Investigation of Airfoil Aero Acoustics due to Different Stall - PowerPoint PPT Presentation

Investigation of Airfoil Aero Acoustics due to Different Stall Mechanism using Large Eddy Simulation (LES) Voo Keng Soon Vince Lim Nee Sheng Winson Tan Chun Hern DSO National Laboratories 05 Nov 2012 Slide 1

Brief Review of Available Experimental Data Slide 2

Literature Containing Experimental Data Brooks, T., Pope, D. and Marcolini, M., “Airfoil Self-Noise and Prediction”, NASA Reference Publication 1218, 1989 • Identification of 5 airfoil self-noise mechanisms • Contains a series of acoustic tests of 2D and 3D NACA0012 airfoils  6 2D blade sections of chord length from 2.54cm to 30.48cm (1in to 12in)  5 3D blade sections of chord length from 5.08cm to 30.48cm (2in to 12in)  wind tunnel speeds up to Mach 0.21 (Re based on chord up to 1.3x10 6 ) Slide 3

NACA0012 Blade Sections [1] 2D models 3D models span: 30.48cm span: 45.72cm chord: varies chord: varies trailing edge: sharp, < 0.05mm thick trailing edge: sharp, < 0.05mm thick [1] Brooks, T., Pope, D. and Marcolini, M., “Airfoil Self-Noise and Prediction”, NASA Reference Publication 1218, 1989 Slide 4

Acoustic Test Setup [1] M8 M5 30° M2 M1 1.22m M7 M4 Free jet nozzle [1] Brooks, T., Pope, D. and Marcolini, M., “Airfoil Self-Noise and Prediction”, NASA Reference Publication 1218, 1989 Slide 5

Brief Review of Different Airfoil Self-Noise Mechanisms Slide 6

Airfoil Self-Noise Mechanisms [1][2] [1] Brooks, T., Pope, D. and Marcolini, M., “Airfoil Self-Noise and Prediction”, NASA Reference Publication 1218, 1989 [2] Brooks,T. And Schlinker, R., “Progress in Rotor Broadband Noise Research”, Vertica, vol. 7, no. 4, 1983, pp. 287-307 Slide 7

Experimental Data [1] Self-noise Spectra for NACA0012 5.08cm chord, aoa 08.4° 5.08cm chord, aoa 15.4° 90 85 Sound Pressure Level (dB) 80 75 70 65 60 55 50 500 10000 Frequency (Hz) [1] Brooks, T., Pope, D. and Marcolini, M., “Airfoil Self-Noise and Prediction”, NASA Reference Publication 1218, 1989 Slide 8

CFD Simulation of NACA0012 at AOA of 8.4° to show Turbulent Boundary Layer Trailing Edge Noise and Boundary Layer Separation Noise Slide 9

Meshing and Testing Conditions • Trimmer mesh of ~11 million cells • cells aligned to air flow • RANS, LES • Mach 0.208 (71.3m/s) • 0.0508m chord, 0.1286m span • Reynolds Number studied: 230,000 • Periodic boundary conditions • Time-Step: 1.0e-5s • Simulated time: 0.1s • Aeroacoustics module: Ffowcs Williams - Hawkings Slide 10

Mid-Span Velocity Distribution (LES) and Streamwise Reynolds Stress <U'U'> normalized by ρ V 2 <U'U'> normalized by ρ V 2 <U'U'> normalized by ρ V 2 1 1 1 0.5 chord 2.0 chord 3.0 chord after TE after TE after TE 0.5 0.5 0.5 Z / chord Z / chord Z / chord 0 0 0 0.000 0.005 0.010 0.000 0.005 0.010 0.000 0.005 0.0 -0.5 -0.5 -0.5 -1 -1 -1 generates turbulent boundary layer trailing edge noise and boundary layer separation noise Slide 11

Acoustic Pressure Local pressure deviation from the ambient atmospheric pressure caused by sound wave Plot shows characteristics of a dipole Slide 12

Sound Pressure Level vs. 3 rd Octave Bands Sound Pressure Level vs. 3rd Octave Bands Experimental Data Observer Point (Corrected for Span) 70 Sound Pressure Level (dB) 60 50 40 500 10000 3rd Octave Frequency Bands (Hz) Simulation Peak SPL: 65.65 dB at Center Frequency: 3171 Hz Experiment Peak SPL: 65.61 dB at Center Frequency: 3171 Hz Reasonably close fit between experimental data and simulation Slide 13

Far-Field Sound Directivity Pattern SPL vs. Receiver Positions 90 105 75 60 120 60 50 135 45 40 150 30 30 20 165 15 10 180 0 0 195 345 210 330 225 315 240 300 255 285 270 Microphones are placed at a distance of 70x chord around airfoil Slide 14

