The Pennsylvania State University
The Graduate School
Department or Aerospace Engineering
NON-HARMONIC ROOT-PITCH INDIVIDUAL-BLADE CONTROL FOR THE
REDUCTION OF BLADE-VORTEX INTERACTION NOISE IN ROTORCRAFT
A Dissertation in
Aerospace Engineering
by
Brendon D. Malovrh
2012 Brendon D. Malovrh
Submitted in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
December 2012
The dissertation of Brendon D. Malovrh was reviewed and approved* by the following:
Farhan S. Gandhi Professor of Aerospace Engineering Dissertation Advisor Chair of Committee
Kenneth S. Brentner Professor of Aerospace Engineering
Edward C. Smith Professor of Aerospace Engineering
Christopher Rahn Professor of Mechanical Engineering
Kevin W. Noonan Aerospace Engineer NASA Langley Research Center
George A. Lesieutre Professor of Aerospace Engineering Head of the Department of Aerospace Engineering
*Signatures are on file in the Graduate School
iii ABSTRACT
One of the greatest obstacles to public acceptance of rotorcraft is the high levels
of noise they produce, particularly in low-speed descent. In this flight condition, the
trailing edge vortex of one blade often passes in close proximity to other blades resulting
in impulsive changes in lift. This Blade-Vortex Interaction (BVI) creates high levels of
both noise and vibration.
The objective of this dissertation is to evaluate the effectiveness of using
physically motivated pulse-type Individual Blade Control for reducing the noise
associated with the BVI. First, the major parameters that affect the severity of the
interaction, such as vortex strength and blade-vortex miss-distance, are analyzed. Second, inputs designed specifically to alter the parameters previously identified as key are explored, resulting in elimination of advancing side noise and overall peak BVI Sound
Pressure Level (BVISPL) reductions of up to 4.6 dB. Lastly, different feedback mechanisms for closed-loop control of IBC are examined to allow implementation of the developed inputs. iv TABLE OF CONTENTS
LIST OF FIGURES ...... vi
ACKNOWLEDGEMENTS ...... xvi
Chapter 1 Introduction ...... 1
1.1 BVI and Important Interaction Parameters ...... 2 1.2 Operational Methods ...... 4 1.3 Passive Methods ...... 7 1.3.1 Advanced Tip Configurations ...... 8 1.3.2 Other Passive Methods ...... 11 1.4 Active Methods ...... 13 1.4.1 Higher Harmonic Control ...... 13 1.4.2 Individual Blade Control ...... 16 1.5 Overview ...... 20 References ...... 21
Chapter 2 Sensitivity of BVI-Induced Noise and Vibration to Variations in Individual Interaction Parameters ...... 29
2.1 Analysis Method ...... 30 2.2 Baseline ...... 33 2.3 Vortex Strength ...... 37 2.4 Core Radius ...... 42 2.5 Miss-Distance ...... 44 2.6 Blade-Shaft Plane Interaction Angle ...... 50 2.7 Rotor Disk Plane Interaction Angle ...... 55 2.8 Spanwise Location of the Interaction ...... 60 2.9 Spanwise Length of the Interaction ...... 64 2.10 Collective Pitch ...... 68 2.11 Comparison of Effects of Various Parameters in Reducing BVI-induced Noise and Hub Vibratory Loading ...... 74 2.11.1 Weak Interaction Parameters ...... 75 2.11.2 Strong Interaction Parameters ...... 76 2.12 Minimum Vibration Case in the HART Test ...... 80 2.13 Conclusions ...... 82 References ...... 86
Chapter 3 Localized Individual Blade Root Pitch Control for BVI Noise Reduction ...... 88
3.1 Description of Analysis and Results for the Baseline Configuration ...... 89 3.2 Localized IBC Inputs to Reduce Interacting Vortex Strength ...... 105 v 3.3 IBC Input Profile ...... 108 3.4 Second Quadrant Vortex Strength Reduction ...... 117 3.5 Third Quadrant Vortex Strength Reduction ...... 128 3.6 Combination of Second and Third Quadrant Vortex Strength Reduction ...... 135 3.7 Effects of Retrimming ...... 142 3.8 Conclusions ...... 145 References ...... 149
Chapter 4 Metrics for BVI noise ...... 151
4.1 Skid Microphones as a Feedback Metric ...... 152 4.1.1 Variable Advance Ratio, 3 deg Backward Shaft Tilt ...... 154 4.1.2 Variable Advance Ratio, 4 deg Backward Shaft Tilt ...... 163 4.1.3 Variable Shaft Tilt, 0.135 Advance Ratio ...... 172 4.1.4 Constant Flight Condition (4 deg backward Shaft Tilt, 0.135 Advance Ratio), Variable IBC Input ...... 179 4.1.5 Constant Flight Condition (3 deg backward Shaft Tilt, 0.17 Advance Ratio), Variable IBC Input ...... 183 4.1.6 Summary of Skid Microphone Feedback ...... 188 4.2 Blade Pressures as a Feedback Metric ...... 189 4.2.1 First Quadrant RMS ...... 189 4.2.2 BPAP ...... 191 4.2.2.1 BPAP Model Development ...... 192 4.2.2.2 BPAP Results ...... 199 4.2.3 Summary Blade Pressure Feedback ...... 203 References ...... 204
Chapter 5 Conclusions and Reccomendations for Future Work ...... 206
5.1 Sensitivity of BVI-Induced Noise and Vibration to Varisations in Individual Interaction Parameters ...... 206 5.2 Localized Individual Blade Root Pitch Control for BVI Noise Reduction .... 209 5.3 Metrics for BVI noise ...... 212 5.4 Recommendations for Future Work ...... 213
vi LIST OF FIGURES
Figure 1.1: High-speed flight, no BVI...... 3
Figure 1.2: Low-speed descent, BVI condition...... 3
Figure 1.3: Lift at 80% radius...... 4
Figure 1.4: BVISPL as a function of shaft tilt and advance ratio [25]...... 6
Figure 1.5: Advanced tip shapes a) ogee tip b) winglet c) tapered tip...... 9
Figure 1.6: BERP blade tip...... 9
Figure 1.7: Anhedral...... 11
Figure 1.8: Configuration for HHC...... 16
Figure 1.9: Configuration for trailing edge flaps...... 17
Figure 1.10: Configuration for IBC...... 19
Figure 2.1: Baseline blade-vortex interaction geometry...... 31
Figure 2.2: Baseline non-dimensional sectional lift distribution over the rotor disk... 34
Figure 2.3: Baseline blade root loads versus azimuth (lift integrated along span, non-dimensionalized by ρacR(ΩR)2)...... 35
Figure 2.4: Frequency spectrum of baseline BVI-induced hub vertical loads...... 36
Figure 2.5: Baseline BVISPL footprint one diameter below the rotor disk plane...... 37
2 Figure 2.6: CNM versus azimuth at 75% radius spanwise station for different vortex strengths...... 39
2 Figure 2.7: d(CNM )/dψ versus azimuth at 75% radius spanwise station for different vortex strengths...... 39
Figure 2.8: Max BVISPL and sensitivity versus vortex strength...... 40
Figure 2.9: Frequency spectrum of BVI-induced hub vertical loads for various vortex strengths...... 41
Figure 2.10: Max BVISPL versus vortex core radius...... 43 vii Figure 2.11: Frequency spectrum of BVI-induced hub vertical loads for various core radii...... 44
Figure 2.12: Max BVISPL and sensitivity versus miss-distance...... 46
Figure 2.13: Max BVISPL versus miss-distance (small values of miss-distance)...... 47
2 Figure 2.14: CNM versus azimuth at 75% radius spanwise station for different blade-vortex miss-distances...... 48
2 Figure 2.15: d(CNM )/dψ versus azimuth at 75% radius spanwise station for different blade-vortex miss-distances...... 49
Figure 2.16: Frequency spectrum of BVI-induced hub vertical loads for various blade-vortex miss-distances...... 50
Figure 2.17: Vortex interacting with the blade at an angle in the blade-shaft plane. .. 52
Figure 2.18: Max BVISPL versus blade-shaft plane interaction angle...... 53
Figure 2.19: Frequency spectrum of BVI-induced hub vertical loads for various blade-shaft plane interaction angles...... 54
Figure 2.20: Vortex interacting with blade at an angle in the rotor disk plane (oblique interaction)...... 56
2 Figure 2.21: CNM versus azimuth for different rotor disk plane interaction angles (parallel and oblique interaction)...... 57
2 Figure 2.22: d(CNM )/dψ versus azimuth for different rotor disk plane interaction angles (parallel and oblique interaction)...... 57
Figure 2.23: Max BVISPL versus rotor disk plane interaction angle...... 58
Figure 2.24: Frequency spectrum of BVI-induced hub vertical loads for various rotor disk plane interaction angles...... 60
Figure 2.25: Max BVISPL and sensitivity versus spanwise location of center of interacting vortex...... 62
Figure 2.26: Blade root loads versus azimuth for various spanwise locations of center of interacting vortex...... 63
Figure 2.27: Frequency spectrum of BVI-induced hub vertical loads for various spanwise locations of center of interacting vortex...... 64 viii Figure 2.28: Max BVISPL versus spanwise location of center of interacting vortex and length of interaction...... 66
Figure 2.29: Non-dimensional 4/rev BVI-induced vertical hub loads versus spanwise location of center of interacting vortex and length of interaction...... 67
Figure 2.30: Non-dimensional 8/rev BVI-induced vertical hub loads versus spanwise location of center of interacting vortex and length of interaction...... 68
Figure 2.31: Blade sectional lift at 75% radius for various collective pitch settings, θ...... 70
Figure 2.32: Interacting vortex trajectories for different initial positions relative to blade, and different blade pitch settings...... 72
Figure 2.33: Max BVISPL versus rotor collective pitch for different initial interacting vortex positions...... 73
Figure 2.34: Max BVISPL versus interaction angles in the rotor disk and blade- shaft planes...... 78
Figure 2.35: Non-dimensional BVI-induced 4/rev hub vertical loads versus interaction angles in the rotor disk and blade-shaft planes...... 79
Figure 2.36: HART test [9] blade-vortex interaction geometry in blade-shaft plane for baseline...... 81
Figure 2.37: HART test [9] blade-vortex interaction geometry in blade-shaft plane for minimum vibration HHC input...... 82
Figure 3.1: Baseline (no IBC case), top-view of blade-vortex interactions and miss-distances for full rotor disk...... 92
Figure 3.2: Baseline (no IBC case), top-view of blade-vortex interactions and miss-distances for first quadrant (advancing side interactions)...... 93
Figure 3.3: Baseline (no IBC case), top-view of blade-vortex interactions and miss-distances for fourth quadrant (retreating side interactions)...... 94
Figure 3.4: Relative positions of reference blade to tip-vortex generation...... 95
Figure 3.5: Azimuthal variation of the strength of the generated vortices (maximum bound circulation)...... 97
Figure 3.6: Baseline (no IBC case) non-dimensional strength of interacting vortices for full rotor disk...... 98 ix Figure 3.7: Baseline (no IBC case) non-dimensional strength of interacting vortices for first quadrant (advancing side interactions)...... 99
Figure 3.8: Baseline (no IBC case) non-dimensional strength of interacting vortices for fourth quadrant (retreating side interactions)...... 99
Figure 3.9: Baseline (no IBC) BVISPL due to full wake (109.2 peak advancing side, 107.9 dB peak retreating side)...... 101
Figure 3.10: Baseline (no IBC) BVISPL due to (k−0) vortex (106.5 dB peak advancing side, 101.4 dB peak retreating side)...... 102
Figure 3.11: Baseline (no IBC) BVISPL due to (k−1) vortex...... 103
Figure 3.12: Baseline (no IBC) BVISPL due to (k−2) vortex...... 104
Figure 3.13: Baseline (no IBC) BVISPL due to (k−3) vortex (107.4 dB peak advancing side)...... 105
Figure 3.14: Tip vortex elements generated over portions of the second quadrant resulting in first quadrant BVI...... 108
Figure 3.15: IBC input profiles...... 109
Figure 3.16: (k−3) vortex BVISPL due to a truncated step IBC input profile ( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚)...... 111
Figure 3.17: (k−3) vortex BVISPL due to a half-period sine IBC input profile ( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚)...... 112
Figure 3.18: (k−3) vortex BVISPL due to a full-period cosine IBC input profile
( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚)...... 114
Figure 3.19: (k−3) vortex BVISPL due to a ramped IBC input profile ( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚)...... 116
Figure 3.20: BVI non-dimensional vortex strength for = 1˚ ramped IBC input centered at ψ = 140˚...... 118
Figure 3.21: BVISPL due to (k−0) vortex for = 1˚ ramped IBC input centered at ψ = 140˚ (103.7 dB peak advancing side, 105.7 dB peak retreating side)...... 119
Figure 3.22: Blade-vortex interactions and miss-distance for = 1˚ ramped IBC input centered at ψ = 140˚...... 120 x Figure 3.23: BVI non-dimensional vortex strength for = 2˚ ramped IBC input centered at ψ = 140˚...... 121
Figure 3.24: BVISPL due to (k−3) vortex for = 2˚ ramped IBC input centered at ψ = 140˚ (98.5 dB peak)...... 122
Figure 3.25: BVISPL due to (k−0) vortex for = 2˚ ramped IBC input centered at ψ = 140˚ (108.5 dB peak)...... 123
Figure 3.26: Blade-vortex interactions and miss-distance for = 2˚ ramped IBC input centered at ψ = 140˚...... 124
Figure 3.27: BVI non-dimensional vortex strength for = 3˚ ramped IBC input centered at ψ = 140˚...... 125
Figure 3.28: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ (106.8 dB peak)...... 126
Figure 3.29: BVISPL due to full wake for = 3˚ ramped IBC input centered at ψ = 140˚ (107.0 dB peak)...... 127
Figure 3.30: BVI non-dimensional vortex strength for = 1˚ ramped IBC input centered at ψ = 225˚...... 130
Figure 3.31: BVISPL due to full wake for = 1˚ ramped IBC input centered at ψ = 225˚ (108.7 dB peak advancing side, 106.8 dB peak retreating side)...... 131
Figure 3.32: BVI geometry and miss-distance for = 1˚ ramped IBC input centered at ψ = 225˚...... 132
Figure 3.33: BVI non-dimensional vortex strength for = 2˚ ramped IBC input centered at ψ = 225˚...... 133
Figure 3.34: BVISPL due to full wake for = 2˚ ramped IBC input centered at ψ = 225˚ (108.7 dB peak advancing side, 105.1 dB peak retreating side)...... 134
Figure 3.35: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ and = 1˚ ramped IBC input centered at ψ = 225˚ (106.0 dB peak)...... 137
Figure 3.36: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ and = 2˚ ramped IBC input centered at ψ = 225˚ (105.5 dB peak)...... 138 xi Figure 3.37: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ and = 3˚ ramped IBC input centered at ψ = 225˚ (102.6 dB peak)...... 139
Figure 3.38: BVI geometry and miss-distance for = 3˚ ramped IBC input centered at ψ = 140˚ and = 2˚ ramped IBC input centered at ψ = 225˚...... 140
Figure 3.39: BVI geometry and miss-distance for = 3˚ ramped IBC input centered at ψ = 140˚ and = 3˚ ramped IBC input centered at ψ = 225˚...... 141
Figure 3.40: BVISPL due to full wake for = 3˚ ramped IBC input centered at ψ = 140˚ and = 3˚ ramped IBC input centered at ψ = 225˚ (104.6 dB peak)...... 142
Figure 3.41: BVISPL due to full wake with rotor retrimmed for the = 3˚ ramped IBC input centered at ψ = 140˚ (103.8 dB peak advancing side, 106.0 dB peak retreating side)...... 144
Figure 3.42: Azimuthal variation of maximum (non-dimensional) bound circulation (representative of the strength of generated vortices), with the case of the retrimmed rotor included...... 145
Figure 4.1: Skid Microphone Locations ...... 153
Figure 4.2: BVISPL at observer microphones for 3 degree backward shaft tilt and variable advance ratio ...... 155
Figure 4.3: BVISPL at observer microphone 1 for 3 degree backward shaft tilt and variable advance ratio ...... 156
Figure 4.4: BVISPL at observer microphone 2 for 3 degree backward shaft tilt and variable advance ratio ...... 156
Figure 4.5: BVISPL at observer microphone 3 for 3 degree backward shaft tilt and variable advance ratio ...... 157
Figure 4.6: BVISPL at observer microphone 4 for 3 degree backward shaft tilt and variable advance ratio ...... 157
Figure 4.7: BVISPL at observer microphone 5 for 3 degree backward shaft tilt and variable advance ratio ...... 158
Figure 4.8: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.135 ...... 159 xii Figure 4.9: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.140 ...... 159
Figure 4.10: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.145 ...... 160
Figure 4.11: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.150 ...... 160
Figure 4.12: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.155 ...... 161
Figure 4.13: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.160 ...... 161
Figure 4.14: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.165 ...... 162
Figure 4.15: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.170 ...... 162
Figure 4.16: BVISPL at observer microphones for 4 degree backward shaft tilt and variable advance ratio ...... 164
Figure 4.17: BVISPL at observer microphone 1 for 4 degree backward shaft tilt and variable advance ratio ...... 164
Figure 4.18: BVISPL at observer microphone 2 for 4 degree backward shaft tilt and variable advance ratio ...... 165
Figure 4.19: BVISPL at observer microphone 3 for 4 degree backward shaft tilt and variable advance ratio ...... 165
Figure 4.20: BVISPL at observer microphone 4 for 4 degree backward shaft tilt and variable advance ratio ...... 166
Figure 4.21: BVISPL at observer microphone 5 for 4 degree backward shaft tilt and variable advance ratio ...... 166
Figure 4.22: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.130 ...... 167
Figure 4.23: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.135 ...... 168
Figure 4.24: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.140 ...... 168 xiii Figure 4.25: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.145 ...... 169
Figure 4.26: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.150 ...... 169
Figure 4.27: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.155 ...... 170
Figure 4.28: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.160 ...... 170
Figure 4.29: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.165 ...... 171
Figure 4.30: BVISPL at observer microphones for an advance ratio of 0.135 and variable shaft tilt...... 172
Figure 4.31: BVISPL at observer microphone 1 for an advance ratio of 0.135 and variable shaft tilt...... 173
Figure 4.32: BVISPL at observer microphone 2 for an advance ratio of 0.135 and variable shaft tilt...... 173
Figure 4.33: BVISPL at observer microphone 3 for an advance ratio of 0.135 and variable shaft tilt...... 174
Figure 4.34: BVISPL at observer microphone 4 for an advance ratio of 0.135 and variable shaft tilt...... 174
Figure 4.35: BVISPL at observer microphone 5 for an advance ratio of 0.135 and variable shaft tilt...... 175
Figure 4.36: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.135 ...... 176
Figure 4.37: BVISPL in a) Skid microphone plane and b) Observer plane for 3.5 degree backward shaft tilt and advance ratio of 0.135 ...... 176
Figure 4.38: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.135 ...... 177
Figure 4.39: BVISPL in a) Skid microphone plane and b) Observer plane for 4.5 degree backward shaft tilt and advance ratio of 0.135 ...... 177
Figure 4.40: BVISPL in a) Skid microphone plane and b) Observer plane for 5 degree backward shaft tilt and advance ratio of 0.135 ...... 178 xiv Figure 4.41: BVISPL in a) Skid microphone plane and b) Observer plane for 6 degree backward shaft tilt and advance ratio of 0.135 ...... 178
Figure 4.42: BVISPL at observer microphones for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input...... 180
Figure 4.43: BVISPL at observer microphone 1 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input...... 181
Figure 4.44: BVISPL at observer microphone 2 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input...... 181
Figure 4.45: BVISPL at observer microphone 3 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input...... 182
Figure 4.46: BVISPL at observer microphone 4 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input...... 182
Figure 4.47: BVISPL at observer microphone 5 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input...... 183
Figure 4.48: BVISPL at observer microphones for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input...... 184
Figure 4.49: BVISPL at observer microphone 1 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input...... 185
Figure 4.50: BVISPL at observer microphone 2 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input...... 185
Figure 4.51: BVISPL at observer microphone 3 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input...... 186
Figure 4.52: BVISPL at observer microphone 4 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input...... 186
Figure 4.53: BVISPL at observer microphone 5 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input...... 187
Figure 4.54: First quadrant RMS tracking with maximum advancing side noise...... 190
Figure 4.55: Relative arrival times at observer location for different radial sources...... 198
Figure 4.56: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 3 backward degree shaft tilt and variable advance ratios...... 200 xv Figure 4.57: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 0.135 advance ratios and variable shaft tilt...... 201
Figure 4.58: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 0.135 advance ratios, 4 degree backward shaft tilt, and variable IBC input...... 202
Figure 4.59: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 0.170 advance ratios, 3 degree backward shaft tilt and variable IBC input...... 203
xvi ACKNOWLEDGEMENTS
The completion of this dissertation would not have been possible without the constant support and encouragement of many people. I am sincerely grateful to my friends and family for helping me through this period of my life.
I would like to thank my advisor, Dr. Farhan Gandhi, for introducing me to the world of graduate level research and teaching me the skills I would need to achieve what
I have, even if our personalities overlap in ways that have extended the duration of this research somewhat more than I may have guessed when beginning it.
I would also like to thank my supervisors at NASA Langley Research Center who have picked up in my education as an engineer and a researcher where Dr. Gandhi left off. They have provided a fantastic work environment that is exactly what I hoped for when I began my education and their support in completing this dissertation is one of the primary reasons for its completion.
My coworkers, both in the Rotorcraft Center of Excellence at Penn State and at
NASA Langley have provided a great deal of emotional and intellectual support. They have always been there to bounce and idea off of or to read over what I have done and I owe them deeply for this.
I would also like to thank the faculty and staff at The Pennsylvania State
University. From the secretaries at the Department of Aerospace Engineering (plural as there have been several during the time I have been here) who were instrumental in keeping we on the correct side of forms and deadlines, to the many professors who have xvii taught me so much, to my dissertation committee who pulled me through the completion of this document, they have all been the heart and soul of my time at Penn State.
Lastly I would like to thank my girlfriend Cathy Carlson, who’s love, support, and incessant nagging helped me beyond words. She did not realize what she had signed up for and I thank her for sticking with me throughout this process.
Chapter 1
Introduction
Helicopters and tiltrotors are unique among Vertical Take Off and Landing
(VTOL) aircraft in their ability to fly at very low speeds and hover for extended periods
of time. This makes them singularly useful in both civilian and military applications.
