y-axis was perpendicular to the ZOX plane and pointing to the right.
Balance coordinate system: the origin was in the center of the balance, the X-axis was parallel to the model axis, pointing in front direction, the Y-axis was inside the model symmetry plane, perpendicular to the X-axis, pointing up, and the Z-axis was perpendicular to the XOY plane, pointing to the right.

Data Storage Format
Three sets of experimental results were given as the form of the six-component data, one set in the form of model body shafting, one set in the form of model wind shafting, one set in mixed shafting, lift coefficient and drag coefficient in model wind shafting, and the other four quantities in model body shafting. The data storage format was stored in 8 columns according to the conventional data format of force measurement experiment: The model body axis coefficients were arranged in the following order: C N , C A , C m , C Y , C n , C l , α and β where C N denotes the normal force coefficient, C A denotes the axial force coefficient, C m denotes the pitching moment coefficient, C Y denotes the transverse force coefficient, C n denotes the yawing moment coefficient, C r denotes the rolling moment coefficient, α denotes the angle of attack, and β denotes the sideslip angle.
The wind axis coefficients of the model were arranged in the following order: C L , C D , C ma , C c , C na , C la , α and β. Where C L denotes the lift coefficient, C D denotes the resistance coefficient, C ma denotes the pitching moment coefficient, C c denotes the lateral force coefficient, C na denotes the yawing moment coefficient, C la denotes the rolling moment coefficient, α denotes the angle of attack, and β denotes the sideslip angle.
The mixed axis coefficients were arranged in the following order: C L , C D , C m , C Y , C n , C l , α and β. Where C L denotes the lift coefficient, C D denotes the drag coefficient, C m denotes the pitching moment coefficient, C Y denotes the transverse force coefficient, C n denotes the yawing moment coefficient, C l denotes the rolling moment coefficient, α denotes the angle of attack, and β denotes the sideslip angle.
As for the final test data, the processing data should also be provided to correct the errors cause by the bracket and the hole in wall. Hence, the following data is needed, they are the results of unfastened bracket and the wall hole interference correction, the correction of the existence of fastened bracket, the results of only hole wall interference correction, and the data of hole wall interference correction and fastened bracket.

Data Processing Methods
The experimental data processing was carried out according to the conventional method, including deducting the initial reading, balance load calculation and angle correction, balance shaft rotation axis, conversion coefficient, fastened bracket interference, body shaft rotation axis, hole wall interference correction and data shafting conversion. There were some differences between the two steps of deducting the initial reading as well as the shaft of the balance shafting turning model, and the standard procedures.

Number of Blows-Initial Reading
The blow number minus initial reading was substituted into the balance formula, and the measured load of the balance was calculated by iteration (generally seven times) according to the requirements, with the unit of N.
It should be noted that, since the model attack angle changed when the wind tunnel worked, the data needed to be compared with that when the wind tunnel stopped. And the difference could be found after interpolation of the data from the balance system and data analysis system. The outputs of the balance included N T (normal force), A T (axial force), m T (pitching moment), Y T (transverse force), n T (yaw moment) and l T (roll moment).

Angle Correction
The angle correction formula is: α denotes the measured real-time angle of inclination sensor A1, β denotes the nominal side slipping angle of the model, Dα e and Dβ e are the balance elastic angles determined by the balance load calculated from the blow number minus the initial reading, respectively. Dα p and Dβ p represents the average flow deflection angles of the wind tunnel.

Rotary Axis of Balance Axis
Step 1: Coordinate translation. The origin of the balance shafting needs to be transformed to the origin of the model: Step 2: Coordinate frame rotation. eliminated the influence of installation error and elastic deformation angle. It should be clear that the model body shafting did not coincide with the balance coordinate system after coordinate rotation. The order was converted to β R , α R and γ R , i.e., where, F t = [-A t N t Y t ] T is the the force measured by the balance after deducting installation error and elastic deformation angle; F Tm = [-A tm N tm Y tm ] T , m mT rep the coordinate rotation transformation matrix, which could be expressed as, And this term is the representation of the moment measured by the balance after deducting the installation error and elastic deformation. By expanding the equation, where, the subscript "t" represented the expression form of the load measured by the balance in the model coordinate system. α totation , β totation and γ totation referred to the included angle between the model coordinate axis and the calibration axis of the balance (the fixed end of the balance), which could be decomposed into three parts, namely, Under the condition of no load, for α A , β A and γ A , the three initial installation angles of the model is measured [4]. And Dα 0 , Dβ 0 , and Dγ 0 denotethe elastic deformation angle caused by gravity of the model, which was calculated according to the initial condition and the state with model. The initial condition is representing that the model is installed in the wind tunnel at zero α A . And the zero reading is that the model is at an attack angle of 0°on the platform.
Dα elasticity , Dβ elasticity and Dγ elasticity denotes the extra angle of attack caused by deformation. When the data of attitude angle of the model and the change of the aerodynamic load of the model is obtained, the elastic deformation angle can be generated, which is as the following: When the ground test is finished, the zero position of the balance should be recorded. Before the measurement, the model should be installed on the platform.

