airframe noise research
TRANSCRIPT
![Page 1: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/1.jpg)
NASA Langley Research Center, Hampton, VA1
Airframe Noise Research
David P. Lockard
Acoustics TWG, NASA LaRC, April 21-22, 2015
![Page 2: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/2.jpg)
NASA Langley Research Center, Hampton, VA2
FUN3D Solutions for a Nose Landing Gearusing Wall Functions
Veer N. Vatsa, Jan-Renee Carlson,
David P. Lockard
And
Mehdi R. Khorrami
![Page 3: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/3.jpg)
3
Assessment of Wall-functions for Aero-acoustic Applications
• Explore the use of the newly developed wall-function capability in the FUN3D code for simulating unsteady flow
• The wall-function approach based on the approach outlined by Knopp et al., was incorporated recently in the unstructured grid flow solver, FUN3D
Knopp, Arrutz and Schwarmborn: Journal of Comp. Physics, vol. 220, pp.19-40, 2006
Carloson, Vatsa and White: AIAA Paper-2015-xxxx, AIAA Aviation 2015 Conference, June 2015
• Currently tested for one-equation model of Spalart-Allmarasand two-equation model of Menter
• Demonstrated to work well on grids where y+ at wall varies from O(1) to O(100)
![Page 4: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/4.jpg)
4
Initial Validation of FUN3D Wall-function Capability for RAE 2822 Airfoil
M=0.73, Alpha=2.8o
![Page 5: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/5.jpg)
5
Simplified Nose Landing Gear: Computational Setup and Surface Pressure
• Grids used in this study were generated for the partially dressed, cavity closed configuration using Pointwise
62 million node grid (PW-62M (y+=1)), near wall spacing is O(1)
45 million node grid (PW-45M (y+=25), near wall spacing is O(25)
Surface grids are identical
![Page 6: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/6.jpg)
6
Sample Grid Cuts at Torque-arm andWheel mid-plane (62 million node grid)
Zoomed view: focus region
![Page 7: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/7.jpg)
7
Comparisons of Time-averaged Velocity and Turbulence Kinetic Energy at Torque-arm cut
45 million node grid: y+=25 62 million node grid: y+=1
![Page 8: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/8.jpg)
8
Effect of using Wall-functions on Power Spectral Density Distributions
![Page 9: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/9.jpg)
9
Effect of using Wall-functions on Far-field Sound Pressure Levels
Mic #4 Mic #7
Mic #9Upstream
Downstream
Overhead
![Page 10: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/10.jpg)
10
Summary
• Demonstrated the wall-function capability of the unstructured grid flow code FUN3D for simulating the unsteady flow over a nose landing gear configuration
• Solution accuracy on wall-function grids comparable to standard integrate-to-wall grids for:
time-averaged and unsteady solutions
surface pressure power spectral density (PSD)
sound pressure levels (SPL) in far-field
• Minimal overhead associated with wall-function approach on a per node basis
Requires less computational resources to obtain solutions with comparable accuracy
![Page 11: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/11.jpg)
NASA Langley Research Center, Hampton, VA11
Comparison of Computational and Experimental Results for the 18%-Scale, Semi-Span Model
in the LaRC 14x22 Tunnel
Mehdi R. Khorrami and David P. Lockard
ERA ITD50A
![Page 12: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/12.jpg)
12
Aeroacoustic Testing of an 18%-Scale Aircraft Model
Multiple entries in NASA LaRC 14’x22’ tunnel
Phase1: October 2010 Aerodynamic measurements
Phase2: February 2013 Aero and acoustic measurements
Phase2 test campaign
Aerodynamic measurements
Acoustic: Microphone array and free-field microphones
Off-surface flow: PIV and LV
PIV measurements documented gear-flap interaction effects
Processing of PIV database completed
Measured off-surface flowfield being used to benchmark
high-fidelity simulations
PIV Configuration - 2013
“On the Accuracy of Simulated Gear-Flap Flow
Interaction”
By Mehdi R. Khorrami, Ramond Mineck, Ehab
Fares;,Chung-Sheng Yao, Luther N. Jenkins
2D-PIV setup capturing gear wake
and gear-flap interactionStereo PIV setup capturing gear
effects on flap tip vortex
![Page 13: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/13.jpg)
13
Gear-Flap Interaction Zone(Simulated vs. Measured Flowfield: Velocity Contours)
Mean velocity components Fluctuating velocity components
![Page 14: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/14.jpg)
14
Gear-Flap Interaction Zone(Simulated vs. Measured Flowfield: Velocity Profiles)
Mean and fluctuating velocity components
![Page 15: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/15.jpg)
15
Gear-Flap Interaction Zone(Simulated vs. Measured Flowfield: Velocity Contours)
Mean velocity components Fluctuating velocity components
![Page 16: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/16.jpg)
16
Gear-Flap Interaction Zone(Simulated vs. Measured Flowfield: Velocity Profiles)
1
6
Mean and fluctuating velocity components
Profiles along Y coordinate Profiles along Y coordinate Profiles along Z coordinate
![Page 17: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/17.jpg)
17
Gear Effects on Inboard Flap Tip Vortex(Simulated vs. Measured Flowfield: Velocity Contours)
Mean velocity components Fluctuating velocity components
![Page 18: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/18.jpg)
18