Helicopters have found roles in the military in both defensive (search and rescue,
particularly over water, and intelligence gathering) and offensive (utilizing terrain for
cover, and insertion and retrieval of units) missions. Their use by police allows for
monitoring situations from afar, thus reducing danger to on-the-ground officers and
civilians, particularly in high speed chases. Helicopters are also utilized in the private sector, particularly by the media, oil companies operating offshore, and as transport in
metropolitan areas. A major drawback of their diverse flight envelope is the complex
mechanics and aerodynamics involved. One of the worst flight conditions for rotorcraft is
in low-speed descent. Here, the tip vortices generated by the blades can pass close to
following blades resulting in a highly impulsive blade loading [1]. These Blade-Vortex
Interactions (BVI) are known to produce high levels of both noise and vibration. Of
particular concern is that the noise generated by BVI events tends to fall into the
frequency spectrum found most annoying to humans [1]. In military applications, a
reduction in noise signature is critical in increasing the survivability of rotorcraft on the
battlefield and enabling helicopters to be used for covert insertion and recovery of troops
[2]. In civilian applications, lower interior noise is key to acceptance of helicopters as a
2 method of transportation while lower approach noise essential to increasing public acceptance of any rotorcraft (helicopters, tiltrotors, or autogyros) use, particularly in an urban environment where the greatest potential for heliports exists [3, 4]. BVI-induced vibrations not only decrease passenger comfort, but also significantly contribute to pilot fatigue. Their effect on the rotorcraft itself increases maintenance costs, reduces component life, and can also cause avionics problems and reduce the accuracy of weapons systems [1]. For these reasons, a considerable amount of effort has been directed towards prediction and alleviation of BVI in recent years.
1.1 BVI and Important Interaction Parameters
In moderate to high speed cruise, the rotor wake, dominated by the strong tip vortices released by the blades, is swept below and aft of the rotor disk, owing to the large forward velocity and downwash through the disk as seen in Figure 1.1. In low speed descent, however, the lower forward speed and significantly reduced downwash allow the released tip vortices to come into close proximity with the blades before passing behind the rotorcraft as shown in Figure 1.2. It is these close passages that cause the highly impulsive changes in blade lift seen in Figure 1.3. Several key parameters of this interaction which contribute to the resulting noise and vibration levels have been identified [5-8]. The relative strengths of these different parameters have been investigated [5] and will be explored in Chapter 2.
3
Rotor Disk V∞
Tip Vortex
Figure 1.1: High-speed flight, no BVI.
Rotor Disk
V∞ Tip Vortex Interaction Site
Figure 1.2: Low-speed descent, BVI condition.
4
800 Primary BVI Event
700
600 lift (N/m)
500
400
0 100 200 300 Azimuthpsi (deg)
Figure 1.3: Lift at 80% radius.
Reductions in BVI-induced vibration and noise can be brought about by methods that alter one, or a combination of several of the interaction parameters. These methods will be broken down into three general categories: operational, passive, and active.
1.2 Operational Methods
Operational methods for BVI noise and vibration reduction act not by changing the properties of the vortex directly, but rather by modifying the entire aerodynamic environment to avoid the condition of BVI. Since the 1970’s the presence of flight conditions that created strong BVI environments had been identified [9]. These may be
5 plotted on a forward speed vs. vertical descent rate or advance ratio vs. shaft tilt graph
(Figure 1.4) to identify regions of the flight envelope that are known to result in high noise levels. Thus, this allows for the simplest method of BVI noise reduction: avoid the problem. By identifying the high noise regions of the flight envelope for a given helicopter through experiment or theoretical calculation, a flight plan that avoids these regions may be conceived [10, 11, 12, 13, 14, 15]. This is a problem due to the inherent difficulty for pilots to hold to the steep descent rate required using traditional instruments
[12]. This concept has received renewed interest recently due to the development of
Differential Global Positioning System (DGPS) [13]. It has been shown that the use of
DGPS coupled to a flight director makes it possible to follow a specific noise abatement approach with a workload similar to a standard Instrument Landing System (ILS) approach [14-16]. Other methods, based on estimating the miss-distance between the blade and vortex and therefore suggesting modifications to the flight path have also been developed [17]. Unfortunately, implementation of these procedures has been hampered by ICAO and FAA regulations. A noise abatement approach would use a multi-segment glidepath to fly “around” the primary BVI regions of the flight envelope, but the current noise certification procedures of ICAO and the FAA, require the helicopter to fly a 6 constant glideslope approach at Vy (best rate of climb speed) [3]. A certification scheme
that allows for advanced noise abatement procedures has been proposed [18], and if
approved would allow the use of noise abatement procedures for BVI alleviation during
certification.
6
Figure 1.4: BVISPL as a function of shaft tilt and advance ratio [20].
Another approach having recently been studied, X-force control, changes the tip- path-plane angle during low speed descent without modifying the flight path per se to alleviate BVI [19]. Here, high noise regions are computed either experimentally or numerically as a function of the longitudinal force acting on the helicopter (the X-force).
Initial studies suggested using simple momentum theory to identify conditions of zero average inflow through the disk. This was thought to represent a state where the tip vortices would remain in the plane of the rotor disk, thus resulting in a high BVI condition. More recent studies have shown that the complex nature of the aerodynamic environment around the rotor does not allow for such a simple assumption, and a more detailed investigation of the wake is required. Once the most detrimental cases are identified, it can be determined whether positive or negative application of X-force is
7 required to tilt the rotor disk in such a way as to reduce the resulting BVI. A recent study has suggested that reductions of more than 10 dB are possible with this method [20].
Devices to control the X-force could consist of drag devices such as air-brake panels deployed during approach, or ducted propellers that could be used to vary the X-force.
Another approach for operational BVI alleviation consists in changing the rotor speed with the flight condition. A flight test on a 5-bladed McDonnell Douglas MD-500E helicopter showed a 10 dB noise reduction during a flyover for a 25% RPM reduction
[21]. However, the reduction of energy stored in the rotor creates a great danger if engine power is lost, and thus the authors emphasized the necessity to design alternate methods for autorotation at low-speed.
1.3 Passive Methods
Passive BVI reduction methods are those that consist of a fixed design of the rotor that reduces BVI noise and/or vibrations. Another possibility, adaptive-passive BVI reduction, will be included here as well, as it may be described as a passive technique that can be modified to suit different flight conditions. Most investigation into passive BVI reduction methods has been in the field of advanced tip configuration, with some attention being given to a variety of other concepts.
8 1.3.1 Advanced Tip Configurations
The primary passive BVI reduction method is the introduction of advanced blade tip shapes. These are able to weaken the tip vortex as it is being produced. One of these designs, the Ogee-Tip (Figure 1.5a), was tested on a full-scale UH-1H rotor and has shown noise reductions as high as 15 dB by diffusing the tip vortex [22]. Another tip design, the BERP (Figure 1.6) (British Experimental Rotor Program) [23], is currently flying on the Westland Lynx and the Agusta/Westland EH101. This design continues to be refined [24], and though it was not intended to reduce BVI noise, it has shown some positive effects. The University of Aachen has developed a winglet (Figure 1.5b) designed to spread the tip vortex, modify its trajectory and generate less BVI noise. It also creates a 7% increase in hover efficiency that comes with the drawback of an increased forward flight rotor drag [25, 26]. Another design that has been shown to reduce BVI noise, though it was originally designed to reduce high-speed noise, is the parabolic tip [27]. This design has is being used in production helicopters such as the NH
90, the Eurocopter Tiger, the Super Puma and the EC 135.
9
Figure 1.5: Advanced tip shapes a) ogee tip b) winglet c) tapered tip.
Figure 1.6: BERP blade tip.
10
Sweep (Figure 1.5c), such as on the UH-60A blade [28], is usually introduced to alleviate compressibility effects on the advancing side at high forward speeds [29], but it has also been extensively tested for its effect on BVI. Flow visualization has shown that sweep can create a weaker vortex inboard, which should help to diffuse the primary tip vortex [30]. It is thought that sweep will also change the angle of interaction in the disk plane, possibly reducing noise. Different numerical simulations have also concluded that sweep has the potential to reduce BVI noise and vibration [31, 32]. However, these predictions have not been borne out in experimental measurements of noise with either forward or rearward sweep [33]. A doubly swept tip has been suggested but the numerical study of its effect on BVI was inconclusive [36]. Anhedral (Figure 1.7) also attempts to create a secondary vortex as well as to displace the primary tip vortex further down [37]. It should also change the angle of interaction in the blade-shaft plane, possibly weakening the interaction. An acoustic experimental study failed to show any improvements over a baseline rectangular blade tip though [33].
A slotted tip was shown to have the ability to diffuse the tip vortex, greatly increasing its core size in a wind tunnel test [34]. Here, forward-facing slots along the blades leading edge divert some of the flow through the blade and vent it at the tip.
Continuing work [35] has supported these findings and noted the potential for BVI noise reduction, though no quantitative results have been published. Several other possible tip shapes exist including such designs such as spoilers, porous tips, and end plates [38].
11
Figure 1.7: Anhedral.
1.3.2 Other Passive Methods
A wide variety of modifications to traditional rotor configurations have been examined for their effect on reducing BVI noise and vibration. Sweep does not have to be constrained to the blade tip, but can be extended further inboard, as far as the 70% radial position to create "dogleg" or "paddle" planforms. A numerical study with an unloaded rotor concluded that large noise reductions could be achieved by these configurations
[39]. Taper has been shown to have an effect on BVI, by insuring a smoother radial decay of the lift distribution on the outboard portion of the blade. An advanced BO-105 rotor, with taper starting at the 80% radial position, was tested in the DNW (Duits-Nederlandse
Windtunnel) in a strong BVI condition and reduced overall noise by 3-4 dB [40]. This design was eventually chosen for the EC 135, although mostly for its high-speed flight
12 noise reductions. Along with advanced airfoils, this rotor has resulted in flyover noise reductions of up to 3 dB [27].
Other passive designs have involved the entire geometry of the rotor such as scissors rotors or the Variable Geometry Rotor (VGR). The scissors (or X-shaped) rotor concept consists in altering the angle between successive blades as well as their height on the hub to create a balanced anisotropic rotor. This has already been numerically examined for its effect upon BVI noise, and it was found that a 5% decrease in two blades of a four bladed scaled UH-60 rotor resulted in 4.8 dB in rotor noise [41]. So far it has been mostly investigated for tail rotors, as on the Apache and the Mi-28 [42]. In the latter design, noise reductions up to 5 dB have been found. This concept will likely significantly modify the trajectory of the tip vortices and thus change the entire nature of the wake and BVI events. The Variable Geometry Rotor is designed to not only accommodate different axial and azimuthal spacing, but also different blade length.
Improved rotor performances were obtained with a modified Sikorsky CH-53A 6-bladed rotor [43]. Measurements in hover and analytical predictions also showed overall noise reductions, but BVI conditions were not specifically studied. However, the authors pointed out that the potential for BVI noise and vibration reductions existed due to the important changes in vortex trajectory observed. The VGR concept has evolved into the
Variable Diameter Tiltrotor (VDTR), an adaptive passive design approach, which is was considered by Sikorsky for the civil tiltrotor application [44].
13 1.4 Active Methods
In contrast to the passive design concepts considered in the previous sections, concepts that require power input during each rotor revolution are termed active.
Actuation can be achieved by numerous means, such as trailing edge flaps, root pitch actuation, and active blade twist. Implementation of these methods, particularly those actuated in the rotating frame, has become more practical with the development of smart materials which can be incorporated into the structure of the blade, such as piezoelectric or magnetostrictive actuators [45]. Active control for BVI alleviation can be divided in
Higher Harmonic Control (HHC) where the blade is actuated from the fixed frame through the swashplate and Individual Blade Control (IBC) where a blade can be individually actuated in the rotating frame.
1.4.1 Higher Harmonic Control
The higher harmonic pitch control method (HHC) uses the existing rotor and adds a higher harmonic pitch control fed through the swashplate (Figure 1.8). This method of actuation limits inputs to N-1, N, and N+1 per rev where N is the number of blades.
Though originally developed to reduce vibrations, HHC has been also shown to be effective for reducing BVI noise. A test conducted in the Langley Transonic Dynamics
Tunnel (TDT) in the late 80s on the Aeroelastic Rotor Experimental System (ARES), in
Freon-12 gas found reductions up to 4.7 dB were in a low speed descent condition with a
4 per rev input [46]. Unfortunately, the inputs designed to reduce noise were found to increase vibration levels. An MBB BO-105 HHC test in the DNW conducted around the
14 same time found similar results [47]. Neither the TDT nor the DNW tests contained any noise directivity results. These two groups combined to conduct a new set of experiments in the DNW in 1991, using a BO-105 model [48]. A phased combination of 1 and 4 per rev actuation of the swashplate (the BO-105 is four bladed) were used to apply 3, 4, and 5 per rev controls. Results showed that not only were similar noise reductions
(approximately 5 dB) obtained as in previous tests, but that the BVI noise directivity is drastically changed. The authors recommended further research consisting of IBC inputs using closed loop control with feedback from body-mounted microphones or pressure sensors on the blades. In later testing, a closed-loop control system showed reductions in noise similar to the open loop, around 5 dB, with the additional advantage of reduced vibrations [49].
In 1994, a new test was conducted with the BO-105 rotor in the DNW, under the program name HART (Higher harmonic control Aeroacoustics Rotor Test). With extensive instrumentation for the measurement of blade surface pressure distribution, blade deformation, acoustic signatures, tip vortex geometry and strength, an unprecedented amount data was obtained [50]. This data has been used to validate many prediction codes and continues to remain a major source of experimental data for the validation of computational results. In the HART test it was found that 3 per rev actuation was able to generate a 6-dB noise reduction at the expense of increased vibrations.
Likewise, the input for minimum vibration was found to increase noise. For other actuation schedules, smaller reductions in noise were achieved simultaneously with vibration reductions. The miss-distance was shown to have increased in the minimum noise case (whereas it had decreased for minimum vibration cases) and was thought to be
15 the main factor in BVI noise reductions [51]. It has been reported that a decrease in miss- distance produced the decrease in vibration in the minimum vibration case, though a later study suggests that the change of inclination of the interaction in the blade-shaft plane was responsible, and that the concept of miss-distance is imprecise in this case [5].
Several experimental helicopters have been flight tested with HHC systems.
A 3-bladed SA-439 Gazelle from Aérospatiale, with both open and closed loop actuation, showed BVI noise reductions around 5 dB but with increased lower frequency noise [52]. A wind tunnel test of the XV-15 rotor showed a 10 dB reduction in peak noise levels [53], though when a closed loop controller was utilized (restricted to using only 2 per rev HHC) only 5.3 dB reductions were achieved utilizing microphone feedback. It is thought that allowing the controller to optimize several frequencies of HHC input would result in greater reductions. Actual results may not be as beneficial as reported, though, as the new peak noise locations may have fallen outside of the microphone sweep area. It is questionable whether microphones at fixed locations attached to the rotorcraft can provide enough feedback information to reduce BVI noise when a common result of
HHC input is a change in the directivity of BVI noise (due changes in the dominant interaction geometry or a secondary interaction becoming dominant).
In 2004 a CFD analysis of several rotors with trailing edge flaps was conducted that determined that changing miss-distance was the primary reason for BVISPL reductions [54]. It was estimated the reductions of up to 5 dB were possible.
16
HHC Actuators
Figure 1.8: Configuration for HHC.
1.4.2 Individual Blade Control
Active control for BVI can also be achieved through Individual Blade Control
(IBC) methods, where, unlike HHC, each blade has the potential to be individually actuated in the rotating system. This can be used to generate harmonic control outside of the range of N-1 through N+1 per rev obtained with HHC actuators in the fixed frame, or a pulsed input over a small portion of the azimuth. Another method of generating IBC is through trailing edge flaps, as seen in Figure 1.9. Primary control (1 per rev) via servo- flaps is already commonly used on Kaman Helicopters, such as the SH-2G Super
Seasprite and the K-Max, where the flap actuation is achieved by a non-rotating cam, which could also be used to input HHC. Other actuation options include the use of smart materials.
McDonnell Douglas Helicopter Systems (MDHS) has extensively tested tip- mounted trailing-edge flaps in the wind tunnel [55-57], achieving a 4 dB BVI noise reduction and showing the need for a careful phasing of the actuation since a noise increase of 8 dB was created with other schedules. The authors concluded that a flap
17 deflection at a particular azimuthal location reduces the strength of the tip vortex that interacts with the succeeding blades. MDHS was able to significantly reduce BVI induced vibration by applying a 5 per rev input [58].
Use of non-harmonic inputs were investigated on a one seventh scale AH1-G rotor and found that a 20 degree negative flap input over a range of 120 degrees was able to reduce the average noise level by 5 dB, but at the cost of a 58% increase in power consumption [59]. This input was less than ideal, however, and the authors stress the importance of correctly placing the input as different azimuthal locations will result in significant noise increases. The correct placement is dependent upon the flight condition.
Trailing Edge Flaps
Figure 1.9: Configuration for trailing edge flaps.
Another method of generating IBC controls is the use of root pitch actuators in the rotating-frame (as opposed to the fixed frame as with HHC) as seen in Figure 1.10. A
BO-105 full-scale rotor fitted with such actuators was tested in the NASA Ames 40 by
80-Foot Wind Tunnel. The actuators were controlled through a hydraulic slipring. 4 per rev inputs were found to be the most effective in reducing vibrations, by as much as 70%, while 5 per rev yielded less reduction. 2 per rev inputs were found to make significant
18 changes to the rotor trim state. For BVI noise reduction, 2 per rev was the best single harmonic input, with noise reductions up to 6.7 dB. Multiple frequency inputs were also tested and a 2 and 5 per rev combination was found to give 8.8 dB noise reductions and a
90% decrease in 4 per rev vibrations [60, 61]. Advanced schedules, such as pulses or wavelets were also tested but failed to give noise reductions. The authors attributed this to the difficulty of designing the proper schedule due to the complex dynamic behavior of the blades [62, 63]. This stresses a need for a detailed understanding of the mechanisms through which these inputs would obtain noise reductions, and, if possible, information on the strength, core size and location of the tip vortex [64]. Flight tests were also conducted on a BO-105 helicopter with hydraulic root pitch actuators and showed the potential for a 6-dBA noise reduction, with 2 per rev and a 3 and 5 per rev combination yielding the most benefits [65]. It was noted, however, that the benefits were highly dependent on the flight condition [66]. A numerical simulation actually calculated the respective benefits of flaps versus root actuation for IBC [67]. The study was conducted on a 5-bladed MD-900 rotor in a flight condition that produced strong BVI events and concluded that at moderate speeds flaps produced greater noise reductions, but root pitch demonstrated a greater ability to control retreating side BVI.
A full-scale test of single-frequency harmonic IBC through root pitch actuation on a UH-60 showed that 2/rev actuation could reduce BVI noise by up to 12 dB at some locations and on average 6-8 dB over the range of observer locations [68]. Reductions of peak BVISPL were not reported. It was again reported that inputs that produced the most dramatic reductions in BVI noise were accompanied by a significant increase in vibratory hub loads. Another study comparing the two effects on a BO-105 determined that they
19 produced similar levels of peak observer-plane noise reduction, though root pitch actuation required more power input [69, 70]. This work was continued by incorporating passive methods for BVI noise and vibration reduction in conjunction with trailing-edge flaps to produce a new rotor designed specifically to minimize BVI [71]. Full-scale testing of a 5-bladed MD 900 with piezoelectrically actuated trailing-edge flaps utilized open loop harmonic IBC to reduce the peak BVISPL on an observer plane [72, 73]. It was found that a 2-7 dB reduction in noise was possible depending upon the flight condition. This study also reported that the ideal case for noise reduction corresponded with a severe increase in vibratory hub loads.
IBC Actuators
Figure 1.10: Configuration for IBC.
Another method for BVI alleviation is the use of air injection over the outboard section of the blade. In one study, run numerically on the MD-900 rotor in low speed descent, blowing and suction are used to modify the "effective surface" of the blade to artificially change airfoil thickness and camber with azimuth or flight condition.
Although the authors recommend further research, this method shows potential to temporally alleviate impulsive leading edge surface pressures [74]. Applying blowing to the blade tip has shown the potential for dramatically increasing the core size of the trailing edge vortex, weakening the peak velocity within the vortex [75, 76, 77]. This is
20 suggested to be likely capable of reducing the resulting BVI noise. Another active flow control method, leading edge blowing, ejects air from the leading edge as a vortex approaches the blade [78]. This has been shown to significantly reduce BVI-induced vibration and is thought to possibly reduce BVI noise. Other methods suggested include a reverse swirl generator or a jet engine at the tip of the blade [38].
Another method for IBC is actively changing the blade’s twist. Piezoelectric materials are embedded into the structure of the blade itself, allowing rabid changes in blade twist. Feasibility studies on active twist blades have predicted 2 to 4 dB reduction in the strongest BVI events, with up to 10 dB reduction in weaker events. These studies assume that up to 2 degrees of twist will be obtained with the blades [79]. Tests have found the blades capable of producing up to 1.4 degrees of twist can reduce BVI noise by up to 3 dB, but this comes at the cost of increased low-frequency noise (up to 7 dB) and increased vibration (up to 100%) [80]. Another analytical study showed that active twist was capable of producing reductions in BVI noise on par with trailing-edge flaps [81].
1.5 Overview
The objective of this dissertation is to evaluate the effectiveness of using physically motivated pulse-type Individual Blade Control for reducing the noise associated with the Blade-Vortex Interactions seen in low-speed descent. This differs from other IBC work for the reduction of BVI noise in that rather than parametrically varying an IBC input to identify the parameters that give the greatest reduction in BVI noise and only afterwards attempting to identify the mechanism of noise reduction, this
21 dissertation first seeks to identify a mechanism for noise reduction and then tailor an IBC input to target that mechanism. A method such as this is necessary to explore the effectiveness of non-harmonic IBC inputs, as there are too many variables to these inputs to be able to simply perform parametric sweeps.
The first step towards this goal is developing an understanding of the major parameters that affect the severity of the interaction as seen in Chapter 2. This allows for the identification of parameters that will produce significant reductions in IBC noise for moderate changes. Second, inputs designed specifically to alter the parameters previously identified as key are explored in Chapter 3. This includes examining different forms of inputs, and the effectiveness of different amplitudes, azimuthal locations, and azimuthal ranges. In order to implement these inputs, a feedback mechanism for use in closed loop control is necessary. Different mechanisms are evaluated in Chapter 4, including a newly developed metric based upon blade pressures.
References
1. Johnson, W., “Helicopter Theory,” Dover Publications, Inc., New York, 1994.
2. Schmitz, F. H., “Rotor Noise” in Aeroacoustics of Flight Vehicles: Theory and Practice, NASA Langley Research Center, edited by H. H. Hubbard, Vol. 1, pp. 65-149, 1991.
3. Leverton, J. W., “The Potential for Quiet Rotorcraft,” First Joint CEAS/AIAA Aeroacoustic Conference, Munich, Germany, June 1995.
4. “Bell/Agusta 609 - Aircraft attributes and Characteristics,” Bell Helicopter, URL http://www.bellhelicopter.textron.com/content/products/tiltRotor/609/ articles/609attrib.html.
22 5. Malovrh, B., Gandhi, F., and Tauszig, L., “Sensitivity of Helicopter Blade- Vortex-Interaction Noise and Vibration to Interaction Parameters,” Journal of Aircraft, Vol. 42, No. 3, pp. 685-697, May 2005.
6. Hardin, J. C., Lamkin, S. L., “Concepts for Reduction of Blade/Vortex Interaction Noise,” Journal of Aircraft, Vol. 24, No. 2, pp. 120-125, Feb. 1987.
7. Gallman, J. M., “Parametric Computational study of Isolated Blade-Vortex Interaction Noise,” AIAA Journal, Vol. 32, No. 2, pp. 232-238, Feb. 1994.
