Figure 2 – uploaded by Aditya Saini

See full PDF downloadDownload figure

Figure 1.2: Air data boom of F-35C combat aircraft. Image source: http://commons.wiki- media.org/wiki/File:Cockpit_and_Air_Data_Boom_F-35C. jpg. — Figure 1 2: Air data boom of F-35C combat aircraft. Image source: http://commons.wiki- media.org/wiki/File:Cockpit_and_Air_Data_Boom_F-35C. jpg.

Related Figures (67)

Figure 1.1: A conventional airdata boom for measuring pressures and flow angles.[14 aircraft. dow angles is shown in Figure 1.1 and Figure 1.2 shows the air data boom of F-35C comba Although undoubtedly successful, these techniques have certain drawbacks. For instance,

Figure 1.3: F-18 Systems Research Aircraft closeup of nose cap showing new flush air data system sensor holes from http://www.dfrc.nasa.gov/Gallery/Photo/F-18SRA/HTML/EC97- 43936-6.html.

Figure 1.4: Estimated freestream velocities (U) using the maximum of the pressure measured at five ports for the S809 airfoil compared with the actual velocities (dashed lines).

2.2.1 Stagnation-Point Flow

2.2.2. Flow Past a Parabola The stagnation-point flow can be mapped onto the physical z-plane, using the transformation:

Figure 2.4: Parabola at an angle of attack. lifference between the surface pressure and the known freestream static pressure can be written

Figure 2.5: A comparison of the pressure distribution on a cambered $809 airfoil obtained from XFOIL with the pressure distribution over a parabola obtained from Eq. 2.25. Here, r and A in the equations are estimated by generating the parabolic curve based on the leading-edge region of $809 airfoil, as discussed in Chapter 3. of reference based on the inclination of the defined parabola that appropriately fits the airfoil

Figure 3.1: Leading-edge parabola fit for a symmetric NACA 0012 airfoil. 1. Start with an initial guess value of r and generate a symmetric parabola.

Figure 3.2: Leading-edge parabola fit for a cambered $809 airfoil.

Table 4.1: Pressure port locations used to extract data for LEFS inputs from CFD data. The airfoils analyzed using the CFD data are relatively thin with thickness ratio, t/c, less than from the CFD data for all the three airfoils are tabulated in Table 4.1.

Figure 4.1: Stagnation-point location, pressure-peak location, and pressure-peak value are compared from the LEFS algorithm (red) are compared with CFD results (blue) for the NACA 0012 at Re = 3 million. The dashed vertical line represents the stall angle of attack.

$D7003 Figure 4.3: Comparison of freestream velocity, angle of attack, lift coefficient, and suction coefficient estimated from LEFS (red) and the actual values from CFD (blue) for SD 7003 at Re = 3 million.

Table 4.2: Pressure port locations used to extract data for LEFS inputs from experimental data. The airfoils considered in the experimental test cases have greater thickness compared to

Figure 4.4: Airfoil profile of NACA 0012, LS 0417, and S809. test cases. Therefore, look-up table approach appears to be the appropriate method for thicker

Figure 4.5: Comparison of velocity, angle of attack and lift coefficient from LEFS and wind tunnel, for Re = 0.75 (a) and 1 million (b) for LS0417 airfoil.

Figure 4.6: Comparison of velocity, angle of attack and lift coefficient from LEFS and wind tunnel, for Re = 0.75 (a) and 1 million (b) for S809 airfoil.

Figure 4.7: Angle of attack estimate from LEFS (red) compared with the actual angle of attack from CFD (blue). The value of azgrg deviates from the linear trend near stall which results in the same deduced azrrs (green marker) being predicted at two different aerodynamic states.

Figure 4.8: Variation of A with angle of attack for NACA 0012 and S809.

Figure 4.9: Angle of attack estimate from LEFS (red) compared with the actual angle of attack from CFD (blue). The value of azgrg deviates from the linear trend near stall which results in the same arpgrs (green marker) being predicted at two different aerodynamic states. The flow separation status at 50% chord can be tracked by sensors mounted at that location.

Figure 4.10: Angle of attack estimate from LEFS (red) and LEFS modified for post stall (green) compared with the actual angle of attack from CFD (blue). Angle of attack estimate from LEFS modified for post stall using the actual separation-point location from CFD is also co-plotted (black) for comparison.