Model Wind Shafting to Model Body Shafting Conversion
The calculation formula of the coefficient was as follows:

Analysis of Test Results
Figs. 6 and 7 show that the lift and drag of different scaled models in different wind tunnels. The results suggest that the experimental results of different scaled standard models in different wind tunnels are in accordance with the theory. C Lmax and α cr increase with the experimental Re number, and C Dmin shows a decreasing trend, because the maximum lift coefficient and the minimum drag coefficient were closely related to Reynolds number.
Figs. 8 to 15 showed the lift and drag characteristics of smooth and riblet-surface model under different experimental wind speed conditions. It could be observed from Fig. 8 that under the incoming flow wind speed of 30 m/s, the smooth surface and the riblet surface had the same trend in lift coefficient. Under a small attack angle, the riblet surface had lift-augmentation effect, with a 6% increase in maximum lift coefficient. It could be identified from Fig. 8 that under the condition where the wind speed was 30 m/s, the flow channel had no obvious drag reduction effect.  Under the attack angle of 6°~16°, the riblet surface had obvious lift-augmentation effect, with a 6% increase in maximum lift coefficient. It can be seen from Fig. 9 that at an attack angle of 8°~10°, the flow riblet had a drag reduction effect, with a 7% decrease in drag.
It could be identified from Fig. 10 that under the incoming flow wind speed of 50 m/s, the smooth surface and the riblet surface had the same law of lift coefficient curve. Under the attack angle of 6°~18°, the riblet surface had obvious lift-augmentation effect, with a 9% increase in maximum lift coefficient. It can be seen from Fig. 11 that at an attack angle of 8°~12°, the flow riblet had a significant drag-reduction effect, with a 7% decrease in drag.  Figs. 16 to 20 showed the experimental results of flow structure around smooth wing surface and riblet wing surface at the same experimental velocity and different attack angles. It could be observed from Fig. 18 that there was a classical flow instability structure in the wake field of smooth wing surface at the attack angle of 0°. In the same experimental condition, the flow instability was inhibited on the riblet wing surface and in the wake field structure.
It could be observed from Fig. 17 that under the attack angle of 8°, the flow structure around the smooth wing surface was clear, the flow was separated from the leading edge of the wing, and the vortex structure in the wake field was clear. In the same experimental condition, the reattachment structure appeared in the leading edge separation line of the riblet wing surface, and the vortex structure in the wake field was stable.      Fig. 20 indicates that under the model's attack angle of 14°, in the wake field of a smooth wing, the vortex rupture was advanced from the trailing edge to the leading edge of the wing. The vortex breakdown position in the wake field of the flow channel wing approach to the trailing edge of the wing.

Conclusions
1) The force measurement experimental results illustrate that the lift-drag characteristics of different models and different wind tunnel transport models were consistent, and the experimental results were reliable.
2) Under the condition of flow speed of 30 m/s, 40 m/s, 50 m/s, 60 m/s and 70 m/s and within the attack angle of 8°~20°, the V-shaped flow riblet wing had an obvious lift-augmentation effect, which increased with Reynolds number. When the flow speed is 70 m/s, the increase of lift coefficient was largest with a 22% increase in maximum lift coefficient, and the angle of attack was in the phase of take-off climb. 4) In the range of attack angle during take-off climb, the V-shaped symmetrical flow riblet surface had obvious inhibitory effect on the wing surface separation and the vortex breakage in the wake field.
Funding Statement: The authors received no specific funding for this study.

Conflicts of Interest:
The authors declare that they have no conflicts of interest to report regarding the present study.