8. Kitaplioglu, C., Caradonna, F. X., and Burley, C. L., “Parallel Blade-Vortex Interactions: An Experimental Study and Comparison with Computations,” Proceedings of the 2nd American Helicopter Society International Aeromechanics Specialists’ Conference, Bridgeport, Connecticut, Oct. 1995.
9. Leighton, K. P., “Research on Model Helicopter Rotor Blade Slap at Moderate Tip Speeds,” Proceedings of the 36th American Helicopter Society Annual Forum, Washington, DC, May 1980.
10. Anon., “Fly Neighborly Guide,” Helicopter Association International, 1991.
11. Schmitz, F. H., Gopalan, G., Sim, B. W.-C., “Flight-path Management/Control Methodology to Reduce Helicopter Blade-Vortex Interaction Noise,” Journal of Aircraft, vol. 39, no. 2, March 2002, p. 193-205.
12. Leverton, J. W., “Twenty-Five Years of Rotorcraft Aeroacoustics: Historical Prospective and Important Issues,” Journal of sound and Vibration, Vol. 133, No. 2, pp. 261-287, 1989.
13. Hindson, W. S., Chen, R. T. N., “Flight Tests of Noise Abatement Approaches for Rotorcraft Using Differential GPS Guidance,” Proceedings of the 51st American Helicopter Society Annual Forum, Ft. Worth, TX, May 1995.
14. Jacobs, E. W., O’Connell, J. M., Conner, D. A., Rutledge, C. K., Shigemoto, F., Fleming, G. G., “Acoustic Flight Testing of a Boeing MD Explorer and a Sikorsky S-76B Using a Large Area Microphone Array,” Proceedings of the Technical Specialists’ Meeting for Rotorcraft Acoustics and Aerodynamics, Williamsburg, VA, Oct. 1997.
15. Jacobs, E. W., Chen, R. T. N., Santa Maria, O. L., “The Development and Flight Test Demonstration of Noise Abatement Approach Procedures for the Sikorsky S- 76,” Proceedings of the Technical Specialists’ Meeting for Rotorcraft Acoustics and Aerodynamics, Williamsburg, VA, Oct. 1997.
16. Janakiram, R. D., O’Connell, J. M., Fredrickson, D. E., Conner, D. A., Rutledge, C. K., “Development and Demonstration of Noise Abatement Approach Flight Operations for a Light Twin-engine Helicopter – MD Explorer,” Proceedings of
23 the Technical Specialists’ Meeting for Rotorcraft Acoustics and Aerodynamics, Williamsburg, VA, Oct. 1997.
17. Abello, J. C. and George, A. R., “Wake Displacement Modifications to Reduce Rotorcraft Blade-Vortex Interaction Noise,” Journal of Aircraft, Vol. 41, No. 2, pp. 290-303, Mar./Apr. 2004.
18. Leverton, J. W., “Helicopter Noise Certification. Past-Present-Future,” Proceedings of the 18th European Rotorcraft Forum, Avignon, France, September 1992.
19. Schmitz, F. H., “Reduction of Blade-Vortex Interaction (BVI) Noise through X- Force Control,” Journal of the American Helicopter Society, Vol. 43, No. 1, pp.14-24, Jan. 1998.
20. Munsky, B. and Gandhi, F., “Analysis of Helicopter Blade-Vortex Interaction Noise with Flight Path or Attitude Modification,” Journal of the American Helicopter Society, Vol. 50, No. 2, pp. 123-137, Apr. 2005.
21. Janakiram, R. D., “NASA/AHS Rotorcraft Noise Reduction Program: McDonnell Douglas Helicopter Company Accomplishments,” Vertiflite, Vol. 43, pp. 64-68, Nov.-Dec. 1988.
22. Mantay, W. R., Shidler, P. A., Campbell, R. L., “Some Results of the Testing of a Full-Scale Ogee-Tip Rotor,” Journal of Aircraft, Vol. 6, No. 3, pp.215-221, Mar. 1979.
23. Perry, F. J., Jones, A. F., Wilby, P. G., “The BERP rotor - How does it work, and what has it been doing lately?” Vertiflite, vol. 44, no. 2, p. 44-48, 1998.
24. Harrison, R., Stacey, S., and Hansford, B., “The Design, Development, and Testing of an Advanced Rotor Blade,” Proceedings of the 64th American Helicopter Society Annual Forum, Montreal, Canada, April 2008.
25. Mueller, R. H. G., “Special Vortices at a Helicopter Rotor Blade,” Journal of the American Helicopter Society, Vol. 35, pp. 16-22, Oct. 1990.
26. Mueller, R. H. G., “Winglets on Rotor Blades in Forward Flight - A Theoretical and Experimental Investigation,” Vertica, Vol. 14, No. 1, pp. 31-46, 1990.
27. Niesl, G., Arnaud, G., “Low Noise Design of the EC 135 Helicopter,” Proceedings of the 52nd American Helicopter Society Annual Forum, Washington, DC, June 1996.
28. Lorber, P. F., “Blade-Vortex Interaction Data Obtained from a Pressure- Instrumented Model UH-60A Rotor at the DNW,” Journal of the American Helicopter Society, Vol. 38, No. 3, pp. 26-43, July 1993.
24 29. Lorber, P. F., “Aerodynamic Results of a Pressure-Instrumented Model Rotor Test at the DNW, ” Proceedings of the 46th American Helicopter Society Annual Forum, Washington, DC, May 1990.
30. Mueller, R. H. G., Nsi Mba, M., Aymard, E., Favier, D., Berton, E., Maresca, C., “Visualization and Measurement of Helicopter Rotor Flow with Swept Back Tip Shapes at Hover Flight Using the 'Flow Visualization Gun' Time Line Technique,” Experiments in Fluids, Vol. 21, No. 3, pp. 161-169, July 1996.
31. De Terlizzi, M., Friedmann, P. P., “Aeroelastic Response of Swept Tip Rotors Including the Effects of BVI, ” Proceedings of the 54th American Helicopter Society Annual Forum, Washington, DC, May 1998.
32. Toulmay, F., Arnaud, G., Falchero, D., Villat, V., “Analytical Prediction of the Rotor Dynamics for Advanced Geometry Blades,” Proceedings of the 52nd American Helicopter Society Annual Forum, Washington, DC, June 1996.
33. Schaffar, M., Haertig, J., Gnemmi, P., “Effect of Non-Rectangular Blade Tips on BVI Noise for a Two-Bladed Rotor,” Proceedings of the 16th European Rotorcraft Forum, Glasgow, Scotland, Sept. 1990.
34. Han, Y. O and Leishman, J. G., “Investigation of Helicopter Rotor-Blade-Tip- Vortex Alleviation Using a Slotted Tip,” AIAA Journal, Vol. 42, No. 3, pp. 524- 535, Mar. 2004.
35. You, J. Y., Kwon, O. J., and Han, Y. O., “Viscous Flow Simulation of Rotor Blades with Tip Slots in Hover,” Journal of the American Helicopter Society, Vol. 54, No. 1, Jan. 2009.
36. Wake, B. E., Egolf, T. A., “Evaluation of a Doubly-Swept Blade Tip for Rotorcraft Noise Reduction,” NASA-CR 189677, NASA Langley Research Center, Oct. 1992.
37. Wells, V. L., Glinka, A. T. IV, “The Flip Tip - A Novel Method for Reducing Blade-Vortex Interaction Noise,” Innovation in Rotorcraft Technology, Proceedings of the Royal Aeronautical Society, London, United Kingdom, June 1997.
38. White, R. P., “An Investigation of the Vibratory and Acoustic Benefits Obtainable by the Elimination of the Blade Tip Vortex,” Journal of the American Helicopter Society, Vol. 18, No. 4, pp. 35-44, Oct. 1973.
39. Baeder, J. D., “Passive Design for Isolated Blade-Vortex Interaction Noise Reduction,” Proceedings of the 53rd American Helicopter Society Annual Forum, Virginia Beach, Virginia, April 1997.
25 40. Schultz, K.-J. Splettstoesser, W. R., “Helicopter Main Rotor/Tail Rotor Noise Radiation Characteristics from Scaled Model Rotor Experiments in the DNW,” Proceedings of the 49th American Helicopter Society Annual Forum, St. Louis, Missouri, May 1993.
41. Tauszig, L., Gandhi, F., “Influence of Blade-To-Blade Dissimilarities on the Alleviation of Blade-Vortex Interactions,” Proceedings of the 55th American Helicopter Society Annual Forum, Montreal, Quebec, Canada, May 1999.
42. Rozhdestvensky, M. G., “Scissors Rotor Concept: New Results Obtained,” Proceedings of the 52nd American Helicopter Society Annual Forum, Washington, DC, June 1996.
43. Mantay, W. R., Rorke, J. B., “The Evolution of the Variable Geometry Rotor,” Proceedings of the American Helicopter Society Symposium on Rotor Technology, Mideast Region, Aug. 1976.
44. Fradenburgh, E. A., Matuska, D. G., “Advancing Tiltrotor State-of-the-Art with Variable Diameter Rotors,” Proceedings of the 48th American Helicopter Society Annual Forum, Washington, DC, June 1992.
45. Chopra, I., “Recent Progress on the Development of a Smart Rotor System,” Proceedings of the 26th European Rotorcraft Forum, The Hague, Netherlands, Sept. 2000.
46. Brooks, T. F., Booth, E. R. Jr., Jolly, J. R. Jr., Yeager, W. T. Jr., Wilbur, M. L., “Reduction of Blade-Vortex Interaction Noise Through Higher Harmonic Pitch Control,” Journal of the American Helicopter Society, Vol. 35, pp. 86-91, Jan. 1990.
47. Splettstoesser, W. R., Lehman, G., Van der Wall, B., “Higher Harmonic Control of a Helicopter Rotor to Reduce Blade-Vortex Interaction Noise,” Proceedings of the 15th European Rotorcraft Forum, Sept. 1989.
48. Brooks, T. F., Booth, E. R. Jr., Boyd D. D. Jr., Splettstoesser, W. R., Schultz, K.- J., Kube, R., Niesl, G. H., Streby, O., “Analysis of a Higher Harmonic Control Test to Reduce Blade Vortex Interaction Noise,” Journal of Aircraft, Vol. 31, No. 6, pp. 1431-1439, Dec. 1994.
49. Kube, R., Achache, M., Niesl, G., Splettstoesser, W. R., “A Closed Loop Controller for BVI Impulsive Noise reduction by Higher Harmonic Control,” Proceedings of the 48th American Helicopter Society Annual Forum, Washington, DC, June 1992.
50. Splettstoesser, W. R., Kube, R., Wagner, W., Seelhorst, U., Boutier, A., Micheli, F., Mercker, E., Pengel, K., “Key Results from a Higher Harmonic Control
26 Aeroacoustic Rotor Test (HART),” Journal of the American Helicopter Society, Vol. 42, No. 1, pp. 58-78, Jan. 1997.
51. Yu, Y., “Miss Distance for Rotor Blade-Vortex Interaction Noise Reduction,” Proceedings of the 2nd AIAA/CEAS Aeroacoustics Conference, State College, PA, May 1996.
52. Polychroniadis, M., “Generalized Higher Harmonic Control, Ten Years of Aérospatiale Experience,” Proceedings of the 16th European Rotorcraft Forum, Glasgow, UK, 1990.
53. Nguyen, K., Betzina, M., and Kitaplioglu, C., “Full-Scale Demonstration of Higher Harmonic Control for Noise and Vibration Reduction on the XV-15 Rotor,” Proceedings of the 56th American Helicopter Society Annual Forum, Virginia Beach, Virginia, May 2000.
54. Aoyama, T. Yang, C, and Saito, S., “Numerical Analysis of BVI Noise Reduction by Active Tab,” Proceedings of the 60th American Helicopter Society Annual Forum, Baltimore, MD, June 2004.
55. Dawson, S., Straub, F., “Design, Validation, and Test of a Model Rotor with Tip Mounted Active Flaps,” Proceedings 50th American Helicopter Society Annual Forum, Washington, DC, May 1994.
56. Dawson, S., Marcolini, M., Booth, E., Straub, F., Hassan, A., Tadghighi, H., Kelly, H., “Wind Tunnel Test of an Active Flap Rotor: BVI Noise and Vibration Reduction,” Proceedings of the 51st American Helicopter Society Annual Forum, Fort Worth, Texas, May 1995.
57. Marcolini, M. A., Booth, E. R. Jr., Tadghighi, H., Hassan, A. A., Smith, C. D., Becker, L. E., “Control of BVI Noise Using an Active Trailing Edge Flap,” Proceedings of the American Helicopter Society Vertical Lift Aircraft Design Conference, San Francisco, California, Jan. 1995.
58. Straub, F. K., “Active Flap Control for Vibration Reduction and Performance Improvement,” Proceedings of the 51st American Helicopter Society Annual Forum, Fort Worth, Texas, May 1995.
59. Charles, B. D., Tadghighi, H., and Hassan, H. H., “Effects of a Trailing Edge Flap on the Aerodynamics and Acoustics of Rotor Blade-Vortex Interactions,” Proceedings of the 14th DGLR/AIAA Aeroacoustics Conference, Aachen, Germany, May 1992.
60. Jacklin, S. A., Blaas, A., Teves, D., Kube, R., “Reduction of Helicopter BVI Noise, Vibration and Power Consumption Through Individual Blade Control,” Proceedings of the 51st American Helicopter Society Annual Forum, Fort Worth, Texas, May 1995.
27 61. Jacklin, S. A., Nguyen, K. Q., Blaas, A., Richter, P., “Full-Scale Wind Tunnel Test of a Helicopter Individual Blade Control System,” Proceedings of the 50th American Helicopter Society Annual Forum, Washington, DC, May 1994.
62. Swanson, S. M., Jacklin, S. A., Blaas, A., Niesl, G., Kube, R., “Acoustic Results from a Full-Scale Wind Tunnel Test Evaluating Individual Blade Control,” Proceedings of the 51st American Helicopter Society Annual Forum, Fort Worth, Texas, May 1995.
63. Swanson, S. M., Jacklin, S. A., Niesl, G., Blaas, A., Kube, R., “Effect of Individual Blade Control on Noise Radiation,” Proceedings of the AGARD Aeroacoustic Conference, Berlin, Germany, Oct. 1994.
64. Swanson, M. S., Jacklin, S. A., Blaas, A., Kube, R., Niesl, G., “Individual Blade Control Effects On Blade-Vortex Interaction Noise,” Proceedings of the 50th American Helicopter Society Annual Forum, Washington, DC, May 1994.
65. Splettstoesser, W. R., Schultz, K.-J., Van der Wall, B., Buchholz, H., Gembler, W., Niesl, G., “The Effect of Individual Blade Pitch Control on BVI Noise - Comparison of Flight Test and Simulation results,” Proceedings of the 24th European Rotorcraft Forum, Marseilles, France, Sept. 1998.
66. Kube, R., van der Wall, B. G., “IBC Effects on BVI Noise and Vibrations A combined Numerical and Experimental Investigation,” Proceedings of the 55th American Helicopter Society Annual Forum, Montreal, Quebec, Canada, May 1999.
67. Charles, B., Tadghighi, H., Hassan, A., “Higher Harmonic Actuation of Trailing- Edge Flaps for Rotor BVI Noise Control,” Proceedings of the 52nd American Helicopter Society Annual Forum, Washington, DC, June 1996.
68. Jacklin, S. A., Haber, A., de Simone, G., Norman, T. R., Kitaplioglu, C., and Shinoda, P., “Full-Scale Wind Tunnel Test of an Individual Blade Control System for a UH-60 Helicopter,” Proceedings of the 58th American Helicopter Society Annual Forum, Montreal, Canada, June 2002.
69. Patt, D., Liu, L., Friedmann, P. P., “Active Flaps for Noise Reduction: A Computational Study,” Proceedings of the 60th American Helicopter Society Annual Forum, Baltimore, MD, June 2004.
70. Patt, D., Liu, L., and Friedmann, P., “Helicopter Noise Reduction by Actively Controlled Flaps,” 11th AIAA/CEAS Aeroacoustics Conference, Mar. 2005.
71. Glaz, B., Friedmann, P., and Bagnoud, F.-X., “Vibration and Noise Reduction of Helicopter Rotors using an Active/Passive Approach,” AHS Specialists Conference on Aerodynamic 2008, San Francisco, CA, Jan. 2008.
28
72. Straub, F. K., Anand, V. R., Birchette, T. S., and Lau, B. H., “Wind Tunnel Test of the SMART Active Flap Rotor,” Proceedings of the 65th American Helicopter Society Annual Forum, Grapevine, TX, May 2009.
73. Janakiram, R. D., Sim, B. W., Kitaplioghu, C., Straub, F., “Blade-Vortex Interaction Noise Characteristics of a Full-Scale Active Flap Rotor,” Proceedings of the 65th American Helicopter Society Annual Forum, Grapevine, TX, May 2009.
74. Hassan, A. A., Straub, F. K., Charles, B. D., “Effects of Surface Blowing/Suction on the Aerodynamics of Helicopter Rotor Blade-Vortex Interactions (BVI) - A Numerical Simulation,” Proceedings of the 52nd American Helicopter Society Annual Forum, Washington, DC, June 1996.
75. Pegg, R. J., Hosier, R. N., Balcerak, J. C., Johnson, H. K., “Design and Preliminary Tests of a Blade Tip Air Mass Injection System for Vortex Modification and Possible Noise reduction on A Full-Scale Helicopter Rotor,” NASA TM X-3314, Dec. 1975.
76. Duraisamy, K., Baeder, J. D., “Control of Helicopter Rotor Tip Vortex Structure using Blowing Devices,” Proceedings of the 60th American Helicopter Society Annual Forum, Baltimore, MD, June 2004.
77. Liu, Y. and Sankar, L. N., “Computational Evaluation of Controlling Flap Edge Vortex and Tip Vortex Effects with Circulation Control Technique,” 45th AIAA Aerospace Sciences Meeting, Reno, NV, Jan. 2007.
78. Weiland, C. and Vlachos, P., “A Mechanism for Mitigation of Blade-Vortex Interaction using Leading Edge Blowing Flow Control,” Experiments in Fluids, Vol. 47, Num. 3, pp. 411-426, Sept. 2009.
79. Chen, P. C., Evans, R. A. D., Niemczuk, J., and Baeder, J. D., “Blade-Vortex Interaction Noise Reduction with Active Twist Smart Rotor Technology,” Smart Materials and Structures, vol. 10, no. 1, Feb. 2001, p. 77-85.
80. Booth Jr., E. R., Wilbur, M. L., “Acoustic Aspects of Active-Twist Rotor Control,” Journal of the American Helicopter Society, Vol. 49, No. 1, pp. 3-10, Jan. 2004.
81. Thepvongs, S., Cesnik, C. E. S., and Voutsinas, S. G., “Numerical Investigation of Integral Twist Actuation for BVI Noise Reduction,” Proceedings of the 62nd American Helicopter Society Annual Forum, Phoenix, AZ, June 2006.
29
Chapter 2
Sensitivity of BVI-Induced Noise and Vibration to Variations in Individual Interaction Parameters
Chapter two presents a comprehensive analysis of the sensitivity of BVI noise and
BVI-induced vibratory loading to changes in various interaction parameters. The interaction parameters considered are (i) strength of the interacting vortex, (ii) vortex core radius, (iii) blade-vortex miss-distance, (iv) angle of interaction in the blade-shaft plane, (v) angle of interaction in the rotor disk plane, (vi) spanwise or radial location of the interaction, (vii) spanwise length of interaction, and (viii) blade lift at the time of interaction. Though the effects of parameters such as vortex strength and miss-distance upon noise and vibration have been studied, less work has been done on understanding the effects of parameters such as the radial location of the interaction and interaction angles. The comprehensive set of results allows a direct comparison of the reductions in
BVI noise and BVI-induced vibratory loading achievable by effecting changes in the various blade-vortex interaction parameters, vis-à-vis the effort (the magnitude of change in the parameter) that is required. These results will then facilitate the tailoring of IBC inputs to most effectively reduce BVI noise and vibration by exploiting a thorough understanding of the contributions of the various factors. For example, with an understanding of the relative influence of change in vortex core size, vortex strength, and blade-vortex miss-distance on BVI noise, a decision could be made whether blade tip design (to increase vortex core size), active rotor control (to decrease vortex strength or 30 increase blade-vortex miss-distance), or change in flight-path (to increase blade-vortex miss-distance), is the most suitable approach for BVI alleviation.
2.1 Analysis Method
Numerical results in the present study are nominally based on a model rotor previously examined at NASA Langley Research Center [1]. The 4-bladed rotor considered has a 2.856 meter diameter, a 9.1 cm chord, and a 217.11 m/s tip speed. In the present simulations, a NACA 0012 symmetric airfoil with zero blade twist is examined.
The rotor shaft is held vertical, and both the forward velocity and collective pitch are set to zero. The rotating blades are then allowed to interact with an imposed straight-line vortex at ψ = 60º (see Figure 2.1). Although the choice of this azimuthal location is arbitrary for the present configuration (zero advance ratio, shaft tilt, and rotor thrust), strong interactions are known to occur around this region in typical low-speed descent conditions that yield strong blade-vortex interactions. Thus, simulating an interaction in this region produces a BVI noise ‘footprint’ that looks qualitatively similar to others commonly seen in the literature. It should be noted that as this rotor is non-lifting, it does not produce its own tip vortices and the present approach of artificially imposing a vortex element makes it possible to vary the interaction-parameters as desired and isolate the effects of the various parameters that govern BVI. A self-generated helicopter vortex- wake system, on the other hand, is highly complex, rendering it virtually impossible to quantify the effects of changes in individual blade-vortex interaction parameters. This has 31 been recognized by other researchers as well, and was the rationale behind their using an externally generated vortex to study variations in a few interaction parameters [2, 3].
180˚
Ω
270˚ 90˚ Interacting Vortex 60˚
Reference Blade
ψ = 0˚
Figure 2.1: Baseline blade-vortex interaction geometry.
The vortex element interacting with the rotor blade is assumed to have a viscous core of radius rc, and following References 4 and 5, the tangential velocity, vθ, at a
distance r from the center of the vortex is given by Equation 2.1. 32
r v r 2 4 4 2.1 rc r
Here, is the strength of the vorticity. The vortex core radius, rc, is assumed to be
20% of the blade chord, unless otherwise stated. Using the above viscous core/vortex velocity profile model, the rotor inflow is calculated. Blade-element theory is then used to predict the lift on the blade as it traverses around the azimuth. The lift over an individual blade is integrated along the spanwise direction to obtain the blade root vertical load. Summing the root loads over all the blades as the rotor undergoes one
revolution yields the BVI-induced hub vertical vibratory loading. The lift data is also input into the acoustic code WOPWOP [6] to obtain acoustic pressure time histories at
several grid points or “observer locations” on a plane one rotor diameter below the disk.
From the acoustic pressure time histories, the BVI Sound Pressure Level (BVISPL) is
calculated by considering the 6th to the 40th harmonics of the blade passage frequency.
The BVISPL plots yield the direction of the radiated BVI noise as well as the peak noise
levels. Changes in BVI-induced vibratory loading and peak noise levels with changes in
interaction-parameters are then examined in detail. Peak BVISPL is used throughout this
dissertation as a gauge of BVI noise. Though it is possible to utilize methods that take
into account sound pressure levels over the entirety of the observer plane, peak BVISPL allows for easy comparison with other work and is independent of the size of the observer
plane itself (so long as the peak level is included) and, as will be seen, corresponds well
with overall noise levels.