Gear Effects on Inboard Flap Tip Vortex(Simulated vs. Measured Flowfield: Velocity Profiles)
Mean and fluctuating velocity components
Profiles along Z coordinate
![Page 19: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/19.jpg)
19
Acoustic Comparisons
• Computations and experiments are in good agreement in the near field
Mean and fluctuating surface pressure
Flow velocities at the locations of PIV data
• Acoustic predictions from CFD use solid surface pressure data + Acoustic Analogy (Ffowcs Williams-Hawkings Equation)
Predicted signal at array center compared with experimental microphone array output
– Good agreement obtained with data from PowerFLOW® and FUN3D CFD codes
New comparisons involve using CFD to predict the signals at all microphone locations and applying the same array processing to those signals
– Signals from CFD do not suffer from extraneous noise but are extremely short in time duration compared with experiment
Experimental and CFD array data has been processed in a similar fashion using conventional beamforming and two deconvolutiontechniques: DAMAS and CLEAN-SC
![Page 20: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/20.jpg)
20
Spectral Comparison
M=0.2, AoA = 3 deg
Gear on
Flap at 39 deg
Array at 90 deg to model
Unified DAMAS V10R2
Tone from a cavity in LG that was covered
in the experiment and FUN3D grid
![Page 21: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/21.jpg)
21
Conventional Beamforming Source Strengths
M=0.2, AoA = 3 deg
Gear on
Flap at 39 deg
Array at 90 deg to model
5 kHz model scale
0.5 kHz full scale
Unified DAMAS V10R2
Experiment PowerFLOW
FUN3D
![Page 22: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/22.jpg)
22
DAMAS Beamforming Source Strengths
M=0.2, AoA = 3 deg
Gear on
Flap at 39 deg
Array at 90 deg to model
2.8 kHz model scale
0.5 kHz full scale
Unified DAMAS V10R2
Experiment PowerFLOW
FUN3D
![Page 23: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/23.jpg)
23
CLEAN-SC Beamforming Source Strengths
M=0.2, AoA = 3 deg
Gear on
Flap at 39 deg
Array at 90 deg to model
5 kHz model scale
0.9 kHz full scale
AVEC Array Processing Software
Version 3.14
Experiment PowerFLOW
FUN3D
![Page 24: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/24.jpg)
24
Summary
• Good comparisons between CFD and PIV in the wake of the landing gear and inboard flap despite differences in configuration
Open-jet experiment and free-field with floor in computations
Flow angularity in experiment
• Synthetic array results look reasonable despite short time records
• Array processing does not improve spectral comparisons between computations and experiment for this semi-span model test
• Beamforming contours of source strength provide an additional means of comparison
![Page 25: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/25.jpg)
25
Kyle A. Pascioni
Prof. Louis Cattefesta
Aeroacoustic Measurements of Slat Noise:
FSU Aeroacoustic Facility
Florida State University/NASA Collaboration
![Page 26: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/26.jpg)
26
Previous Measurements: Open-Jet Setup
80-channel
phased microphone array
![Page 27: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/27.jpg)
27
Recent Measurements: Closed-Wall Setup
![Page 28: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/28.jpg)
28
Mean Surface Pressure Measurements:Comparison with Open Air CFD Simulations
• Adequate simulation of aerodynamic environment at approach-like
angle of attack
• Closed-wall configuration required to obtain reasonable pressure
distribution
CFD (Open Air 5.5 deg)Wind Tunnel (8 deg)
![Page 29: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/29.jpg)
29
Kevlar Wall Setup
![Page 30: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/30.jpg)
30
Characterizing the Kevlar Wall
![Page 31: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/31.jpg)
31
Kevlar Wall Acoustic Attenuation: No Flow
• Δ𝑆𝑃𝐿 is difference between speaker to microphone transmission with and without Kevlar walls present
Attenuation due to transmission through either one (top) or two (bottom) walls
![Page 32: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/32.jpg)
32
Kevlar Wall Acoustic Attenuation: With Flow
• Δ𝑆𝑃𝐿 is the difference between the no flow condition and various tunnel speeds, both with two Kevlar walls.
• Greater reduction with tunnel speed due to scattering effects of higher amplitude turbulent fluctuations
![Page 33: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/33.jpg)
33
Plasma Acoustic Point Source for Kevlar Characterization
• Use a laser-generated acoustic point source to measure the array’s point spread function (PSF)
• With the measured PSF, deconvolution more accurate for advanced beamforming techniques such as DAMAS
Conduct w/ and w/out Kevlar wall: difference defining acoustic transmission loss
Conduct w/ and w/out flow, two Kevlar walls present: difference defining boundary layer losses
![Page 34: Airframe Noise Research](https://reader030.vdocuments.site/reader030/viewer/2022032314/6230a86f56b8fa43295b0d1c/html5/thumbnails/34.jpg)
34
Summary
• Kevlar wall configuration developed for airframe noise testing in the FSU facility
• Initial characterization of the transmission losses across the Kevlar wall established
Lift also changes requiring a modification to the angle to attack
• More advanced plasma source to be used to measure the in-situ array response (point spread function)
• Testing of 30p30n slat noise to follow