Figure 4.12: Vorticity contours from CFD simulations to illustrate the critical LEV events. algorithm in identifying the critical LEV events, as depicted in Fig. 4.12

‘able 4.3: Unsteady flow simulation test cases for NACA 0012 airfoil at Re = 3 million. Case 1

Figure 4.13: Onset of LI from CFD and LI HE'S for EV formation for case 1 (NACA 0012). Comparison of LESP parameter the entire motion is shown in subplot (a). The quantity A from LEFS is shown in subplot (b). The marker represents the time instant when the LEV formation initiates, for which the vorticity and velocity contours are shown in (c) and (d).

Figure 4.14: CFD and LI LEV detachment for case 1 (NACA 0012). Comparison of LESP parameter from EF'S for the entire motion is shown in subplot (a). The quantity A from LEFS is shown in subplot (b). The marker represents the time instant when the LEV detachment occurs, for which the vorticity and velocity contours are shown in (c) and (d).

Figure 4.15: Termination of LEV shedding for case 1 (NACA 0012). Comparison of LESP parameter from CFD and LEFS for the entire motion is shown in subplot (a). The quantity A from LEFS is shown in subplot (b). The marker represents the time instant when the LEV shedding terminates, for which the vorticity and velocity contours are shown in (c) and (d).

Figure 4.16: Trailing-edge stall for case 2 (NACA 0012). Comparison of LESP parameter from CFD and LEFS for the entire motion is shown in subplot (a). The quantity A from LEFS is shown in subplot (b). The marker represents the time instant showing trailing-edge separation, for which the vorticity and velocity contours are shown in (c) and (d).

Figure 4.17: Onset of LI from CFD and LI HE'S for EV formation for case 3 (NACA 0012). Comparison of LESP parameter the entire motion is shown in subplot (a). The quantity A from LEFS is shown in subplot (b). The marker represents the time instant when the LEV formation initiates, for which the vorticity and velocity contours are shown in (c) and (d).

Figure 4.18: CFD and LI LEV detachment for case 3 (NACA 0012). Comparison of LESP parameter from EFS for the entire motion is shown in subplot (a). The quantity A from LEFS is shown in subplot (b). The marker represents the time instant when the LEV detachment occurs, for which the vorticity and velocity contours are shown in (c) and (d).

Figure 4.19: The research turbine with instrumented blade mounted in the wind-tunnel test section. uantities for numerous test sequences. Figure 4.19 shows the two bladed turbine in the wind-

Figure 4.20: Pressure taps and probe locations on the wind turbine blade used in the NRI UAI EF) experiemental campaign [92]. kl

Table 4.4: Mean absolute error at different span locations for two freestream velocity cases. The performance of the LEFS algorithm for a rotating blade demonstrates the algorithm’s ability to sense the local flow velocity at different blade sections using just five pressure meaz-

Figure 5.1: Microfence structures ((a) top view, (b) isometric view, (c) isometric view shown with higher magnification, (d) single microfence structure and four fiducial marks).

Figure 5.2: Schematic (A) and photograph (B) of the shear stress-sensor-prototype microscope for optical tracking.

Figure 5.3: Flat plate model in the wind tunnel ((a) drawing in which the fairing is not shown for clarity, (b) view of the model in test section, (c) view of the fairing under the test section housing the camera and microscope assembly).

Figure 5.4: Raw image (left) from the video recorded in the experiments compared to the enhanced version (right) with better contrast after intensity adjustment.

Table 5.1: Microfence deflections observed for different freestream velocity cases.

in micrometers, um.

Figure 5.7: Variation of microfence tip deflections with freestream velocity.

Figure 5.8: The 3D printed wedge with 30 degrees angle used to create flow reversal (left). A modified flat-plate setup with the wedge mechanism installed on the top surface, shown here in the unactuated position (right). orresponding micropillar response at critical flow conditions is presented in Figure 5.9 for a

Figure 5.10: The deflection (in pixels) of the microfence sensor with time as the velocity is increased and the wedge is translated towards the sensor location and then moved away.

Table A.1: Parameters used for generating the parabolic profile for different airfoils

Figure A.1: Comparison of freestream velocity, angle of attack, lift coefficient, and suctio1 coefficient estimated from LEFS (red) and the actual values from CFD (blue) for NACA 2301: airfoil at Re = 3 million. The dashed vertical line represents the stall angle of attack.

Figure A.2: Estimated velocity from LEFS algorithm (markers) compared with the actual velocity (dotted lines) for LS0417 airfoil at different Reynolds numbers.

Figure A.5: Estimated velocity from LEFS algorithm (markers) compared with the actual velocity (dotted lines) for S809 airfoil at different Reynolds numbers.

Figure A.6: Comparison of the deduced a and C; from LEFS and wind tunnel, for Re = 1.25 (a) and 1.5 million (b) for S809 airfoil. The dashed vertical line represents the stall angle of attack.