To obtain the high-resolution BVI airloads, 640 azimuthal stations and 45 radial
stations are used. It has been shown that such an azimuthal resolution of nearly half 33 degree is adequate to accurately capture the impulsive BVI event [7]. Also, for high aspect ratio blades in the absence of shock, lifting line theory produces accurate predictions of far field BVI noise [8].
2.2 Baseline
The baseline BVI event considered is at an azimuthal location of 60˚ with a vortex parallel to the blade intersecting it over the outer half of the blade’s length (R/2 to R, see
Figure 2.1) with zero miss-distance, a vortex core radius of rc = 0.2 chords, and a non-
dimensional vortex strength, /(cVtip), of 0.101. This non-dimensional vortex strength corresponds to a circulation of Γ = 2 m2/s (in the HART program [9], the strengths of the interacting vortices on the advancing and retreating side varied between 1.1 – 2.8 m2/s).
The baseline orientation of the vortex and blade represent a worst-case scenario (zero miss-distance and perfectly parallel), but was chosen as it falls in the middle of the range of angles in the blade-shaft plane, angles in the disk plane, and miss-distances considered.
The nominal values of the interaction parameters given above are used for all further cases, unless otherwise stated. Figure 2.2 illustrates the baseline non-dimensional sectional lift distribution over the rotor disk (non-dimensionalized by ρac(ΩR)2). As the pitch of this uncambered untwisted blade is zero, and the change in inflow as the blade passes the vortex is the only contributor to the angle of attack, the lift distribution appears symmetric about ψ = 60˚. It is also seen that the azimuthal gradients in lift (dL/d) are largest near the interaction location, and the rapidly reduce away from ψ = 60˚. Due to the nature of the interaction, the vertical root loads (obtained by integration of the BVI- 34 induced lift along the blade length) are highly impulsive, as seen in Figure 2.3. As a consequence, the BVI-induced vibratory hub vertical loading, obtained by integrating the blade root loads over the azimuth and summing over the number of blades, is observed to have a very strong higher harmonic content (Figure 2.4). In fact, the magnitude of the
8/rev component exceeds that of the 4/rev component. There are no steady (zeroth harmonic) loads as the blade generates no lift other than that due to inflow from the imposed vortex.
=180 100
75
50
25
= 270 0.0004 0 =90
0.0 013 0 X(%Radius) -25 .00 30 0 .01 35 -50
-75 =0 -100 100 50 0 -50 -100 Y(%Radius)
Figure 2.2: Baseline non-dimensional sectional lift distribution over the rotor disk. 35
0.01
0.0075
0.005
0.0025
0
-0.0025
-0.005
-0.0075Non-Dimensional Blade Root Loads
-0.01 0 90 180 270 360 (deg)
Figure 2.3: Baseline blade root loads versus azimuth (lift integrated along span, non- dimensionalized by ρacR(ΩR)2).
36
0.0009
0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
Non-Dimensional Hub0.0001 Vertical Loads
0 4 8 12 16 20 Freq (/rev)
Figure 2.4: Frequency spectrum of baseline BVI-induced hub vertical loads.
The BVISPL footprint is then examined on an observer plane one rotor diameter below the rotor disk to locate the maximum far-field BVI noise generated in this plane
(Figure 2.5). The baseline case produces a maximum of 129.5 dB of noise. Though this is significantly larger than normally experienced by a rotor in standard operating conditions, a perfectly parallel interaction occurring over the entire outer half of the blade length with zero miss-distance represents an extremely severe BVI event (an idealization, and one of the worst scenarios conceivable). 37
BVISPL (dB) 130 126 = 180 122 100 118 114 110
50
=270 0 X(%radius)
-50
=0 -100 100 50 0 -50 -100 -150 Y(%radius)
Figure 2.5: Baseline BVISPL footprint one diameter below the rotor disk plane.
2.3 Vortex Strength
In this section, variation in maximum BVI noise and the BVI-induced hub vibratory loading is examined as a function of the interacting vortex strength, Γ. The strength of the interacting vortex is one of the easier parameters to control, as it is directly related to the lift on the blade upstream at the time the interacting vortical elements are 38 generated (this would be over specific azimuthal ranges in the second quadrant of the rotor disk for parallel interactions occurring in the first quadrant). Figure 2.6 shows the
2 variation of the normal force coefficient, CNM , and Figure 2.7 shows its azimuthal
2 derivative, dCNM /d, at 75% radius for various non-dimensional vortex strength values.
It can be deduced that a change in the vortex strength does not affect the azimuthal range over which the effects of the interaction extend, or its fundamental nature, but simply
linearly scales the values of loading. The variation in peak far-field BVI noise, as a
function of vortex strength, Γ, as well as the sensitivity of this noise variation, versus Γ is
shown in Figure 2.8. Note that the units of peak BVI noise and the sensitivity of that
noise to changes in the interaction parameter have the same units since the interaction
parameter is, itself, unitless. From the figure it is observed that the maximum BVI noise
reduces with decreasing values of Γ, as expected. It is also interesting to note that the
peak noise levels are highly sensitive to changes in Γ when the baseline value of Γ is
small, but are less sensitive to changes in Γ when the baseline value is larger. Thus, if the
interacting vortex is stronger, much larger reductions in vortex strength, Γ, are required to
produce comparable reductions in peak BVI noise. 39
0.1
0.05 2
M 0 N C 0.051 -0.05 0.077 0.101
-0.1 50 55 60 65 70 (deg) 2 Figure 2.6: CNM versus azimuth at 75% radius spanwise station for different vortex strengths.
9
6 )/d 2 3 M N d(C
0
-3 50 55 60 65 70 (deg)
2 Figure 2.7: d(CNM )/dψ versus azimuth at 75% radius spanwise station for different vortex strengths. 40
600 135
500 130
400 125
Max BVISPL 300
120 d(Max______BVISPL) d(/(cV )) tip 200 Sensitivity (dB) Max BVISPL (dB)
115 100
110 0 0 0.05 0.1 0.15 0.2
/(cVtip)
Figure 2.8: Max BVISPL and sensitivity versus vortex strength.
Figure 2.9 shows the BVI-induced vibratory hub loading for several different values of vortex strength. It is observed that variations in Γ change the magnitude of the vibratory loading, but have no effect on the distribution over the different harmonics, as would be expected since only the magnitude of the lift distribution was changed. It is also evident that the magnitude of the loading (at any harmonic) varies linearly with Γ, so 41 reducing the value of Γ by one half reduces the BVI-induced vibratory hub loading to one half of the baseline amplitude.
0.010 0.051 0.002 0.101 0.152 0.0018 0.202
0.0016
0.0014
0.0012
0.001
0.0008
0.0006
0.0004
Non-Dimensional Hub0.0002 Vertical Loads
0 48121620 Freq (/rev)
Figure 2.9: Frequency spectrum of BVI-induced hub vertical loads for various vortex strengths. 42
2.4 Core Radius
A small vortex core radius (tight core) implies higher peak velocities in the vortex, which produces more impulsive changes in blade loading and higher noise intensity when the vortex passes in close proximity to the blade. Variation in peak BVI noise level as a function of vortex core radius is presented in Figure 2.10. It is observed that the noise decreases almost linearly with increasing core radius, provided the miss- distance between the blade and vortex (vortex above blade) is small. For larger values of miss-distance (typically larger than the vortex core radius), the peak BVI noise levels are insensitive to vortex core size. 43
140 Miss-Distance 135 0.0 chords 0.5 chords 1.0 chords 130
125
120
115 Max BVISPL (dB)
110
105
100 0.10.20.30.4 Vortex Core Radius (chords)
Figure 2.10: Max BVISPL versus vortex core radius.
Effects of variation in vortex core radius on the BVI-induced hub vibratory loading levels are shown in Figure 2.11 for a zero blade-vortex miss-distance. It is observed that for a very small core size, the extremely impulsive nature of the interaction produces hub vibratory loads that have very large higher harmonic content. In fact, for a core radius of 0.05 chords, the amplitudes of the harmonics increase from 4/rev up through 20/rev. In general, increasing core radii result in reductions in vibratory load levels, with the reductions being more significant in the higher harmonics. 44
0.05 chords 0.0012 0.10 chords Core Radius 0.20 chords 0.0011 0.30 chords 0.40 chords 0.001
0.0009
0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
Non-Dimensional Hub0.0001 Vertical Loads
0 4 8 12 16 20 Freq (/rev)
Figure 2.11: Frequency spectrum of BVI-induced hub vertical loads for various core radii.
2.5 Miss-Distance
One of the most important blade-vortex interaction parameters is the separation between the vortex and the rotor blade in the blade-rotor-shaft plane. This is referred to as the “miss-distance” (see for example, Reference 7). Figure 2.12 shows the variation in 45 peak far-field BVI noise, as well as the sensitivity of this variation, as a function of the miss-distance. It is observed that when the miss-distance increases from zero (vortex passing right through the blade) to a value of half a chord, a 14.4 dB reduction in BVI noise is obtained. From the sensitivity curve in Figure 2.12 it is observed that except at very small values of miss-distance (less than the vortex core radius), the sensitivity of
BVI noise to changes in miss-distance reduces for increasing miss-distances, with low sensitivity beyond a miss-distance of one chord. On the other hand, for miss-distances smaller than the vortex core radius, the sensitivity (change in BVISPL for small perturbation in miss-distance) is again small. This implies that as long as the vortex core is intersecting the blade, changes in the separation between the vortex center and the blade have only a small influence on BVI noise. In Figure 2.13 the variations in BVI noise for small values of miss-distance are examined in greater detail. It is again evident that when the vortex core sizes are larger (see the 0.4 chords curve) and the initial miss- distances are small, modest changes in the miss-distance have little impact on BVI noise levels. However, for smaller vortex core sizes (see the 0.2 chords curve), the blade quickly comes out of the vortex core for similar modest increases in miss-distance, producing reductions in BVI noise. 46
130 40
125 35 Max BVISPL 120
______d(Max BVISPL) 30 115 d(Miss-Distance)
25 110
105 20 Sensitivity (dB) Max100 BVISPL (dB) 15
95 10 90
00.511.52 Miss-Distance (chords)
Figure 2.12: Max BVISPL and sensitivity versus miss-distance.
47
130
129
128 Core Radius 0.2 chords 127 0.3 chords 126 0.4 chords
125
124
123 Max BVISPL (dB) 122
121
120
119 0 0.05 0.1 0.15 0.2 Miss-Distance (chords)
Figure 2.13: Max BVISPL versus miss-distance (small values of miss-distance).
2 Figure 2.14 shows the variation of the normal force coefficient, CNM , and
2 Figure 2.15 shows its azimuthal derivative, dCNM /d, at 75% radius for various blade-
vortex miss-distance values. As the miss-distance decreases, the peak loading levels
increase while the azimuthal interval between the negative and positive peaks appear to
be unaffected, resulting in a more impulsive loading. Figure 2.16 shows the impact of
variations in miss-distance on the BVI-induced vibratory hub loads. It is observed that for small miss-distances (producing more impulsive sectional loads and blade root loads) a 48 relatively large portion of the hub vibration energy is in the higher harmonics. As the miss-distance increases (and the loads become less impulsive, see Figure 2.14), the total hub vibration energy decreases (in all the harmonics), with the greatest reductions seen in the higher harmonics.
0.1
0.05 2
M 0 N
C Miss-Distance 0.0 chords -0.05 0.1 chords 0.2 chords
-0.1 50 55 60 65 70 (deg)
2 Figure 2.14: CNM versus azimuth at 75% radius spanwise station for different blade- vortex miss-distances.
49
9
Miss-Distance 6 0.0 chords
0.1 chords )/d
2 0.2 chords 3 M N d(C
0
-3 50 55 60 65 70 (deg)
2 Figure 2.15: d(CNM )/dψ versus azimuth at 75% radius spanwise station for different blade-vortex miss-distances.
50
0.0 chords 0.001 0.5 chords Miss-Distance 1.0 chords 0.0009 1.5 chords 2.0 chords 0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
Non-Dimensional Hub0.0001 Vertical Loads
0 4 8 12 16 20 Freq (/rev)
Figure 2.16: Frequency spectrum of BVI-induced hub vertical loads for various blade- vortex miss-distances.
2.6 Blade-Shaft Plane Interaction Angle
For the results presented in the previous sections, the interacting vortex was contained in the rotor disk plane (zero miss-distance) or in a plane parallel to the disk plane (non-zero, but constant miss-distance over the blade-vortex interaction length). In 51 this section, the effect of the vortex interacting with the blade at some non-zero inclination angle in the blade-shaft plane is examined, as depicted in Figure 2.17 (vortex length is R/2 and intersects the blade at a radial location of 3R/4). It should be noted that the term “miss-distance”, as used in the literature, has little meaning when the vortex is inclined in the blade-shaft plane relative to the blade, since the vertical distance between a point on the vortex and the corresponding radial station on the blade will clearly vary with radial position. Figure 2.18 shows that significant reductions in BVI noise are obtained as the vortex tilts relative to the blade in the blade-shaft plane. For a relative inclination of as little as 10˚, a significant reduction in peak noise of 7.1 dB is realized.
The inclination of the vortex in the blade-shaft plane reduces the BVI noise, as it effectively reduces the interaction length, with regions of the vortex further away from the blade contributing little to the impulsive loads. Also of note is that noise reductions are independent of the sign of the angle in the blade-shaft plane. 52
Ω Vortex 3R/4 Rotor Blade
Inclination Angle in Blade-Shaft Plane R/2
Rotor Shaft
Figure 2.17: Vortex interacting with the blade at an angle in the blade-shaft plane.
53
130
125
120
Max BVISPL115 (dB)
110
-45 -30 -15 0 15 30 45 Shaft Plane Interaction Angle (deg)
Figure 2.18: Max BVISPL versus blade-shaft plane interaction angle.
Figure 2.19 shows the BVI-induced vibratory hub loading for various blade- vortex interaction angles in the blade-shaft plane. For a zero relative inclination angle between the blade and the vortex in the blade shaft plane (a perfectly parallel interaction), the interaction is very impulsive. Thus, the vibratory loading levels are the highest and there is a large amount of energy in the higher harmonics. For higher blade-shaft plane interaction angles (increasing vortex tilt relative to the blade in the blade-shaft plane) the 54
“effective interaction length” is reduced and the interaction is less impulsive. Thus, the
BVI-induced total hub vibration energy (in all harmonics) decreases, with the largest reductions observed in the higher harmonics.
-45 0.001 -25 Blade-Shaft Plane Interaction Angle 0 0.0009 25 45 0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
Non-Dimensional Hub0.0001 Vertical Loads
0 4 8 12 16 20 Freq (/rev)
Figure 2.19: Frequency spectrum of BVI-induced hub vertical loads for various blade- shaft plane interaction angles. 55
2.7 Rotor Disk Plane Interaction Angle
For the results considered so far, the blade-vortex interaction was always perfectly parallel in the rotor disk plane at = 60˚. In this section, the effect of varying the interaction angle between the vortex and the blade in this plane is examined. As seen in
Fig. 2.20, the vortex of length R/2 (from mid-span to the blade tip) is rotated in the disk
plane about the 3R/4 point to produce less parallel (more oblique) interactions (the miss-
distance is set at zero). This causes the blade to interact with the vortex over a longer
2 azimuthal duration. Figure 2.21 shows CNM , and Figure 2.22 shows its derivative versus
at three different spanwise stations. For a parallel interaction the peaks at the different
spanwise stations occur at the same azimuthal location. For an oblique interaction,
however, the peaks at the three spanwise stations occur at different azimuthal locations.
As oblique interactions do not focus noise as effectively as parallel interactions [10, 11],
the BVI noise is expected to decrease as well. Figure 2.23 shows the reductions in noise
obtained for increasingly oblique interactions. Unlike the inclination of the vortex in the
blade-shaft plane, positive and negative angles in the disk plane produce different effects.
This is not entirely unexpected since BVI noise is known to be dependent on the “type of
interaction” [10, 11] and positive angles in the disk plane produces a different interaction
(moving from inboard to outboard locations along the blade span) from that seen with
negative angles (interaction moves inboard from the blade tip). From Figure 2.23 it is
observed that for small angles between the vortex and blade (less than 10°) relatively
modest noise reductions are obtained. However, if the interaction becomes more oblique 56 so that the angles between the vortex and blade are in the 20-25° range, then significant
reductions in peak BVI noise levels (of over 10 dB) can be obtained.
Ω Vortex 3R/4 Rotor Blade Disk Plane Rotor R/2 Hub Interaction Angle
Figure 2.20: Vortex interacting with blade at an angle in the rotor disk plane (oblique interaction).
57
0.1
0.05 2
M 0 N Parallel (0)Interaction C 60% radius 75% radius -0.05 90% radius Oblique (30)Interaction 60% radius -0.1 75% radius 90% radius
40 50 60 70 80 (deg)
2 Figure 2.21: CNM versus azimuth for different rotor disk plane interaction angles (parallel and oblique interaction).
10 Parallel (0) Interaction 60% radius 75% radius
90% radius
)/d Oblique (30)Interaction 2 5 60% radius M 75% radius N 90% radius d(C
0
40 50 60 70 80 (deg)
2 Figure 2.22: d(CNM )/dψ versus azimuth for different rotor disk plane interaction angles (parallel and oblique interaction). 58
130
128
126
124
122
120
Max118 BVISPL (dB)
116
114
-45 -30 -15 0 15 30 45 Disk Plane Interaction Angle (deg)
Figure 2.23: Max BVISPL versus rotor disk plane interaction angle.
An increase in the angle between the blade and the interacting vortex in the disk plane could be achieved by sweeping the outboard regions of the blade (either forward or backward). Introduction of a backward blade sweep would produce a positive angle between blade and vortex (for a vortex that would have struck the unswept blade in a parallel manner). Conversely, introduction of a forward blade sweep would produce a 59 negative angle. Figure 2.23 suggests that if only small sweep angles of 10° or less were being considered, a forward sweep (resulting in a negative angle between blade and vortex) would be more beneficial than backward sweep for BVI noise reduction.
However, if larger sweep angles (15° – 25°) were permissible, a rearward sweep
(resulting in a positive angle between blade and vortex) would produce greater reductions in BVI noise. It should be noted, though, that other factors such as performance in high- speed flight and aeroelastic stability play an important role in determining the magnitude and direction of the tip sweep.
Figure 2.24 shows the BVI-induced vibratory hub load levels for different interaction angles in the rotor disk plane. Clearly, perfectly parallel interactions occur over very short time durations, are highly impulsive, and produce vibratory hub loadings that are of the largest magnitudes and have a significant amount of energy in the higher harmonics. As the interaction becomes more oblique (occurs over a longer azimuthal interval and is less impulsive), the BVI-induced vibratory hub load levels reduce in magnitude, and the higher harmonic components decrease in prominence. 60
-45 0.001 -25 Rotor Disk Plane 0 Interaction Angle 0.0009 25 45 0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
Non-Dimensional Hub0.0001 Vertical Loads
0 4 8 12 16 20 Freq (/rev)
Figure 2.24: Frequency spectrum of BVI-induced hub vertical loads for various rotor disk plane interaction angles.
2.8 Spanwise Location of the Interaction
The influence of the radial location of the BVI event on the noise and vibration levels is examined next. For the simulations in this section, angles of the interaction in the blade-shaft and rotor disk planes are again zero, as is the blade-vortex miss-distance. 61
The vortex of length R/2 was moved in increments along the blade span, from the rotor hub to the blade tip (so that the vortex center traversed from a radial location of R/4 to
3R/4). In Figure 2.22 it was already seen that for a parallel interaction in the rotor disk plane, the blade sectional loading at outboard stations is more impulsive in nature, relative to the loading at inboard spanwise stations. Correspondingly, it is observed in
Figure 2.25 that the peak BVI noise increases as the center of the interacting vortex moves outboard. It is also seen from the figure that the sensitivity of BVI noise to changes in interaction location is greater for inboard rather than outboard interactions.
Thus, if a BVI event occurring in the outboard portion of the blade moves slightly inboard, the noise reductions are smaller than those obtained when a BVI event occurring in the inboard portion moved further inboard by a comparable amount. 62
55
50 125 45
40 Max BVISPL d(Max BVISPL) 120 ______35 d(Radial Position) Sensitivity (dB) Max BVISPL (dB) 30
25 115 20
30 40 50 60 70 Radial Postion of Vortex Center (% radius)
Figure 2.25: Max BVISPL and sensitivity versus spanwise location of center of interacting vortex.
Figure 2.26 shows the impulsive blade root loads due to the blade-vortex interaction occurring at different locations along the span. It is observed that for the inboard interactions the magnitude of the impulsive load is lower, this due to the lower tangential velocities and the longer duration of interaction. Conversely, the outboard interactions produce very high intensity impulsive changes in blade root loads, and these result in significant higher harmonic content in the BVI-induced vibratory hub loading 63
(see Fig. 2.27). For mid-span and inboard interactions, which produce less impulsive blade root loads, a larger percentage of the vibration energy is in the lower harmonics. In fact, mid-span interactions actually produce a larger 4/rev component of BVI-induced vibratory hub loading than do the outboard interactions.
0.01
0.0075
0.005
0.0025
0
-0.0025 Radial Position of Vortex Center -0.005 25% radius 50% radius 75% radius
-0.0075Non-Dimensional Blade Root Loads
-0.01 30 45 60 75 90 (deg)
Figure 2.26: Blade root loads versus azimuth for various spanwise locations of center of interacting vortex.
64
0.25 radius 0.38 radius 0.001 Location of Center of Interacting Vortex 0.50 radius 0.63 radius 0.0009 0.75 radius
0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
Non-Dimensional Hub0.0001 Vertical Loads
0 4 8 12 16 20 Freq (/rev)
Figure 2.27: Frequency spectrum of BVI-induced hub vertical loads for various spanwise locations of center of interacting vortex.
2.9 Spanwise Length of the Interaction
The spanwise length of the interaction is intrinsically tied to the spanwise location of the interaction. When the vortex length is reduced, which portion of the blade is no longer affected by the vortex is as important as how much less of the blade span is 65 affected, in terms of changes to the BVI noise and vibration results. Figure 2.28 illustrates the change in maximum BVI noise levels with simultaneous variations in these two parameters. If the vortex center is held fixed and its length progressively decreased
(moving down lines parallel to the y-axis), which is tantamount to reducing the blade area affected by the interacting vortex at the fringes of the vortex, BVI noise levels are seen to reduce. The reductions in noise are qualitatively similar to those seen in Figure 2.18, for increasing inclinations of the vortex in the blade-shaft plane. This is to be expected since increasing the inclination in the blade-shaft plane has the effect of reducing the length of the blade affected by the vortex (spanwise length of the interaction), with the fringes of the vortex having reduced contributions. It is seen that the BVI induced-noise is more sensitive to changes in the spanwise location of the interaction for longer interactions, and less sensitive for shorter interactions. Likewise, the noise is more sensitive to changes in the interaction length at more outboard spanwise stations, and less sensitive at inboard stations. The largest reductions in noise are obtained when the vortex length is reduced and the vortex center is simultaneously moved inboard. This has the net effect of reducing the interaction of the blade with the vortex in the outboard locations, which produce more impulsive sectional loads (Figure 2.22). 66
60 Nominal Interaction Parameters
50 Outer End of Vortex Reaches Blade Tip 40
BVISPL (dB) 130 127.5 30 125 122.5 120 117.5 115 20 112.5 110 107.5 Interacting Vortex Length (% radius) 105
10 40 50 60 70 80 Radial Position of Vortex Center (% radius)
Figure 2.28: Max BVISPL versus spanwise location of center of interacting vortex and length of interaction.