Comparison of Coefficients At 8.4°AOA Drag Coefficient, C D Lift Coefficient, C L Published Data [3] for NACA0012 0.01748 0.81798 at Re 0.36x10 6 RANS 0.02481 0.83163 LES 0.02047 0.93481 Reference values for simulations: • reference area = 6.5068E-3 m² • reference velocity = 71.3 m/s • reference density = 1.17683 kg/m³ LES reported slightly higher C L . Mesh is coarse for LES. To consider modelling eddies in turbulent boundary layer (TBL) using RANS and the outer-flow with LES => Detached Eddy Simulation (DES) [3] Robert E. Sheldahl, Paul C. Klimes, “Aerodynamic Characteristics of Seven Symmetrical Airfoil Sections Through 180-Degree Angle of Attack for Use in Aerodynamic Analysis of Vertical Axis Wind Turbines”, Sandia National Laboratories, SAND80-2114, 1981 Slide 15

CFD Simulation of NACA0012 at AOA of 15.4° to show Large Scale Separation (Deep Stall) Noise Slide 16

Meshing and Testing Conditions • Trimmer mesh of ~11 million cells • cells aligned to air flow • URANS, LES • Mach 0.208 (71.3m/s) • 0.0508m chord, 0.1286m span • Reynolds Number studied: 230,000 • Periodic boundary conditions • Time-Step: 2.0e-5s • Simulated time: 0.2s • Aeroacoustics module: Ffowcs Williams - Hawkings Slide 17

Mid-Span Velocity Distribution (URANS) and Streamwise Reynolds Stress <U'U'> normalized by ρ V 2 <U'U'> normalized by ρ V 2 <U'U'> normaliz 1 1 1 0.5 chord 2.0 chord 3.0 chord after TE after TE after TE 0.5 0.5 0.5 Z / chord Z / chord Z / chord 0 0 0 0.000 0.050 0.100 0.000 0.050 0.100 0.000 0.050 -0.5 -0.5 -0.5 -1 -1 -1 large scale separation stall generating noise Slide 18

Mid-Span Velocity Distribution (LES) and Streamwise Reynolds Stress <U'U'> normalized by ρ V 2 <U'U'> normalized by ρ V 2 <U'U'> normaliz 1 1 1 0.5 chord 2.0 chord 3.0 chord after TE after TE after TE 0.5 0.5 0.5 Z / chord Z / chord Z / chord 0 0 0 0.000 0.050 0.100 0.000 0.050 0.100 0.000 0.050 -0.5 -0.5 -0.5 -1 -1 -1 large scale separation stall generating noise Slide 19

Acoustic Pressure Local pressure deviation from the ambient atmospheric pressure caused by sound wave Plot shows characteristics of a dipole Slide 20

Sound Pressure Level vs. 3 rd Octave Bands Sound Pressure Level vs. 3rd Octave Bands Experimental Data Observer Point (Corrected for Span) 90 Sound Pressure Level (dB) 80 70 60 500 10000 3rd Octave Frequency Bands (Hz) Simulation Peak SPL: 84.6 dB at Center Frequency: 1010 Hz Experiment Peak SPL: 88.9 dB at Center Frequency: 1010 Hz Similar trend between experimental data and simulation Slide 21

Far-Field Sound Directivity Pattern SPL vs. Receiver Positions 90 105 75 70 120 60 60 135 45 50 40 150 30 30 20 165 15 10 180 0 0 195 345 210 330 225 315 240 300 255 285 270 Microphones are placed at a distance of 70x chord around airfoil Slide 22

Comparison of Coefficients At 15.4°AOA Drag Coefficient, C D Lift Coefficient, C L Published Data [3] for NACA0012 0.21114 0.64258 at Re 0.36x10 6 URANS 0.31031 0.92369 LES 0.17406 1.15898 Reference values for simulations: • reference area = 6.5068E-3 m² • reference velocity = 71.3 m/s • reference density = 1.17683 kg/m³ LES reported higher C L . Mesh is coarse for LES. To consider DES. [3] Robert E. Sheldahl, Paul C. Klimes, “Aerodynamic Characteristics of Seven Symmetrical Airfoil Sections Through 180-Degree Angle of Attack for Use in Aerodynamic Analysis of Vertical Axis Wind Turbines”, Sandia National Laboratories, SAND80-2114, 1981 Slide 23

Results and Conclusions • For the case involving 8.4° AOA,  a reasonably close fit between the experimental data and simulation  similar peak SPL and center frequency  LES reported a slightly higher C L . • For the case involving 15.4° AOA,  comparable trend between the experimental data and simulation albeit loose fit  a difference of ~5dB in the peak SPL, similar center frequency  LES reported a much higher C L . • Mesh for a fully resolved LES has to be sufficiently fine to resolve the small eddies • Considerations for future work:  Refining the existing mesh  Switching to Detached Eddy Simulation for additional comparison Slide 24