Figures 2.29 and 2.30 show the variation in BVI-induced 4/rev and 8/rev vertical hub loads, respectively, for variations in the spanwise length of interacting vortex, as well as the spanwise location of the center of the vortex. The vertical hub loads are far more sensitive to changes in length of the interaction than to changes in the spanwise location.
Though a reduction in the length of the interaction reduces both 4/rev and 8/rev vibratory loads, it is seen that an inboard motion of the vortex actually increases the 4/rev 67 component, and can either increase or decrease the 8/rev component. The large reductions in loads observed for reductions in interacting vortex length echo the reductions seen with increase in blade-shaft plane interaction angles (Figure 2.19).
60 Nominal Interaction Parameters
50 Outer End of Vortex Reaches Blade Tip 40
Non-Dimensional 4/rev Vertical Hub Load 30 0.0012 0.001 0.0008 0.0006 0.0004 20 0.0002 0 Interacting Vortex Length (% radius)
10 40 50 60 70 80 Radial Position of Vortex Center (% radius)
Figure 2.29: Non-dimensional 4/rev BVI-induced vertical hub loads versus spanwise location of center of interacting vortex and length of interaction.
68
60 Nominal Interaction Parameters
50 Outer End of Vortex Reaches Blade Tip 40
Non-Dimensional 8/rev Vertical Hub Load 30 0.0012 0.001 0.0008 0.0006 0.0004 20 0.0002 0 Interacting Vortex Length (% radius)
10 40 50 60 70 80 Radial Position of Vortex Center (% radius)
Figure 2.30: Non-dimensional 8/rev BVI-induced vertical hub loads versus spanwise location of center of interacting vortex and length of interaction.
2.10 Collective Pitch
All the previous results examining the influence of various BVI parameters on
BVI noise and BVI-induced vibratory hub loads were obtained for a non-lifting rotor. In the present section the influence of collective pitch (or rotor thrust) variation is examined. 69
A perfectly parallel interaction extending from R/2 to R, with zero miss-distance, is once again considered. Using hybrid blade-element/momentum theory, an equivalent inflow through the disk is calculated for different values of collective pitch (inflow is assumed to be uniform around the azimuth, and a self-generated wake comprising trailing vortices and tip vortices are not considered). Thus, the lift produced on any blade section (at a given radial station) is due to the resultant angle of attack from the non-zero collective pitch, minus the total inflow angle (comprising a downwash due to the lifting rotor and the impulsive change in inflow due to the interacting vortex).
Figure 2.31 shows azimuthal variation in blade lift at 0.76R for different values of blade collective pitch setting. It is seen that the total lift increases with blade pitch setting, but the impulsive variation, which produces the BVI noise, is unchanged. The solid line on the graph corresponds to a linear aerodynamic model and the dashed line corresponds to a model that includes quasi-static stall effects. For the linear aerodynamic model, BVI noise is insensitive to the blade lift (or blade pitch angle). It is observed that while the total lift increases with increasing collective pitch values, for higher pitch settings the passage of the vortex caused the blade section angle of attack to exceed stall levels. When a static stall (table lookup) model is used, the impulsive change in lift is reduced, and this actually produces a reduction in BVI noise. 70
0.35 Stall Threshold 0.3 =7 0.25
0.2 =3.5 2 0.15 M N
C 0.1
0.05 =0 0 Linear C -0.05 Table Lookup (Static Stall) -0.1 20 40 60 80 100 (deg)
Figure 2.31: Blade sectional lift at 75% radius for various collective pitch settings, θ.
These results appear to contradict previous reports implying that increase in blade lift at the location of the BVI event increases noise, and vice-versa [9, 12]. However, it should be noted that for the simulations in Figure 2.31, the vortex was held fixed (as the blade traverses around the azimuth and interacts with it in the vicinity of = 60°). When
the interacting vortex is initially placed at the = 60° azimuthal location, but is then allowed to convect freely as the lifting blade passes close to it, the results obtained are 71 different. Figure 2.32 shows the motion of the free vortex due to the influence of the bound vorticity of the lifting blade. The motion is plotted in blade-fixed coordinates, and the interacting free vortex is shown moving over a range extending from 60° before the interaction to 60° after. It is seen that the bound vorticity of the blade convects the interacting vortex upward before, and downward after, the interaction. When the free vortex starts below the blade this effectively decreases the miss-distance. The larger the bound vorticity (the greater the blade pitch), the more pronounced is this effect. However, when the free vortex initially starts in the plane of the blade or above, the miss-distance at the time of interaction, due to the bound vorticity on the blade, is increased. The results in
Figure 2.32 suggest that BVI noise may be reduced by locally changing the lift on the blade at the time of interaction through reducing the blade pitch, if the vortex is initially sailing below the blade, or increasing the blade pitch, if the vortex is initially in the disk plane or sailing above the blade. In either case, the resulting increase in the miss-distance would be expected to be beneficial from a BVI noise reduction standpoint. This is clearly seen in Fig. 2.33, which shows the variation in peak BVI noise with variation in blade pitch for three different initial positions of the interacting free vortex relative to the blade. 72
Interacting Vortex 0.6 Above Blade
0.4 Trajectory of Vortex for of 4 0.2 Interacting Vortex 8 in Plane of Blade
0 Bound Vorticity of Blade -0.2
Vertical Distance (chords) Interacting Vortex -0.4 Below Blade
-60 -30 0 30 60 Relative Azimuthal Location (deg)
Figure 2.32: Interacting vortex trajectories for different initial positions relative to blade, and different blade pitch settings.
73
130
125 Initial Position of Free Vortex 0.5 Chords Above Blade In the Disk Plane 0.5 Chords Below Blade 120 Max BVISPL (dB) 115
110 02468 Rotor Collective Pitch (deg)
Figure 2.33: Max BVISPL versus rotor collective pitch for different initial interacting vortex positions.
However, when a self-generated free-wake was considered and local changes in blade pitch around the time of interaction were examined to reduce BVI noise, neither positive nor negative changes in pitch (of varying amplitudes) had much effect on the
BVI noise (no results provided). This is because in addition to the bound vorticity
(changing whose strength could in principle move the interacting free vortex as explained in the previous paragraph), the trailed vorticity and the tip vortex of the interacting blade, 74 and the rest of the vorticity in the entire rotor wake also influence the movement of the interacting free vortex. For example, an increase in the strength of the bound vorticity
(through local change in the blade pitch) would also produce a corresponding increase in the strength of the tip vortex. While the stronger bound vorticity would tend to convect the free vortex upward, the stronger tip vortex would tend to convect the free vortex downward (inside of the rotor disk), thus negating in part the effect of the change in bound vorticity. It should be noted that Reference 13 has reported that changes in blade flap setting (for the purpose of changing blade lift) around the time of blade-vortex interaction had very little effect on changing the blade-vortex miss-distance (the mechanism by which noise reductions would actually have been obtained). On the other hand, no evidence has been presented in the literature that reasonable changes in blade pitch (or lift) around the time of interaction can actually reduce BVI noise, though this idea is frequently suggested.
2.11 Comparison of Effects of Various Parameters in Reducing BVI-induced Noise and Hub Vibratory Loading
This section examines the comparative reductions in BVI-induced noise and vibratory loading achievable through “moderate” variations in the magnitude of the different interaction parameters previously considered. The purpose is to compile information that would assist an engineer in deciding which of the parameters it would be most advantageous to attempt to manipulate (resulting in largest reductions in BVI noise and BVI-induced hub vibrations for relatively modest changes in the parameter). 75
2.11.1 Weak Interaction Parameters
A 50% increase in core radius from a baseline value of 0.2 chords to 0.3 chords produced a reduction in peak BVI noise of about 5 dB, but little reduction in 4/rev BVI- induced hub vibratory loads. If the spanwise location of the BVI event moved inboard from the tip by about 10% radius (vortex center moving from 75% to 65%) the reduction in BVI noise is less than 3 dB, and the BVI-induced 4/rev vibratory hub loads actually increased. A much larger inboard movement of about 25% radius (center moving from
75% to 50%) would be required to produce a substantial BVI noise reduction, of the order of 7 dB; and such a large spanwise movement of the BVI event may be difficult to achieve. Causing the BVI event to move to a more inboard location (to reduce noise) is unable to significantly reduce BVI-induced hub vibratory loading, and in many cases causes it to increase. A 10% change in the length of the interaction (from 50% radius to
45% radius) only reduces the BVI noise by 0.3dB and the 4/rev hub load by about 10%.
A full 50% reduction in length results in only about a 5 dB decrease in noise, though it reduces 4/rev vibration by almost 50%. The above interaction parameters (vortex core radius, spanwise location of the event, and spanwise length of the interaction) appear to have less effect in reducing BVI noise (for moderate changes in the magnitude of the parameter), compared to some of the other interaction parameters such as blade-vortex miss-distance, vortex strength, and the interaction angles in the blade-shaft and the rotor disk planes. Further, they may not even be candidate parameters for change if simultaneous reduction of BVI noise and BVI-induced vibration is desired. 76
2.11.2 Strong Interaction Parameters
Compared to an interaction wherein the vortex passes right through the blade, a miss-distance value of 0.5 chords reduces the BVI noise by a very significant 15 dB and the 4/rev hub vibratory loading by a more modest 15%. Reducing the interacting vortex strength by a factor of one half (from = 2 m2/s to 1 m2/s) decreased the BVI noise by 7
dB and reduced the hub vibrations by 50%. When the angle between the blade and the
interacting vortex in the blade-shaft plane increases from zero to ±10˚, a 7 dB reduction in noise is observed. When the blade-shaft plane inclination angle increases to ±20˚, a 12 dB noise reduction and a reduction in 4/rev BVI-induced vibratory hub loading of around
30% can be achieved. When the angle between the blade and the interacting vortex in the disk plane increases from zero (perfectly parallel interaction) to about ±10˚, reductions in
BVI noise in the range of only 1 dB to less than 3 dB (depending on the sign of the angle) are observed. However if the disk-plane angle increases to about ±20 deg, BVI noise reductions in the range of 6-10 dB (depending on sign of angle) and very modest reductions in 4/rev BVI-induced vibratory hub loading (of the order of 10%), are achievable. Of these four parameters (miss-distance, vortex strength, angle in blade-shaft plane, and angle in disk plane) it appears that increasing the miss-distance is the most effective in reducing noise whereas decreasing the vortex strength is most effective in reducing BVI induced vibratory loading. Reductions in vortex strength may be the best option if simultaneous reductions in BVI noise and vibratory loading are important.
Vortex strength may also be the easiest to control by using active methods such as blade 77 pitch variation or flap deflection upstream, at the time of generation of the vortical elements that convect downstream and result in parallel interactions.
Although it is widely accepted that more oblique, rather than parallel, interactions produce less BVI noise, the results in this chapter indicate that a small increase in the angle between the blade and the vortex in the blade-shaft plane is much more effective in reducing the BVI noise than a comparable increase in the disk plane angle. This point is reinforced in Figure 2.34, which shows contours of constant peak noise levels versus variations in blade-shaft plane as well as rotor disk plane interaction angles. Clearly, changes in the blade-shaft plane angle produce greater reductions in BVI noise for modest changes in the value of the parameter, than do changes in the rotor disk plane angle. Increasing the blade-shaft plane inclination angle has the added benefit of reducing vibration. From Figure 2.35, reductions in BVI-induced 4/rev vertical loads are far greater for inclinations in the blade-shaft plane than inclinations in the rotor disk plane.
This would suggest that anhedral or dihedral (which can potentially affect the interaction angle in the blade-shaft plane) could be more effective in alleviating BVI than blade sweep (which seeks to affect the interaction angle in the rotor disk plane). 78
30
20
BVISPL (dB) 130 127.5 10 125 122.5 120 117.5 0 115 112.5 110 107.5 -10 105
-20 Blade-Shaft Plane Interaction Angle (deg) -30 -30 -20 -10 0 10 20 30 Rotor Disk Plane Interaction Angle (deg)
Figure 2.34: Max BVISPL versus interaction angles in the rotor disk and blade-shaft planes.
79
30
20 Non-Dimensional 4/rev Vertical Hub Load 0.00083 0.00079 10 0.00075 0.00071 0.00067 0.00063 0 0.00059 0.00055 0.00051 0.00047 -10 0.00043
-20 Blade-Shaft Plane Interaction Angle (deg) -30 -30 -20 -10 0 10 20 30 Rotor Disk Plane Interaction Angle (deg)
Figure 2.35: Non-dimensional BVI-induced 4/rev hub vertical loads versus interaction angles in the rotor disk and blade-shaft planes.
The above comparisons point towards an increase in miss-distance and interaction angle in the blade-shaft plane, and a decrease in the interacting vortex strength as the most influential of the parameters considered for reducing BVI-induced noise and vibration. 80
2.12 Minimum Vibration Case in the HART Test
Researchers reporting the results of the HART test [9], and several others subsequently examining Higher Harmonic blade pitch Control (HHC) for BVI alleviation, have indicated that the HHC schedule that produced minimum vibration decreased the miss-distance, relative to the baseline; and an adequate explanation has not been provided to date for this observation. However, the results in the present study show that smaller values of miss-distance produce larger, not smaller, BVI-induced hub vibratory loading (see Figure 2.16). Resolving this question requires taking a closer look at the HART data and in particular the geometry of the interacting vortices, relative to the blade, for both the baseline as well as the minimum vibration cases (see Figures 2.36 and 2.37). It is seen that although the vortices pass through the blade for the minimum vibration case, they have a very large inclination in the blade-shaft plane (compared to the baseline case). From Figure 2.18 it was observed that such an increase in inclination would indeed produce significant reduction in BVI-induced hub vibratory loading. Thus, it is hypothesized that for the HART minimum vibration case it was a larger blade-shaft plane angle that resulted in lower vibrations. Furthermore, it should be reiterated that it is somewhat misleading to use the concept of miss-distance in reference to vortices that are significantly inclined relative to the blade in the blade-shaft plane, since every point will have a different miss-distance from the blade. The miss-distance concept is useful only when vortices are generally contained in planes parallel to the rotor disk plane. 81
3
2 Vortex, Blade 4 Vortex, Blade 3 1
0
-1 Distance in Blade-Shaft Plane (Z/c) Plane Blade-Shaft in Distance
-2 0 0.2 0.4 0.6 0.8 1 Blade Span (r/R)
Figure 2.36: HART test [9] blade-vortex interaction geometry in blade-shaft plane for baseline.
82
3
2
1
0 Upper Vor., Bl. 4
-1 Upper Vor., Bl. 3 Lower Vor., Bl. 4
Distance in Blade-Shaft Plane (Z/c) Plane Blade-Shaft in Distance Lower Vor., Bl. 3 -2 0 0.2 0.4 0.6 0.8 1 Blade Span (r/R)
Figure 2.37: HART test [9] blade-vortex interaction geometry in blade-shaft plane for minimum vibration HHC input.
2.13 Conclusions
This chapter comprehensively examines the influence of various blade-vortex interaction parameters on BVI noise and BVI-induced vibratory hub loading. An externally imposed vortex is allowed to interact with a rotor and key interaction parameters such as vortex strength, core-radius, the blade-vortex miss-distance, the spanwise location and extent of the interaction, the angles between the vortex and the blade in the blade-shaft plane and the disk plane, and finally the blade lift at the time of interaction are varied. The corresponding reductions achieved in BVI noise and BVI- 83 induced vibratory hub loading are evaluated. From the results presented in this chapter, the following conclusions can be drawn:
1. Reduction in vortex strength, Γ, reduces the peak BVI noise. Larger noise
reductions (for a given reduction in Γ) are obtained when the baseline vortex
strength is moderate to low. Reduction in BVI-induced vibratory hub loading is
directly proportional to reduction in vortex strength.
2. Increasing the core radius reduces BVI noise provided the miss-distance is small
(less than the core radius, so the blade is intersecting the vortex core). For larger
miss-distances (that place the blade outside the vortex core) change in core radius
has little effect on the peak BVI noise. For moderate baseline values of vortex
core radius, reduction in core radius has little effect in reducing BVI-induced
vibratory hub loading.
3. As long as the vortex core is not passing through the blade, increase in miss-
distance has a dramatic effect on reducing BVI noise. The largest reductions in
BVI noise are observed when the initial miss-distance is small, whereas the
reductions for comparable increases in miss-distance are smaller than when the
initial miss-distance is larger (beyond one chord length). Increased miss-distance
reduces both the overall amplitude of the BVI-induced vibratory hub loading as
well as the higher harmonic content.
4. Modest reductions in BVI noise can be obtained as the event moves to a more
inboard location and decreases in length. A more inboard interaction can actually 84
increase the amplitude of the 4/rev BVI-induced vibratory hub loading slightly,
although the percentage of vibration energy in the higher harmonics reduces.
5. Even a modest inclination (10-20°) of the vortex relative to the blade, in the
blade-shaft plane, can produce large reductions in BVI noise, as well as BVI-
induced vibratory hub loading. Inclination in the blade-shaft plane also reduces
the percentage of vibration energy contained in the higher harmonics. This
suggests that introducing modest amounts of anhedral or dihedral at the blade tips
may be attractive.
6. Inclination of the vortex in the disk plane, from a more parallel to a more oblique
orientation, also reduced the BVI noise. However, only very modest noise
reductions (~ 2-3 dB) are obtained for small inclination angles (of up to 10°).
Even for larger inclination angles, the noise reductions are smaller than those due
to comparable inclinations between the blade and vortex in the blade-shaft plane.
This suggests that forward or backward sweep of the outboard regions may be
less effective than blade anhedral/dihedral. The noise reductions are dependent on
the sign of the inclination angle. Reductions in BVI-induced vibratory hub
loading are also smaller than those due to inclination in the blade-shaft plane.
7. When the blade pitch is changed but the interacting vortex is held fixed relative to
the blade, there is only a change in the steady component of lift, but no change in
the impulsive lift. However, when the interacting vortex is free to move, the
bound vorticity of the blade causes it to convect upward. By increasing the blade
pitch, the magnitude of the bound vorticity increases, so a vortex that is passing 85
below the blade gets pulled toward it by the bound vorticity and a vortex in the
disk plane or above the blade gets pushed further upward, and away from the
blade. In practice, however, it is difficult to achieve noise reductions by changing
the blade lift around the time of interaction as the effect of the change in bound
vorticty is negated by other influences such as the change in tip vorticity of the
blade.
8. The results in the present chapter suggest that the notion that the miss-distance
reduced for the minimum vibration case in the HART test is misleading. The
angle between the vortex and blade in the blade-shaft plane was significantly
increased for the minimum vibration case, and the results in this chapter suggest
that such an increase could indeed be expected to produce a reduction in BVI-
induced hub vibration levels. 86
References
1. Martin, R. M., Elliott, J. W., and Hoad, D. R., “Experimental and Analytical Predictions of Rotor Blade Vortex Interaction,” Journal of the American Helicopter Society, Vol. 31, (4), October 1986.
2. Gallman, J. M., “Parametric Computational Study of Isolated Blade-Vortex Interaction Noise,” AIAA Journal, Vol. 32, (2), Feb. 1994, pp. 232-238.
3. Kitaplioglu, C., Caradonna, F. X., and Burley, C. L., “Parallel Blade-Vortex Interactions: An Experimental Study and Comparison with Computations,” Journal of the American Helicopter Society, Vol. 42, No. 3, July 1997, pp. 272- 281.
4. Vatistas, G. H., Kozel, V., and Mih, W. C., “A Simpler Model for Concentrated Vortices,” Experiments in Fluids, Vol. 11, 1991.
5. Bagai, A., Leishman, J. G., “Flow Visualization of Compressible Vortex Structures using Density Gradient Techniques,” Experiments in Fluids, Vol. 15, 1993.
6. Brentner, K. S., “Prediction of Helicopter Rotor Discrete Frequency Noise - A Computer Program Incorporating realistic Blade Motions and Advanced Acoustic Formulation,” NASA TM 87721, 1986.
7. Gandhi, F., and Tauszig, L., “A Critical Evaluation of Various Approaches for the Numerical Detection of Helicopter Blade-Vortex Interactions,” Journal of the American Helicopter Society, Vol. 45, (3), July 2000.
8. Brentner, K. S., Burley, C. R., Marcolini, M., A., “Sensitivity of Acoustic Predictions to Variation of Input Parameters,” Journal of the American Helicopter Society, Vol. 39, (3), July 1994.
9. Splettstoesser, W. R., Kube, R., Wagner, W., Seelhorst, U., Boutier, A., Micheli, F., Mercker, E., and Pengel, K., “Key Results from a Higher Harmonic Control Aeroacoustic Rotor Test (HART),” Journal of the American Helicopter Society, Vol. 42, (1), January 1997.
10. Schmitz, F., and Sim, B. W., “Radiation and Directionality Characteristics of Advancing Side Blade-Vortex Interaction (BVI) Noise,” 6th AIAA/ CEAS Aeroacoustics Conference, Lahaina, Hawaii, June 12-14, 2000.
11. Gandhi, F., and Tauszig, L., “Influence of Individual Interactions on Helicopter Blade-Vortex Interaction Noise,” Journal of the American Helicopter Society, October 2003. 87 12. Hardin, J. C., and Lamkin, S. L., “Concepts for Reduction of Blade/Vortex Interaction Noise,” Journal of Aircraft, Vol. 24, (2), February 1987.
13. Charles, B. D., Tadghighi, H., and Hassan, A. A., “Effects of a Trailing Edge Flap on the Aerodynamics and Acoustics of Rotor Blade-Vortex Interactions,” Proceedings of the 14th DGLR/AIAA Aeroacoustics Conference, Aachen, Germany, May 1992. 88
Chapter 3
Localized Individual Blade Root Pitch Control for BVI Noise Reduction
Exploiting both understanding of the BVI noise-generation mechanisms and the tremendous flexibility associated with IBC, this section focuses on localized reductions in blade root pitch, over specific azimuthal ranges specially designed to reduce the strength of those tip vortex elements that, after convecting downstream, produce the noise- generating parallel blade-vortex interactions. Unlike the harmonic IBC root pitch inputs previously considered, the present inputs are intuitive and physically motivated. Whereas the BVI noise reduction mechanisms (for an optimum phase) were explained for the BO-
105 flight test [1, 2] after an analysis of the results, it is the noise reduction mechanism that determines or ‘tailors’ the IBC input in the present effort – and this represents a different approach. Localized inputs are also preferred over harmonic inputs that extend over the entire azimuth based on the following philosophy. Pitch reductions over portions of the front of the rotor disk serve the very specific purpose of weakening the interacting vortices, and thus the parallel BVI events, in the first and fourth quadrants of the rotor disk. But why, then, is it necessary to vary the rotor blade pitch elsewhere around the azimuth (as with harmonic inputs that extend over the entire azimuth)? The location and
(a more limited) duration of the input are carefully selected based on analysis. Different
IBC input profiles and amplitudes are considered, and the effects of second quadrant and third quadrant inputs in reducing BVI noise are examined, individually, and in combination. 89
3.1 Description of Analysis and Results for the Baseline Configuration
To evaluate the changes in helicopter BVI noise levels associated with IBC inputs, a rotorcraft aeroelastic analysis [3] is used in conjunction with the free-wake code developed at the Pennsylvania State University by Tauszig and Gandhi [4, 5]. In the aeroelastic analysis, the rotor blades are assumed to undergo elastic flap bending and elastic torsion deformations. Evaluation of the free-wake geometry, blade response, and controls are carried out iteratively in a coupled wake-response-trim calculation procedure. To examine the blade-vortex interactions in detail, a system of overlapping low- and high-resolution azimuthal grids is used. For computational efficiency, the free- wake geometry is explicitly evaluated only at 80 azimuthal stations over each of four rotor revolutions (low-resolution grid). After computing the converged free-wake geometry (along with the corresponding rotor blade response and trim controls), the reference blade is allowed to time-march around the azimuth in very small increments; and the wake geometry is calculated at each of the 640 points of this high-resolution grid by linearly interpolating the geometry explicitly evaluated at the low-resolution grid points. The blade loading is calculated using this high-resolution grid (which allows impulsive changes due to BVI to be captured).
The Tauszig freewake code has several advantages over production codes, such as
CAMRAD II, that make it particularly useful for BVI analyses. The Tauszig code specifically tracks and records when blades pass in proximity to tip vortices. This allows 90 for the wake structure relative to the blades to be monitored, which gives a unique view of where and how interactions are taking place. The code also allows for the tip vortices of one blade at a time to be analyzed. In this mode the other vortices are “turned off”, thus allowing for the monitoring of specific vortices, and thus specific interactions, separate from others. This is particularly useful when an interaction has a beneficial influence on one interaction and a detrimental influence on another.
Numerical simulations in the present study are conducted on a 4-bladed model hingeless rotor similar to the model BO-105 rotor used in the first HART test [6]. The rotor has a 4 m diameter, a 12.1 cm chord, a 218 m/s tip speed, and a NACA 23012 airfoil with linear twist. The twist has been reduced from 8˚ for the HART rotor to 4˚ in the present simulations in order to avoid negative loading near the blade tip when IBC inputs are introduced. This avoids creation of dual tip vortices, which significantly complicate the problem. The rotor thrust coefficient is 0.0044, and a nominal advance ratio of 0.135 and a backward shaft tilt of 5˚ are considered in the simulations as these correspond to a moderate descent-rate, severe BVI condition.
The rotor is trimmed to a zero first harmonic flapping and then an IBC blade root pitch input is superposed over the collective and cyclic pitch inputs from the swashplate.
Thus, the total blade pitch may be represented as:
01csIBCcos 1 sin 3.1
Where,
f (, , ) for IBC 12 12 3.2 IBC 0 for 1 or 2 91 In Equation 3.2, is the amplitude of the IBC pitch input, the function f describes the profile, and ψ1 and ψ2 are the azimuthal locations corresponding to the beginning and end of the input. The new distorted wake geometry and high-resolution airloads due to the IBC inputs are then calculated, without retrimming the rotor. This allows the direct effects of the IBC inputs to be separated from those produced due to any change in the collective and cyclic pitch necessary to maintain trim. In the section entitled ‘Effects of Retrimming’, however, it is verified that the benefits obtained with localized IBC pitch inputs are essentially preserved when the rotor is retrimmed.
As the reference blade marches in increments on the high-resolution azimuthal grid, if any tip-vortex element in the rotor wake intersects the blade-rotor-shaft plane within a prescribed distance above or below the blade (a distance of three chord lengths is used in this study), this event is recorded. The vertical distance between the blade and the intersection point of the vortex segment on the blade-rotor-shaft plane is termed the miss- distance [5]. Events recorded in close proximity generate bands that represent the passage of tip-vortices close to the blade as it undergoes a revolution, with a radial or nearly radial band representing a parallel interaction with a vortex (see Figures 3.1 through 3.3 for the baseline – no IBC – case), which produces impulsive loads and significant BVI noise. In Figure 3.1 (and all similar figures in the current chapter) the following scheme is used: Interaction sites marked in black indicate the vortex is passing above the blade, while those in gray indicate the vortex is passing below the blade, and the larger the symbol, the smaller the miss-distance. Miss-distances of one chord length or larger are represented by the smallest symbols. Note that when symbols change between dark and light, that indicates that the wake is passing through the disk plane at that point. The 92 parallel interactions thus identified can be attributed to specific generating tip-vortices
[4]. For the 4-bladed rotor considered in this study, the tip-vortex from the blade preceding the reference blade, k, is referred to as the (k−1) vortex, that from the blade opposite the reference blade is the (k−2) vortex, that from the preceding blade is the
(k−3) vortex, and that from the reference blade itself is the (k−0) vortex. This is seen in
Figure 3.4.
ψ = 180°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords ψ = 270° ψ = 90°
(k−3) (k−0) Vortex Above1 Blade 0 -1Below Blade ψ = 0°
Figure 3.1: Baseline (no IBC case), top-view of blade-vortex interactions and miss- distances for full rotor disk.
93
ψ = 90°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords
(k−3) (k−0)
ψ = 0°
Figure 3.2: Baseline (no IBC case), top-view of blade-vortex interactions and miss- distances for first quadrant (advancing side interactions).
94
ψ = 270°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords (k−0)
ψ = 0°
Figure 3.3: Baseline (no IBC case), top-view of blade-vortex interactions and miss- distances for fourth quadrant (retreating side interactions).
95
(k-1)
(k-2) Reference Blade
Ω (k-0)
(k-3)
Figure 3.4: Relative positions of reference blade to tip-vortex generation.
For the baseline – no IBC – case, the dominant parallel interactions (radial bands; with large symbols, indicating small miss-distance) can be identified in the first and fourth quadrants of the rotor disk in Figures 3.2 and 3.3 respectively. In the first quadrant there appear to be two significant BVI events due to interactions with the (k−0) and the
(k−3) vortices. In the fourth quadrant, only a single significant parallel BVI event, with the (k−0) vortex, is observed. Figure 3.5 depicts the maximum bound circulation on a blade, which is representative of the maximum loading at every azimuthal location. The impulsive loading associated with the parallel interactions in the first and fourth quadrants is clearly evident (thin line). The vortex elements released in the front of the rotor disk (second and third quadrants) convect downstream and produce the advancing side (first quadrant) and retreating side (fourth quadrant) interactions, respectively. Since 96 the severity of the blade-vortex interactions depends, among other factors, on the strength of the striking vortex, the non-dimensional strengths, (Γ/(cVt)), are also monitored for all interactions (Figure 3.6). It is seen in Figures 3.6, 3.7, and 3.8 that the strength of the striking vortices progressively increases from the advancing side to the retreating side of the disk, and this is consistent with the increase in bound vorticity seen in Figure 3.5 over the 90˚ to 270˚ range (it is this vorticity convected downstream that produces BVI). There is a portion of the (k-3) interaction seen in Figures 3.6 and 3.7 that appears to be missing points. The large gap between points is an artifact of only detecting the miss-distance once per vortical element. In those cases where the interaction is very parallel this results in detection points that are spaced out. Thus, this can be used as an indication of a particularly parallel interaction. It should be noted that the scales on Figures 3.6, 3.7, and
3.8 are each different. 97
0.5 Baseline -3 Ramped IBC Input
0.4 Azimuthal region of advancing side interacting vortex generation 0.3
0.2
0.1 Advancing side BVI 0 Strength of Generated Vortex (Max.0 Bound Circ.) 90 180 270 360
Figure 3.5: Azimuthal variation of the strength of the generated vortices (maximum bound circulation).
98
ψ = 180°
Vortex Strength
0.45 0.43 0.41 0.39 0.37 0.35 ψ = 270° 0.33 0.31 0.29 0.27 0.25 0.23 0.21 0.19 0.17 0.15
ψ = 0°
Figure 3.6: Baseline (no IBC case) non-dimensional strength of interacting vortices for full rotor disk.
99
ψ = 90° Vortex Strength 0.20 0.19 0.18 0.17 0.16 0.15 0.14 0.13 0.12 0.11 0.10 0.09 0.08 0.07 0.06 (k−3) 0.05 (k−0) ψ = 0°
Figure 3.7: Baseline (no IBC case) non-dimensional strength of interacting vortices for first quadrant (advancing side interactions).
ψ = 270°
Vortex Strength 0.45 0.43 0.41 0.39 0.37 0.35 0.33 0.31 0.29 0.27 (k−0) 0.25
ψ = 0°
Figure 3.8: Baseline (no IBC case) non-dimensional strength of interacting vortices for fourth quadrant (retreating side interactions). 100 The high-resolution airloads from the analysis are used as input to the rotorcraft acoustic code WOPWOP [7], which provides the acoustic pressure time history over one rotor revolution at any specified observer location. From the acoustic pressure signal, the
BVI sound pressure level (BVISPL) is calculated by considering the 6th – 40th harmonics of the blade passage frequency, and this is used as the metric for BVI noise. The BVISPL is examined over a 19 by 24 grid of observer locations on a plane one rotor diameter below the rotor disk, to obtain information on both directivity as well as intensity of the total radiated BVI noise (Figure 3.9). In Figure 3.9 three high noise areas are visible. Two are the advancing and retreating side lobes; the third is due to interference effects between these two noise sources. The interference peak is not considered when determining peak advancing or retreating side noise. The approach developed by Gandhi and Tauszig [8], is used to determine the contributions from individual vortices to the total BVI noise. In this method, the inflow due to only one of the four tip vortices (either
(k−0), (k−1), (k−2), or (k−3)) is used to calculate loading on the reference blade. From
Figures 3.10-3.13 it is seen that interactions with the (k−0) and (k−3) vortices produce significant BVI noise on the advancing side, while noise on the retreating side is dominated by a single interaction with the (k−0) vortex. While this could be qualitatively deduced from Figures 3.1-3.3, Figures 3.10-3.13 provide detailed information on the magnitude and directivity of the radiated noise due to the various advancing and retreating side interactions, and the contribution of each to the total BVI noise in
Figure 3.9. Although the individual interactions have their own noise directivity there is some superposition of the noise lobes, so the total peak noise level on the advancing side
(109.2 dB in Figure 3.9) is somewhat higher than the peak levels due to the primary 101 (k−3) interaction (107.4 dB peak, Figure 3.13) and the secondary (k−0) interaction
(106.5 dB peak, Figure 3.10) on the advancing side. The peak noise level on the retreating side (107.9 dB in Figure 3.9) is considerably higher than that from the (k−0) retreating side interaction (101.4 dB in Figure 3.10) due to contributions from the (k−3) interaction. Plots such as those in Figures 3.10-3.13 are useful in examining and understanding the influence of a given IBC input on the key blade-vortex interactions.
4
90 3 Adv. 92 side 2 94 lobe 96 1 98 100 102 108 0 104 X(m) 108 -1 106
-2 Ret. 104 102 side 102 -3 Interference 100 lobe 104 -4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.9: Baseline (no IBC) BVISPL due to full wake (109.2 peak advancing side, 107.9 dB peak retreating side).
102
4
88 3 Adv. 90 side 2 92 94 lobe 1 96 102 0 98 104 X(m) 100 104 106 -1
-2 Ret. side -3 Interference lobe -4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.10: Baseline (no IBC) BVISPL due to (k−0) vortex (106.5 dB peak advancing side, 101.4 dB peak retreating side).
103
4 80 80 3 82 2 80
82 84 1 86 88 0 90 X(m)
-1 92
-2
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.11: Baseline (no IBC) BVISPL due to (k−1) vortex.
104
4
84 3 88 88 86 86 2 88
90 1 92
94 0
X(m) 96 -1
-2 98
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.12: Baseline (no IBC) BVISPL due to (k−2) vortex.
105
4
86 90 88 3 94 98 92 96 102 100 2 106 104
1
0 X(m)
-1
-2
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.13: Baseline (no IBC) BVISPL due to (k−3) vortex (107.4 dB peak advancing side).
3.2 Localized IBC Inputs to Reduce Interacting Vortex Strength
Blade pitch inputs that reduce the severity of the impulsive blade loading associated with blade-vortex interaction can be expected to reduce BVI noise. The impulsive blade loading can be reduced through blade pitch variations that change the geometry of the interaction (primarily by increasing the blade-vortex miss-distance), or blade pitch reductions that reduce the strength of the interacting vortex. The present study 106 focuses on the second type of input. This was chosen because a reduction in interacting vortex strength results in BVI noise reductions over a wide range of conditions [ 2]. In contrast, IBC inputs that ‘blow the vortex away’ from the blade, or increase the miss- distance in one operating condition, may actually exacerbate the BVI noise in another condition. Further, since multiple parallel interactions can occur, IBC inputs that increase the blade-vortex miss-distance for a primary event can potentially reduce the miss- distance for a secondary BVI event. In general, when the IBC input seeks to change the geometry of the interaction, its amplitude and location (for a localized input), or phase
(for a harmonic input), require variation with operating condition – making a closed-loop controller indispensable. It should be noted that while it is often suggested in the literature that reducing blade loading at the time of interaction could potentially reduce noise, there is evidence that such a reduction has little impact on BVI noise [9].
The tip vortex elements that result in the dominant advancing side parallel interactions (in the first quadrant of the rotor disk) are released from the rotor blades in a preceding revolution in the second quadrant, whereupon they convect downstream and eventually produce the BVI event in the first quadrant. Similarly, tip vortex elements that result in the retreating side parallel interactions (in the fourth quadrant of the rotor disk) are released in the third quadrant. Thus, any effort to reduce the strength of the interacting vortices must involve pitch reductions over specific sections of the second and third quadrants of the rotor disk (the front of the rotor disk). The reduced lift on the blade over these sections reduces the strength of the tip vortex elements released, and this in turn results in milder BVI events downstream in the first and fourth quadrants. 107 In Figure 3.5 it is seen that an IBC pitch reduction in the second quadrant does, in fact, produce milder advancing side BVI events (the thick line shows less impulsive variation in maximum blade loading in the first quadrant, relative to the baseline case).
Figure 3.14 shows how the azimuthal range over which pitch reduction is required in the second quadrant can be deduced based on the location of the dominant first quadrant interactions. In a similar manner, the azimuthal range over which pitch reduction is required in the third quadrant can be deduced from the location of the dominant fourth quadrant BVI events. It should be noted that changes in the local lift also change the local inflow, and this can, in principle, be expected to have some effect on the blade-vortex interaction geometry as well. 108
Range of IBC ψ = 180° pitch reduction
Tip vortices convecting downstream
BVI events
ψ = 0° Figure 3.14: Tip vortex elements generated over portions of the second quadrant resulting in first quadrant BVI.
3.3 IBC Input Profile
In Higher Harmonic Control the pitch input is always sinusoidal in form. In the case of Individual Blade Control, in addition to the possibility of introducing localized pitch inputs to target specific physical phenomena (such as producing weaker interacting 109 vortices), there is also broad flexibility in the pitch input profile. Four different pitch input profiles are examined in this section, as illustrated in Figure 3.15 – (i) a truncated step, (ii) a half-period sine pulse, (iii) a full period cosine pulse, and (iv) a ramp down – hold – ramp up profile.
Truncated Full Period Step Cosine
Ramp Down – Hold – Half Period Ramp Up Sine
Δψr
Figure 3.15: IBC input profiles.
The first IBC input profile considered is the truncated step (Figure 3.15). Here,
θIBC jumps to its target value, − , instantaneously, holds that value for the entire duration of the input, and then jumps back to zero. Thus,
f (, 12 , ) 1 3.3
Such an input is impossible to achieve in practice due to IBC actuator bandwidth and blade inertia considerations. However, if the goal is to reduce blade pitch in order to reduce loading and therefore decrease vortex strength over some azimuth range, 110
12, the truncated step profile provides the maximum overall reduction over the entire duration. Thus, such an input serves as a benchmark to which the effectiveness of the other profiles (Figure 3.15), which produce the largest reductions only over a portion
of the azimuthal range, 12, can be compared.
The influence of the truncated step input, with amplitude = 1˚, 1 = 120˚, and
2 = 160˚ (a range of 40˚ centered at 140˚ aimed at reducing the strength of the advancing side interactions), on the BVISPL associated with the primary (k−3) advancing side interaction is seen in Figure 3.16. This input reduces the peak advancing side BVISPL by 4.7 dB, from 107.4 dB to 102.7 dB (compare Figures 3.13 and 3.16).
However, it should be noted that the sudden changes in blade lift at the edges of the truncated step themselves act as a noise source – hereafter referred to as IBC noise. This noise can be clearly seen directed towards the upper region of Figure 3.16, and becomes more pronounced with increasing amplitude, (results for higher are not shown). 111
4
92 3 94 96 98 100 2 102
1
102 0
X(m) Peak -1 102.7 dB
-2 94 92 -3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.16: (k−3) vortex BVISPL due to a truncated step IBC input profile ( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚).
To alleviate IBC noise, the half-period sine pulse (Figure 3.15), without sudden changes in pitch, was implemented. Thus,
1 f (,12 , ) sin 3.4 21
The values of ψ1 and ψ2 (range of the input) and amplitude are the same as those for the truncated step. Figure 3.17 shows that the peak advancing side BVISPL of the primary (k−3) interaction was reduced by 4.3 dB, to 103.1 dB. The peak BVISPL 112 reduction is slightly smaller than that obtained with the truncated step input, due to a lesser overall reduction in interacting vortex strength that this input produces. There is little evidence of IBC noise (as was seen with the truncated step input in Figure 3.16).
However, an increase in amplitude, , or decrease in the range of the IBC input, (ψ2 −
ψ1), will result in sharper, or more impulsive, changes in pitch, which will in turn yield
IBC noise.
4 84 92 3 86 96 88 92 94 90 98 100 2 102
1
0 102 X(m) Peak -1 103.1 dB 94 -2 92 90 -3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.17: (k−3) vortex BVISPL due to a half-period sine IBC input profile ( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚). 113 The next input profile considered was a full period cosine pulse (Figure 3.15).
Here,
1 2 1 f (,12 , ) 1cos 3.5 2 21
This is the smoothest waveform considered, without any sudden change in pitch or pitch rate. However, for a specified azimuthal range, (ψ2 − ψ1), and pitch amplitude,
, it produces the smallest overall reduction in pitch, relative to the other inputs
(compare the various curves in Figure 3.15). Thus, the reduction in vortex strength is also the smallest and, as seen in Figure 3.18, the peak advancing side BVSPL of the primary
(k−3) interaction is reduced by only 3.5 dB, to 103.9 dB. This input produced no visible
IBC noise. 114
4
86 3 84 90 98 88 92 92 102 96 2 100
1
0 X(m) Peak -1 103.9 dB -2 92 90 -3 88
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.18: (k−3) vortex BVISPL due to a full-period cosine IBC input profile ( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚).
The last input profile considered is the ramp down – hold – ramp up profile
(Figure 3.15), represented mathematically as Equation 3.6:
1 for 11 r r f ,,12 1 for 1 rr 2 3.6 1for2 r 22r r 115
Here Δψr is the azimuthal duration of the ramp down (or ramp up) section of the input (nominally taken to be one quarter of the total duration, ψ2 – ψ1). Such an input alleviates the sudden changes in pitch and the resulting IBC noise associated with the truncated step input, but maintains the peak target value, − , over a large portion of its range of application. Thus, compared to the sine or cosine pulse profiles in Figure 3.15, this input produces a larger overall pitch reduction over the range ψ1 to ψ2, which in turn delivers a larger overall reduction in the strength of the interacting vortices producing the advancing side interactions. Thus, the ramp down – hold – ramp up input (hereafter referred to simply as the ramped input) combines the advantages of the truncated step
(large overall reductions in strength over the entire length of the interacting vortex) with those of the sine or cosine input profiles (little IBC noise and easier to implement).
Figure 3.19 shows that with this input the peak advancing BVI noise of the (k−3) interaction is reduced to 101.9 dB, a reduction of 5.5 dB from the baseline (compare to
Figure 3.13). Only the faintest traces of IBC noise are visible at the top of Figure 3.19, and the levels do not approach those produced by the truncated step. While BVI noise reductions with the ramp input are expected to be larger than those due to the sine or cosine inputs, it is surprising that the reductions are even greater than those produced by the truncated step (compare Figures 3.16 and 3.19) which, in principle, should produce the largest overall reduction in interacting vortex strength. The greater reduction in advancing side BVI noise with the ramp input is attributed to two factors. First, the IBC noise associated with the truncated step input superposes with the advancing side BVI noise lobe. Second, the IBC inputs cause subtle changes in interaction geometry, with the 116 ramped input happening to produce more favorable changes compared to the truncated step input.
4 88 3 90 92 96 94 100 98 2
1
0
X(m) Peak -1 101.9 dB 94 -2 92 90 88 -3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.19: (k−3) vortex BVISPL due to a ramped IBC input profile ( = 1˚, ψ1 = 120˚ and ψ2 = 160˚, or ψcenter = 140˚ and ψrange = 40˚).
For the subsequent simulations in this chapter, ramped IBC inputs are used. As in the present section, the ramped inputs maintain their peak value for 50% of their range, with ψr = (ψ2 – ψ1)/4, though this can in principle be tailored to maintain an ideal balance between effectiveness of the IBC input in reducing the interaction strength and resulting
IBC noise (which is dependent upon the rate of change of pitch). 117 By its nature, the ramped input minimizes the actuation rate necessary to achieve the desired amplitude within a specified azimuthal range. This does not, however, minimize the load necessary to actuate the blade. In fact, the sharp corners present in the ramped model represent an instantaneously infinite load, and in practice will be somewhat rounded, the degree of which will depend upon the capacity of the actuators used.
3.4 Second Quadrant Vortex Strength Reduction
This section examines the effectiveness of ramped inputs in the second quadrant, intended to reduce the strength of the tip vortex as it is generated, and thereby weaken the advancing side interactions. Investigation of the vortex wake showed that the center of the vortex section involved in the primary (k−3) BVI was generated around 140˚ azimuth, so the second quadrant inputs are centered here. It is sometimes more convenient to describe the IBC inputs in terms of range (ψrange) and center (ψcenter), which can easily be converted back to ψ1 and ψ2 as follows:
range 1 center 2 3.7 range 2 center 2
Since the second quadrant inputs are meant to reduce the strength of the advancing side interacting vortices, these will be compared to the corresponding strengths for the baseline configuration (no IBC inputs) in Figure 3.7. When a = 1˚ amplitude was applied over a range of 40˚ (the pitch input only reaches the peak value over a 20˚ 118 azimuthal range), the strength of the interacting vortices, shown in Figure 3.20, is considerably reduced. As already seen in Figure 3.19, BVI noise associated with the primary (k−3) advancing side event is reduced by 5.5 dB. For this input, BVI noise due to interactions with the (k−0) vortex is shown in Figure 3.21. The (k−0) interaction is the secondary advancing side interaction, and the peak noise is reduced by 2.8 dB (compare to Figure 3.10). The (k−0) retreating side noise, however, has a larger peak value and is radiated over a larger area, relative to the baseline. This is due to the (unintended) influence the IBC input has on the miss-distance of the (k−0) retreating side interaction, which is clearly seen to have decreased in Figure 3.22, relative to the baseline
(Figure 3.3), especially over the 60-80% span range.
ψ = 90° Vortex Strength 0.20 0.19 0.18 0.17 0.16 0.15 0.14 0.13 0.12 0.11 0.10 0.09 0.08 0.07 0.06 0.05 (k−3) (k−0) ψ = 0°
Figure 3.20: BVI non-dimensional vortex strength for = 1˚ ramped IBC input centered at ψ = 140˚. 119
4
90 3 92
2 94 96
98 1 96 100 102 0 98 X(m) 100 -1 100 102 -2 98 104 96 -3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.21: BVISPL due to (k−0) vortex for = 1˚ ramped IBC input centered at ψ = 140˚ (103.7 dB peak advancing side, 105.7 dB peak retreating side).
120
ψ = 270°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords (k−0)
Vortex Above1 Blade 0 -1Below Blade ψ = 0°
Figure 3.22: Blade-vortex interactions and miss-distance for = 1˚ ramped IBC input centered at ψ = 140˚.
When the IBC amplitude is increased to = 2˚, Figure 3.23 shows that the strengths of the interacting vortices in the first quadrant are further reduced. As a consequence, BVI noise due to the (k−3) interaction (Figure 3.24) is reduced to 98.5 dB
(an 8.9 dB reduction), rendering this interaction insignificant. For this IBC input, BVI noise due to interactions with the (k−0) vortex is shown Figure 3.25. Further reductions in advancing side noise are observed (compare Figure 3.25 to the 1˚ IBC case in
Figure 3.21, and to the baseline in Figure 3.10). The retreating side noise, on the other hand, has significantly worsened. This due, again, to worsening interaction geometry
(compare Figures 3.22 and 3.26). The interaction has become more parallel, as seen from the large spaces between symbols in the interaction “band” between 60% and 80% span 121 [5]. Additionally, the radial location where the vortex actually crosses the blade has moved further outboard.
ψ = 90° Vortex Strength 0.20 0.19 0.18 0.17 0.16 0.15 0.14 0.13 0.12 0.11 0.10 0.09 0.08 0.07 0.06 0.05 (k−3) (k−0) ψ = 0°
Figure 3.23: BVI non-dimensional vortex strength for = 2˚ ramped IBC input centered at ψ = 140˚.
122
4 92 3 94 96 2 98
1 92 90 88 0 86 X(m)
-1 92 88 -2 90
-3 94
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.24: BVISPL due to (k−3) vortex for = 2˚ ramped IBC input centered at ψ = 140˚ (98.5 dB peak).
123
4 86 92 3 94 86 92 2 90 88 96 98 1 94 100 0 102 X(m) 104 -1 94 106 92 -2 108
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.25: BVISPL due to (k−0) vortex for = 2˚ ramped IBC input centered at ψ = 140˚ (108.5 dB peak).
124
ψ = 270°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords (k−0)
Vortex Above1 Blade 0 -1Below Blade ψ = 0°
Figure 3.26: Blade-vortex interactions and miss-distance for = 2˚ ramped IBC input centered at ψ = 140˚.
Figure 3.27 shows even further reductions in the strengths of the interacting vortices in the first quadrant when the IBC amplitude is increased to = 3˚ (compare to
Figures 3.7, 3.20, and 3.23). Noise due to primary (k−3) interaction on the advancing side had already been reduced to insignificance with the 2˚ IBC input (Figure 3.24), so no additional benefits were obtained for this interaction (BVI noise results not shown).
Figure 3.28 shows the noise due to interactions with the (k−0) vortex. The advancing side interaction with the (k−0) vortex is no longer acoustically significant (compare with baseline, Figure 3.10). In addition, the retreating side noise associated with this vortex, while greater than the baseline, is lower than that seen with the 2˚ IBC input
(Figure 3.25). Figure 3.29 shows the total BVI noise levels, considering all the vortices in 125 the wake. Although there is only a 2.2 dB reduction in the peak overall BVISPL
(compare with Figure 3.9, for the baseline), this is due to retreating side noise. The advancing side noise has been completely eliminated, as was expected with the IBC input. Retreating side noise has also been reduced from the baseline despite a rise in the
(k−0) component as the contributions from the (k−3) interaction were eliminated. It is again emphasized that the second quadrant IBC inputs considered in this section did not target the retreating side noise. Since both of the advancing side interactions have been eliminated with the 3˚ input, further increases in IBC amplitude would have no benefit, and are not examined. In fact, some traces of IBC noise are observed on the upper left portion of Figures 3.28 and 3.29 due to the increased amplitude of the IBC input.
ψ = 90° Vortex Strength 0.20 0.19 0.18 0.17 0.16 0.15 0.14 0.13 0.12 0.11 0.10 0.09 0.08 0.07 0.06 0.05 (k−3) (k−0) ψ = 0°
Figure 3.27: BVI non-dimensional vortex strength for = 3˚ ramped IBC input centered at ψ = 140˚. 126
4
96 84 86 3 98 88 96 94 92 90 2 98
1 94 90 0 96
X(m) 100 98 94
-1 104 102
-2 90 106
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.28: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ (106.8 dB peak).
127
4 98 96 88 3 98 90 92 2 94
96 1 98 92 100 0 102 X(m) 104 -1 106 -2
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.29: BVISPL due to full wake for = 3˚ ramped IBC input centered at ψ = 140˚ (107.0 dB peak).
From the results in this section, some of the second quadrant IBC inputs, intended to weaken the vortex elements producing the advancing side interactions, appear to be inadvertently increasing peak retreating side noise and size of the retreating side BVI noise lobe through changes in wake geometry. It is interesting to note that Reference 10 reported a parallel observation – that a third quadrant flap input (targeting the retreating side interactions) increased the intensity of the advancing side BVI. Although such phenomena are likely dependent on both the configuration and operating conditions, it 128 suggests that effects of IBC inputs targeted towards specific interactions must be examined over the entire rotor disk, as secondary interactions for the baseline can be unintentionally exacerbated.
3.5 Third Quadrant Vortex Strength Reduction
Just as the second quadrant IBC inputs in the previous section reduced the advancing side BVI noise, third quadrant IBC inputs could similarly be expected to reduce the retreating side BVI noise. By reducing the blade pitch over some azimuthal range in the third quadrant, the strength of the tip vortices released is reduced, and the weaker tip vortex segments then convect downstream to produce milder retreating side
BVI events in the fourth quadrant. This section examines the effectiveness of ramped
IBC inputs in the third quadrant in reducing the retreating side BVI noise.
From Figure 3.3 it is seen that for the baseline there is only one dominant blade- vortex interaction on the retreating side, the (k−0) interaction, with the interacting vortex elements released by the reference blade itself, in a preceding revolution. The retreating side interacting vortex strengths for the baseline (no IBC) are shown in Figure 3.8, as a reference. An IBC input centered at 225˚ was used to weaken the strength of the tip vortex segment producing the parallel (k−0) interaction in the fourth quadrant.
When a = 1˚ input was applied over a range of 40˚ (with a 10˚ ramp down, 20˚ hold, 10˚ ramp up profile, as in the case of the second quadrant inputs) the strength of the interacting vortices on the retreating side (shown in Figure 3.30) is reduced, as expected.
The BVI noise due to the full wake is shown in Figure 3.31. Comparison with Figure 3.9 129 (no IBC input) reveals a 1.1 dB reduction in the retreating side noise, significantly less than achieved on the advancing side through use of a second quadrant input of the same magnitude. The noise reductions due to reduced vortex strength were mitigated, in part, by a change in the interaction geometry (compare Figure 3.32 and Figure 3.3). It is evident that the (k−0) vortex, which was passing below the blade in the 65-80% radial range for the baseline, remains much closer to the blade over this radial range, when the
IBC input is introduced (note the larger symbols, forming a thicker interaction band in
Figure 3.32). Overall, the interaction occurs over a larger radial range, occurs more outboard relative to the baseline, and is also more parallel (as seen by the spaces between symbols in the 40-60% span range). All of these factors are detrimental from a noise generation standpoint [11]. 130
ψ = 270°
Vortex Strength 0.45 0.43 0.41 0.39 0.37 0.35 0.33 0.31 0.29 (k−0) 0.27 0.25
ψ = 0°
Figure 3.30: BVI non-dimensional vortex strength for = 1˚ ramped IBC input centered at ψ = 225˚.
131
4 90 3 92 94 2 96 98 100 1 102
0 104 108 X(m) 106 106 -1 108 104 102 -2 100
-3 98
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.31: BVISPL due to full wake for = 1˚ ramped IBC input centered at ψ = 225˚ (108.7 dB peak advancing side, 106.8 dB peak retreating side).
132
ψ = 270°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords
(k−0) Vortex Above1 Blade 0 -1Below Blade ψ = 0°
Figure 3.32: BVI geometry and miss-distance for = 1˚ ramped IBC input centered at ψ = 225˚.
When the amplitude of the IBC input is increased to = 2˚, further reduction in retreating side interacting vortex strength is observed (compare Figure 3.33 with
Figures 3.30 and 3.8), and this results in a further reduction in the peak BVISPL on the retreating side, as seen in Figure 3.34 (compared to the noise levels seen with the = 1˚ amplitude input in Figure 3.31 and baseline in Figure 3.9). A 2.8 dB peak BVISPL reduction is achieved (from 107.9 dB to 105.1 dB). It is evident that for comparable IBC input amplitudes, third quadrant inputs do not reduce retreating side BVI noise to the degree second quadrant inputs reduce advancing side BVI noise. On closer examination, it can be seen that although the actual reductions in vortex strength are comparable to those on the advancing side, the baseline values are much higher on the retreating side 133 (see Figures 3.5 and 3.6). This observation was also reported in Reference 12.
Consequently, the third quadrant input results in a reduction in strength of the order of
20%, compared to reductions of the order of 60% associated with a second quadrant IBC input of the same amplitude ( = 3˚). It has previously been reported that a specified absolute reduction in vortex strength is more effective in reducing BVI noise when the initial interaction strength is moderate-to-low, and less effective when the initial vortex strength is high [11]. The present results are consistent, in this regard.
ψ = 270°
Vortex Strength 0.45 0.43 0.41 0.39 0.37 0.35 0.33 0.31 0.29 0.27 (k−0) 0.25
ψ = 0°
Figure 3.33: BVI non-dimensional vortex strength for = 2˚ ramped IBC input centered at ψ = 225˚.
134
4 86 86 3 90 92 2 94
1 96 98 108 0 100 106 X(m) 102 -1 104 102 104 -2 100 102 98 -3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.34: BVISPL due to full wake for = 2˚ ramped IBC input centered at ψ = 225˚ (108.7 dB peak advancing side, 105.1 dB peak retreating side).
In summary, the effectiveness of localized third quadrant IBC inputs in reducing the interacting vortex strength and thus retreating side BVI noise is limited, compared to the reductions in advancing side BVI noise due to second quadrant inputs. This is due to both the higher baseline vortex strengths on the retreating side, and the accompanying changes in interaction geometry, which partially negate the relatively modest benefits due to strength reduction. In general, it can be said that reducing vortex strength could reduce
BVI noise in a predictable manner provided interaction geometry changes are not 135 significant, or the strength of the interacting vortex is reduced to such levels as to become negligible.
3.6 Combination of Second and Third Quadrant Vortex Strength Reduction
This section examines the effects of simultaneously introducing IBC inputs in the second and third quadrants, which were individually considered in the last two sections.
The = 3˚ amplitude input over 40˚ centered at 140˚, examined in the section “Second
Quadrant Vortex Strength Reduction”, is used as the reference, and third quadrant inputs of increasing magnitude (which only had a limited effect when considered by themselves, in the last section) are successively added to the second quadrant input. For the = 3˚ second quadrant input alone recall that the advancing side interactions were completely eliminated, and the (k−0) retreating side interaction was the dominant noise source. By adding a third quadrant IBC input of = 1˚ amplitude, the strength of the fourth quadrant interaction is reduced slightly, and this results in a modest 0.8 dB reduction of the (k−0) peak retreating side noise (compare Figure 3.35 to Figure 3.28). Increasing the third quadrant IBC amplitude to = 2˚ results in further reductions in the vortex strength, and the peak BVISPL of the retreating side (k−0) interaction is reduced by an additional 0.5 dB (Figure 3.36). When the fourth quadrant amplitude is increased to =
3˚, however, an additional sharp drop of nearly 3 dB is observed for the (k−0) retreating side interaction noise (compare Figure 3.37 to Figure 3.36). In addition to the further reduction in the vortex strength of the (k−0) retreating side interaction, the sharp reduction in noise is attributed to a favorable change in BVI geometry. Comparing 136 Figures 3.38 and 3.39 (corresponding to 2˚ and 3˚ third quadrant inputs, respectively), it is seen that for the 3˚ input (Figure 3.39) the blade-vortex miss-distance is greater over the 60-80% span range, where the interaction is very parallel. In the 80-90% radius range, where the vortex actually intersects the blade, the 3˚ input has resulted in the interaction becoming more oblique (compared to the 2˚ input, where it remained more parallel). The addition of a 3˚ third quadrant input to the second quadrant input resulted in a total 4.2 dB reduction in the peak (k−0) retreating side BVISPL (compare Figure 3.37 to Figure 3.28, for only the second quadrant IBC input). 137
4 86 98 3 96
2 88 90 92 1 98 94 96 0 100
X(m) 102
-1 104
-2
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.35: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ and = 1˚ ramped IBC input centered at ψ = 225˚ (106.0 dB peak).
138
4
98 86 96 3 86
2 88 90 92 1
98 94 96 0 100
X(m) 92 102 -1 104 88
-2
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.36: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ and = 2˚ ramped IBC input centered at ψ = 225˚ (105.5 dB peak).
139
4 86 98 3 88
96 96 2 92
1 90
0 92 X(m) 96 -1 94 100 98 -2 102 90
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.37: BVISPL due to (k−0) vortex for = 3˚ ramped IBC input centered at ψ = 140˚ and = 3˚ ramped IBC input centered at ψ = 225˚ (102.6 dB peak).
140
ψ = 270°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords (k−0) Vortex Above1 Blade 0 -1Below Blade ψ = 0°
Figure 3.38: BVI geometry and miss-distance for = 3˚ ramped IBC input centered at ψ = 140˚ and = 2˚ ramped IBC input centered at ψ = 225˚.
141
ψ = 270°
Miss-Distance . > 1 chord . 0.5 chords . 0 chords (k−0) Vortex Above1 Blade 0 -1Below Blade ψ = 0°
Figure 3.39: BVI geometry and miss-distance for = 3˚ ramped IBC input centered at ψ = 140˚ and = 3˚ ramped IBC input centered at ψ = 225˚.
The effect of the combined second and third quadrant inputs (each of 3˚ amplitude) on the total noise levels, when considering the entire wake and all the interactions, is shown in Figure 3.40. The peak BVISPL of 104.6 dB in Figure 3.40 corresponds to a 2.4 dB reduction relative to the case when only a second quadrant IBC input was used (Figure 3.29), and a full 4.6 dB reduction relative to the baseline
(Figure 3.9). 142
4 90
3 98 96 92
2 94 94
1 100
96 0 98 X(m) 102 100 -1 104 94 -2
-3 92 -4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.40: BVISPL due to full wake for = 3˚ ramped IBC input centered at ψ = 140˚ and = 3˚ ramped IBC input centered at ψ = 225˚ (104.6 dB peak).
3.7 Effects of Retrimming
All previous cases were examined without retrimming the rotor to the nominal thrust coefficient of 0.0044 and zero first harmonic flapping in order to isolate the effects of the IBC inputs from those of modified control inputs to maintain trim. For the = 3˚
IBC input over 40˚ azimuth, centered at 140˚ case, the rotor is now retrimmed to examine 143 what effects this has on the total BVI noise level. Retrimming the rotor caused the control inputs to go from θ0 = 10.3˚, θ1c = 2.7˚, θ1s = −3.8˚ to θ0 = 10.3˚, θ1c = 2.2˚, θ1s = −3.4˚.
Figure 3.41 shows the BVISPLs due to the full wake. It is seen that the advancing side noise is no longer completely eliminated as it had been (compare to the baseline in
Figure 3.9 and the unretrimmed case in Figure 3.29), though a reduction of 5.4 dB on the advancing side is still achieved. It is seen that retrimming actually reduces the retreating side noise. Whereas the peak retreating side BVISPL had decreased by 0.9 dB without retrimming the rotor, with retrimming it was reduced 1.9 dB. It should be noted that this effect was unintentional and will likely change for different flight conditions. Figure 3.42 illustrates why the advancing side peak BVISPL benefits were reduced by retrimming. In order to compensate for the reduced lift over the azimuthal range of the IBC input, the pitch, and therefore lift, were increased over the advancing side of the rotor disk. This reduced the effective pitch change due to the IBC input, and therefore the benefits to noise. Though the magnitude of the resulting noise reduction has been affected, the mechanism by which the BVI is weakened (strength of the interacting vortex) is preserved, and trends seen in the unretrimmed cases hold when the rotor is retrimmed. 144
4 90 88 90 3 92 92 94 2 94 1 96 98 98 0 100 X(m) 102 102 -1 104 100
-2 98
-3 96
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 3.41: BVISPL due to full wake with rotor retrimmed for the = 3˚ ramped IBC input centered at ψ = 140˚ (103.8 dB peak advancing side, 106.0 dB peak retreating side).
145
0.5 Baseline -3 Ramped IBC Input -3 Ramped IBC Input 0.4 (retrimmed) Azimuthal region of advancing side interacting vortex generation 0.3
0.2
0.1 Advancing side BVI 0 Strength of Generated Vortex (Max.0 Bound Circ.) 90 180 270 360
Figure 3.42: Azimuthal variation of maximum (non-dimensional) bound circulation (representative of the strength of generated vortices), with the case of the retrimmed rotor included.
3.8 Conclusions
This chapter examines the influence of Localized Individual Blade Control root pitch actuation inputs on Blade-Vortex Interaction noise. IBC inputs (pitch reductions) 146 were specifically considered over portions of the second quadrant of the rotor disk to reduce the strength of the vortex elements that convect downstream and produce advancing side BVI in the first quadrant. Similarly, pitch reductions over portions of the third quadrant were designed to reduce the strength of the vortex elements that produced retreating side BVI in the fourth quadrant of the rotor disk. The localized IBC pitch inputs were introduced over a limited 40˚ azimuthal range, and the second and third quadrant inputs were considered both individually, as well as in combination. Different
IBC pitch input amplitudes and profiles – such as a truncated step, a ramped input, a half- period sine pulse, and a full-period cosine pulse – were considered. Simulations were conducted for a 4 meter diameter model rotor similar to that used in the HART test, for a condition representative of a low-speed descent characterized by high BVI noise. Based on the results presented in this chapter, the following conclusions can be drawn:
1. The ramp-down-hold-ramp-up IBC input profile was found to be the most
effective of those considered in reducing BVI noise. For a given IBC pitch input
amplitude, , this profile resulted in the pitch being reduced to the maximum
value (of − ) over a substantial azimuthal range, compared to the half-period
sine pulse or the full-period cosine pulse. However, since there is no sharp or
sudden change in pitch (as with the truncated step input), it is implementable, and
since there is no instantaneous change in lift, the IBC input does not itself become
a significant source of noise (IBC noise).
147 2. Ramped IBC inputs in the second quadrant were highly effective in reducing the
advancing side BVI noise (from first quadrant BVI events). Input amplitudes of
= 1˚, 2˚, and 3˚ were considered. For the 1˚ input, the advancing side noise was
substantially reduced (up to 5.5 dB for the primary (k−3) interaction), and for the
2˚ and 3˚ inputs, advancing side noise was reduced to the background level.
However, the second quadrant inputs resulted in changes in wake and retreating
side interaction geometry so as to exacerbate the noise associated with the
dominant (k−0) retreating side interaction. The overall effect of the 3˚ second
quadrant input was to eliminate the advancing side noise lobe (from a baseline
level of 109.2 dB) and produce a peak retreating side noise level of 107.0 dB
(compared to 107.9 dB for the baseline). Thus, it is seen that when a localized
input (second quadrant input, in this case) is effective in its objective (reducing
advancing side noise), BVI noise levels due to sources that this input did not
address (retreating side BVI) gain prominence. This limits the total noise
reductions (from 109.2 dB for the baseline on the advancing side, to 107.0 dB
with the second quadrant IBC input, on the retreating side).
3. Ramped IBC inputs in the third quadrant have a more limited effect in reducing
retreating side BVI noise than the second quadrant inputs have on the advancing
side noise. Although the actual reductions in the strength of the interacting
vortices are comparable for second and third quadrant inputs of similar
amplitudes, the percentage reduction in strength is much smaller with the third
quadrant inputs. This is due to the larger circulation on the retreating side of the 148 disk (to compensate for lower dynamic pressures). In addition, the third quadrant
IBC inputs that reduce the strength of the interacting vortices appear to produce
detrimental changes in BVI geometry and miss-distance that partially negate the
benefits associated with the reductions in vortex strength. For a = 1˚ third
quadrant IBC input, the peak retreating side BVI noise level reduced by 1.1 dB,
and for = 2˚ the reduction was 2.8 dB.
4. When a ramped third quadrant IBC input is added to a ramped second quadrant
input, the largest overall reductions in BVI sound pressure levels are obtained.
When a = 3˚ amplitude is used for both the second and third quadrant inputs,
the advancing side BVI noise lobe is eliminated and the retreating side peak BVI
noise is reduced to 104.6 dB. Comparing to the baseline (109.2 dB peak
advancing side, and 107.9 dB peak retreating side), this represents a 4.6 dB net
reduction in peak BVI noise.
5. Localized IBC pitch reductions would require retrimming of the rotor.
Retrimmining the rotor in the presence of a second quadrant IBC input results in
an overall increase in lift over the advancing side and the front of the rotor disk to
compensate for the reduction in lift due to the localized IBC input. Although this
increase partially negates the localized lift reduction associated with the IBC
input, the overall mechanics and the effects of the IBC input are preserved. For a
3˚ amplitude second quadrant input, an advancing side peak BVI noise reduction 149 of 5.4 dB was still obtained, although the BVI noise was not completely
eliminated.
References
1. Splettstoesser, W. R., Schlutz, K. –J., van der Wall, B. G., Buchholz, H., Gembler, W., and Niesl, G., “The Effect of Individual Blade Pitch Control on BVI Noise – Comparison of Flight Test and Simulation Results,” Proceedings of the 24th European Rotorcraft Forum, Marseilles, France, Sept. 15-17, 1998.
2. Kube, R., van der Wall, B. G., Schlutz, K. –J., Splettstoesser, W. R., “IBC Effects on BVI Noise and Vibrations. A Combined Numerical and Experimental Investigation,” Proceedings of the 55th Annual Forum of the American Helicopter Society, Montreal, Canada, May 25-27, 1999.
3. Munsky, B., “An Analysis of Helicopter Blade-Vortex Interaction Noise With Flight Path or Attitude Modification,” MS Thesis, Department of Aerospace Engineering, The Pennsylvania State University, 2002.
4. Tauszig, L., “Numerical Detection and Characterization of Blade-Vortex Interactions Using a Free Wake Analysis,” MS Thesis, Department of Aerospace Engineering, The Pennsylvania State University, 1998.
5. Gandhi, F., and Tauszig, L., “A Critical Evaluation of Various Approaches for the Numerical Detection of Helicopter Blade-Vortex Interactions,” Journal of the American Helicopter Society, Vol. 45, No. 3, pp. 179-190, July 2000.
6. Splettstoesser, W. R., Kube, R., Wagner, W., Seelhorst, U., Boutier, A., Micheli, F., Mercker, E., and Pengel, K., “Key Results from a Higher Harmonic Control Aeroacoustic Rotor Test (HART),” Journal of the American Helicopter Society, Vol. 42, No. 1, Jan. 1997, pp. 58-78.
7. Brentner, K. S., “Prediction of Helicopter Rotor Discrete Frequency Noise - A Computer Program Incorporating Realistic Blade Motions and Advanced Acoustic Formulation,” NASA TM 87721, 1986.
8. Gandhi, F., and Tauszig, L., “Influence of Individual Interactions on Helicopter Blade-Vortex Interaction Noise,” Journal of the American Helicopter Society, Vol. 48, No. 4, pp. 287 – 299, Oct. 2003.
9. Charles, B. D., Tadghighi, H., and Hassan A. A., “Effects of a Trailing Edge Flap on the Aerodynamics and Acoustics of Rotor Blade-Vortex Interactions” 150 Proceedings of the 14th DGLR/AIAA Aeroacoustics Conference, Aachen, Germany, May 1992.
11. Malovrh, B., and Gandhi, F., “Sensitivity of Helicopter BVI-Induced Noise and Vibratory Loading to Variations in Individual Interaction-Parameters,” accepted for publication, Journal of Aircraft.
10. Dawson, S., Marcolini, M., Booth, E., Straub, F., Hassan, A., Tadghighi, H., Kelly, H., “Wing Tunnel Test of an Active Flap Rotor: BVI Noise and Vibration Reduction,” Proceedings of the 51st Annual Forum of the American Helicopter Society, Fort Worth, Texas, May 9-11, 1995, pp. 631-648.
12. Honert, H., van der Wall, B. G., Fritzsche, M., Niesl, G., “Realtime BVI Noise Identification from Blade Pressure Data,” Proceedings of the 24th European Rotorcraft Forum, Marseilles, France, 15-17 September, 1998, Paper No. AC08. 151
Chapter 4
Metrics for BVI noise
IBC inputs that would reduce BVI noise in one flight condition can increase it in others. Other inputs, such as reduction in bound circulation at the time of vortex generation, though they may reduce noise in all examined flight conditions, need to be tailored to an individual flight condition for maximum effectiveness. As the number of parameters that affect the flight condition, and therefore the effectiveness of a given IBC input, are large, open-loop control is deemed impractical in most cases. If closed-loop control is desired, though, there needs to be a feedback mechanism in place that can accurately gauge the resulting BVI noise.
Several different methods for feedback control have been considered. The most obvious metric is actual measured noise levels on an observer plane below the rotor.
While this can be useful to optimize an active control scheme for a given flight condition in a wind tunnel test [1, 2] it is obviously impractical on a real vehicle. Another study conducted by Zhang [3] utilized, in part, the measured perturbation of velocity around the airfoil. Again, this is not suitable for use on a flying rotorcraft.
Two practical candidates for onboard feedback of BVI noise on flying aircraft are skid microphones [4, 5] and blade pressure sensing [2, 6]. In this chapter the effectiveness of these two feedback methods to correlate with observed BVI noise levels is examined, and a new method for relating blade pressures to BVI noise is developed. 152 4.1 Skid Microphones as a Feedback Metric
Though earlier studies have shown that the measured noise at a microphone on a boom projecting far from the rotorcraft in the primary direction of BVI noise directivity is capable of correlating well with observer-plane noise [7, 8] a production rotorcraft will need a more feasible microphone location. Thus, the existing skid locations were chosen for microphones. A 2005 study on a 5-bladed MD-900 rotor was conducted using the rear of the skid as a microphone feedback location [9] for trailing-edge flap IBC. A 3 dB peak
BVISPL reduction on the advancing side was observed. Similar results were obtained for a 4-bladed BO-105 [10]. In neither case was an analysis done to determine if this corresponded to the minimum obtainable peak BVISPL or was simply a local minimum at the microphone location. A flight test on a Dauphin with microphones mounted on the exterior of the rotorcraft body (the Dauphin has no skids) showed that they could effectively predict the location of the source of the interaction [11]. It was not examined, however, if they were capable of acting as a metric of BVI noise.
In order to identify the effectiveness of skid microphones as a feedback source, the BVISPLs calculated at several theoretical microphone locations by WOPWOP are compared with the maximum advancing side BVISPL on an observer plane as well as the overall maximum BVISPL on that plane, also computed with WOPWOP. Five locations for microphones were chosen based upon the position of a BO105 skid scaled to the rotor. The rotor used is identical to that in Chapter 3. It has a 4 m diameter, a 12.1 cm chord, a 218 m/s tip speed, and a NACA 23012 airfoil with 4˚ linear twist. This skid is located 1.2 meters below the rotor and 0.5 meters from the centerline of the aircraft. 153 Microphone locations were spaced evenly every 0.3 meters from 0.4 meters behind the center of the rotor to 0.8 meters in front of it as shown in Figure 4.1. The observer plane for these comparisons is the same as that considered in previous chapters, 4 meters below the disk plane. Interference between the rotor and the fuselage is not modeled in the current study.
4
3
2
1 Skid Mics 1 2 3 0 4
X(m) 5 -1 Rotor Hub
-2
-3
-4 2 1 0 -1 -2 -3 -4 Y(m)
Figure 4.1: Skid Microphone Locations 154 The ability of skid microphones to track with observer-plane noise levels will be examined for changes in both flight condition and IBC input. The correlation is examined for:
1) Constant Shaft angle (3 deg backward tilt), changing advance ratio.
2) Constant Shaft angle (4 deg backward tilt), changing advance ratio.
3) Constant advance ratio (0.135), changing shaft tilt.
4) Constant flight condition (4 deg backward shaft angle, 0.135 advance ratio),
variable IBC input.
5) Constant flight condition (3 deg backward shaft angle, 0.17 advance ratio),
variable IBC input.
At each of these conditions the peak advancing side noise will be compared with the noise levels at each of the skid microphone locations as calculated by WOPWOP. The
BVI noise level will also be presented for both the observer plane and the skid microphone plane in each case to identify which results did or did not track well.
4.1.1 Variable Advance Ratio, 3 deg Backward Shaft Tilt
For the following case, the shaft angle was held at 3 degrees and only the advance ratio was changed. There were no IBC inputs in this case. In Figure 4.2, both the maximum advancing side BVISPL in the observer plane and the overall maximum
BVISPL is compared with the BVISPL at each of the six microphone stations. In Figures
4.3 to 4.7, the comparison is made with each of the microphone locations individually. In each figure, the primary Y axis relates to the observer plane BVISPL and the secondary 155 Y axis relates to the microphone plane BVISPL. Note that there is no need for the absolute values of the skid microphones to track with the observer plane noise, only for the trends to match in order for this to act as an effective feedback mechanism.
112
110
108 Max. Max. Adv. 106 Mic 1 Mic 2 c 104 Mic 3
BVISPL (dB) Mic 4 Mic 5 102
100
98 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 0.17 Advance Ratio
Figure 4.2: BVISPL at observer microphones for 3 degree backward shaft tilt and variable advance ratio
156
112 104.5
104 111 103.5 110 103 109 102.5 Max. Adv. 108 102 Mic 1
101.5 107
101 (dB) BVISPL Mic Skid 106 WOPWOP BVISPL (dB) 100.5 105 100
104 99.5 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 0.17 Advance Ratio
Figure 4.3: BVISPL at observer microphone 1 for 3 degree backward shaft tilt and variable advance ratio
112 106
111 105
110 104
109 103 Max. Adv. 108 Mic 2 102 107
101 Skid Mic BVISPL (dB)
WOPWOP BVISPL (dB) 106
105 100
104 99 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 0.17 Advance Ratio
Figure 4.4: BVISPL at observer microphone 2 for 3 degree backward shaft tilt and variable advance ratio
157
112 108
111 107 110
106 109 Max. Adv. 108 105 Mic 3
107 104 Skid Mic BVISPL (dB)
WOPWOP BVISPL (dB) 106 103 105
104 102 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 0.17 Advance Ratio
Figure 4.5: BVISPL at observer microphone 3 for 3 degree backward shaft tilt and variable advance ratio
112 110
111 109
110 108
109 107 Max. Adv. 108 c 106 Mic 4
107 105 SkidMic BVISPL (dB) 106 104 WOPWOP BVISPL (dB)
105 103
104 102 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 0.17 Advance Ratio
Figure 4.6: BVISPL at observer microphone 4 for 3 degree backward shaft tilt and variable advance ratio
158
112 111
111 110
110 109
109 108 Max. Adv. 108 107 Mic 5
107 106 SkidMic BVISPL (dB) 106 105 WOPWOP BVISPL (dB)
105 104
104 103 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 0.17 Advance Ratio
Figure 4.7: BVISPL at observer microphone 5 for 3 degree backward shaft tilt and variable advance ratio
It can be seen that none of the microphones tracked well with observer-plane data.
None capture the significant reduction in peak observer plane BVISPL seen at an advance ratio of 0.15, with most indicating a significant noise increase for this condition.
Also, when increasing advance ratio from 0.15 to 0.17, the maximum advancing side noise increases, but every skid microphone location shows a significant reduction over some portion of this range. The plots of BVISPL in both the observer and microphone planes are presented for each of the examined advance ratios in order to identify why the skid microphones were unable to follow the trend of the observer plane data. 159
1.5 4
B
d B B d 0 4 dB 2d 9 0 0 106 B 0 1 2 d 1 0 1 9 10 4 d dB d B B 2 B d 10 8 9 1 B 9 8 d 6 dB Skid Mics 10 dB 2 4 10 1 dB 6 B 10 0 d 11 B 0.5 2 d 2 11
dB 0 114 3 B
X(m) X(m) d 4 10 B 0 d 2 B 0 4 1 0d 0 B 1 d 8 9 5 -2 -0.5
9
6
d
B
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.8: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.135
1.5 4 B B d 94 d B 1 2 d dB 00 0 B 0 1 4 10 dB 0 1 6d dB 0 B 6 B 4 dB d 9 d 10 1 08 1 8 9 B 1 2 d 10 Skid Mics 106 dB dB 2 0 1 11
0.5 dB 2 2 11 B d B B 0 6 d d 0 2 3 1 0 X(m) 4 X(m) 1 11 0 dB 4 00 1 B B B d d d
6 6 8 9 9 1 -2 5 1 -0.5
B d 2 1 1 B 0d 1 1 -1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.9: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.140
160
1.5 4
B B B B B d d d d d 4 6 0 2 4 9 0 9 1 0 0 1 1 B B d d B d 8 B 6 9 d 0 08dB 98 2 1 1 0 1 Skid Mics 1 104 dB 2 1 110 dB 10 6 dB 1 0.5 2 04 dB B B 112 d d 6 0 1 3 0
X(m) X(m) B dB d 14 4 1 dB 0 0 8 1 10 B B 4 d d 0 2 0 0 1 1
11 -2 B 5 6 0 d -0.5 dB 11
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.10: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.145
1.5 4 B d B B B B 8 B d d d B 9 d 4 d d B 2 0 d 0 0 06 4 6 1 1 9 0 1 9 0 B 0 1 d B 1 d 08 1 8 9 B 1 d 2 Skid Mics 0 dB 2 1 0 1 11
dB 0.5 4 2 10
B d 0 4 0 3 1 X(m) X(m) 0 B 2 d 4 11 B d 6 dB 0 2 1 10 5 -2 100 dB B -0.5 114 d 98 dB 9 6 d B 9 4 d B -1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.11: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.150
161
1.5 4 B B d d B dB B 6 8 d 92 d 9 2 dB 9 B B 0 B 0 0 10 d 1 d d B 4 0 4 B d 0 1 0 d 0 4 B 1 1 6 9 9 d 0 1 6 9 dB 02 dB 1 1 6 Skid Mics 10 B dB 2 d 8 110 1 0 1
B 0.5 d 2 2 1 B 1 dB d 14 8 1 0 3 0 1 X(m) X(m) B d 0 6 dB 0 4 116 1 dB B 06 d 1 4 0 1 -2 B d 5 8 0 -0.5 1 B d 2 0 1
B d 0 0 -1 -4 1 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.12: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.155
1.5 4 B B d B d B 8 d d 0 9 0 2 B 9 0 B 9 d 1 d B 2 4 d 0 9 6 B 1 9 d 2 0 dB 1 B 8 B 1 d 9 B d 4 d 6 0 0 B Skid Mics 1 0 0 dB 1 d 1 2 4 0 108 1 1 B B d d 6 B 8 0 d 0 1 2 1 B 1 0.5 d 1 2 0 1 1 B d 0 14 3 1 dB X(m) 6 X(m) 11 0 4
B 4 d B 10 118 dB -2 d 5 6 0 -0.5 1 B 0 d B 10 d 2 0 1
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.13: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.160
162
1.5 4
B B B d d d B 2 8 0 d 9 B 9 0 2 dB d 1 0 94 0 1 96 dB 0 dB 1 4 10 B d B 1 4 8d B B 0 9 d d Skid Mics 1 B 2 06 B d 0 1 d 8 2 1 6 0 1 0 1 1 dB 8 B 0 d 1 2 B 1 d 1 B 0.5 0 d 2 1 0 1 11 dB dB 14 16 0 3 1 1 X(m) X(m) 0 B 8 d dB 4 11 6 10 B B d -2 d 06 8 1 5 0 B 1 d -0.5 B 02 d 1 2 0 1 dB 0 10
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.14: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.165
1.5 4 B 92 dB d B B B d d d B 6 6 8 0 B 9 0 d 9 9 1 2 00 d 0 1 B 1 dB d 94 B 4 d 0 B 8 1 1 d 9 B dB 04 d 8 Skid Mics 1 02 0 B 1 1 B d 2 B d 8 d 6 0 6 1 0 1 0 1 1 dB B d 10 2 1 B 1 0.5 d 1 2 0 1 1 B d B 6 d 1 0 4 1 3 1 X(m) 1 X(m) 0 dB 8 B 4 11 d 06 B 1 d 8 0 B 1 -2 d B 5 6 d 0 4 -0.5 1 10 1 0 2 1 d 0 B 0
d B -1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.15: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.170 163 The increase in noise seen for all microphones between an advance ratio of 0.145 and 0.15 while the peak observer plane noise decreases, as seen in Figures 4.10 and 4.11, appears to be function of directivity. The microphone plane advancing lobe in
Figure 4.11 widens while decreasing in magnitude as compared to Figure 4.10. This results in higher noise at the microphone locations, but lower peak noise. The problems that arise with skid microphones due to directivity are highlighted by comparing the 0.15 and 0.17 advance ratio cases (Figures 4.11 and 4.15). Between these cases it can be seen that the peak noise, as well as overall noise levels, in the observer plane increase. The peak advancing side noise in the skid microphone plane behaves similarly, but the directivity moves away from the skid microphones, resulting in a decreased noise level at the microphones.
4.1.2 Variable Advance Ratio, 4 deg Backward Shaft Tilt
As in the previous case, the shaft tilt is once again held constant, though this time at 4 degrees nose up, and the advance ratio is changed. Figure 4.16 compares the maximum advancing side noise level and overall noise level in the observer plane with the noise level from each of the skid microphones together, while Figures 4.17 through 4.21 compare the observer plane noise with each of the microphones individually. 164
114
112
110
108 Max. 106 Max. Adv. Mic 1 104 Mic 2 102 Mic 3
BVISPL (dB) 100 Mic 4 Mic 5 98
96
94
92 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 Advance Ratio
Figure 4.16: BVISPL at observer microphones for 4 degree backward shaft tilt and variable advance ratio
110.5 101
110 100 109.5 99 109
98 Max. 108.5 Adv. Mic 1 108 97
107.5 96 (dB) BVISPL Mic Skid
WOPWOP BVISPL (dB) 107 95 106.5
106 94 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 Advance Ratio
Figure 4.17: BVISPL at observer microphone 1 for 4 degree backward shaft tilt and variable advance ratio
165
110.5 101
110 100 109.5 99 109
98 Max. 108.5 Adv. Mic 2 108 97
107.5 96 (dB) BVISPL Mic Skid
WOPWOP BVISPL (dB) 107 95 106.5
106 94 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 Advance Ratio
Figure 4.18: BVISPL at observer microphone 2 for 4 degree backward shaft tilt and variable advance ratio
110.5 104
110 103
109.5 102
109 101 Max. 108.5 100 Adv. Mic 3 108 99
107.5 98 Skid Mic BVISPL (dB)
WOPWOP BVISPL(dB) 107 97
106.5 96
106 95 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 Advance Ratio
Figure 4.19: BVISPL at observer microphone 3 for 4 degree backward shaft tilt and variable advance ratio
166
110.5 108
110 106 109.5 104 109
102 Max. 108.5 Adv. Mic 4 108 100
107.5 98 SkidMic BVISPL (dB)
WOPWOP BVISPL (dB) 107 96 106.5
106 94 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 Advance Ratio
Figure 4.20: BVISPL at observer microphone 4 for 4 degree backward shaft tilt and variable advance ratio
110.5 109
110 108
109.5 107 106 109 105 Max. 108.5 Adv. 104 Mic 5 108 103 107.5 102 (dB) BVISPL Mic Skid WOPWOP BVISPL (dB) 107 101
106.5 100
106 99 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 Advance Ratio
Figure 4.21: BVISPL at observer microphone 5 for 4 degree backward shaft tilt and variable advance ratio 167 It is observed that the skid microphones all follow similar trends, and generally follow the observer plane data. Microphone 1 (Figure 4.17) has poorer correlation, but the others are able to capture the trends in much of the range considered. Specifically, all but microphone 1 predict the decrease in noise with an increase of advance ratio from
0.13 to 0.135. Only microphones 4 and 5 capture the continuous increase in noise from
0.135 to 0.15 advance ratio. Yet, every skid microphone was able to follow the decrease in noise from 0.15 to 0.155 advance ratio. All of the microphones identified the peak seen at and advance ratio of 0.15, but neither 1 nor 2 were able to identify the minimum at
0.135, with microphone 1 even predicting a local maximum. To better identify the causes for this, the BVISPLs in the microphone plane and observer plane are compared in
Figures 4.22 through 4.29.
1.5 4 B d B B B B d d B d d 2 4 B d 8 B 9 9 B 0 d 4 6 9 d B 6 d 0 4 9 9 0 d 9 1 0 0 2 1 1 0 B 1 d B 6 dB B d 0 98 d 6 0 1 2 1 1 B 0 B 1 Skid Mics d d 4 0 0 1 2 1 1 1 B d dB 8 108 B 0 dB d 1 B 4 4 d 1 0 0.5 2 1 1 2 1 1
0 2 dB 3 116 dB 10 X(m) X(m) 1 06 dB 0 100 4 dB dB 08 B 1 d 8 5 -2 9 -0.5
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.22: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.130
168
1.5 4 B B d d 90 96 B dB d 6 B dB 2 9 d 8 9 B 0 9 d 0 B 4 1 d 9 1 0 0 dB Skid Mics 1 dB 4 10 2 98 dB 1 02 102 dB B 1 d 6 0 1 106 dB 0.5 2 B d B 0 4d B 1 0 d 1 8 1 0 1 0 3 B
X(m) X(m) B d d B 6 d 0 4 2 1 0 0 11 1
4 B B d d 4 1 0 1 0 1 -2 B
B
5 d d
2 -0.5 8
0
9
1
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.23: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.135
1.5 4
B B B d d d B 6 B B d d d 0 2 6 9 0 9 9 9 8 0 9 1 B B d d B B 04 4 d d 1 B 8 0 d 9 9 0 1 B 4 1 d 0 2 1 B Skid Mics 10 B d 2 d 6 B 2 0 d 10 1 1 8 10 dB 6 0 B 1 d dB 0.5 2 8 2 B 1 0 d 1 1 0 1 1 0 3 B
X(m) d X(m) 4 dB 1 4 1 10 0 B B d dB 16d 2 2 1 0 B 10 4 1 d 4 0 1 100 dB -2 5 98 -0.5 dB
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.24: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.140
169
1.5 4 B B d B B d 8 d d 4 9 2 8 9 0 B dB 9 B 1 0 d 2 d 9 9 B B B 6 d d 9 d 00 6 1 0 B B 0 d 9 d 1 6 B B 4 1 B 0 d d 0 Skid Mics d 1 94 2 1 2 10 104 B d 1 B 6 d 0 8 dB B 0 1 10 d 11 08 0.5 2 1 B B d d 4 2 11 1 0 B 1 d
3 8 X(m) X(m) B 0 B d d 1 4 0 0 B B 0 0 106 d 1 1 6 d 11 1 4 08 B d B B B d d d 6 2 9 8 0
9 5 -2 1 -0.5
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.25: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.145
1.5 4 B B B B d B d B d 0 B d d d 0 d 0 0 0 6 8 9 1 B 9 0 9 B d 4 9 1 B d 4 9 9 d 10 B 4 4 6 d d 9 9 4 B B 0 B d 1 d 2 8 0 1 B B 9 1 d d Skid Mics 2 08 B 0 B 1 2 6 d 1 d 10 6 1 0 1 B B d d 12 0 1 1 0.5 1 08 dB 2 1
B 4 d 0 11 B 3 d X(m) X(m) 6 0 106dB 1 0 dB 16 1 B 1 08 d 4 dB 4 0 1 110 dB B d
2 -2 0 5 1 -0.5
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.26: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.150
170
1.5 4
B B B d B B d d d dB 2 8 90 B 4 8 0d d 9 B 8 B 9 9 0 8 d 1 d 9 6 4 B 9 d 9 2 dB 02 dB 9 1 B 4 B 1 d 10 d dB 2 6 0 Skid Mics 0 9 0 1 2 1 B 1 B d B d 4 dB d 8 0 6 6 0 1 0 0 1 1 1 0.5 2
B d 0 dB 0 1 2 B 3 1 11 d X(m) X(m)
6
0
1 0 B d
4 4 0 B dB 1 d 4 4 1 0 1 1 dB -2 6 B 5 1 6d B -0.5 1 10 d 2 0 1 B
d
0
0
1
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.27: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.155
1.5 4
B B B B d 92 d B d 8 d 4 d 0d B B B 9 d 4 9 d 9 6 2 9 2 9 9 0 B 1 d 0 B 0 d B 1 1 d B 96 dB Skid Mics d 0 4 B 4 2 0 0 d 0 1 1 8 1 B 1 9 B d d B 6 2 d 0 0 8 1 1 B 0 d 1 6 0.5 2 0 1 B d dB 1 0 2 0 1 8d 1 11 0 3 B X(m) X(m)
dB 0 4 11 4 B d B 04 B d 1 4 d 6 10 11 -2 5 B B d -0.5 d 02 6 1 0 1 B d 0 0 1
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.28: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.160
171
1.5 4 B d B dB 8 B B B d 4 8 d d d 2 9 0 2 6 9 9 9 9 B B dB d d 6 4 0 9 9 0 1 1 B B dB d d 104 Skid Mics 98 8 B 2 9 dB d 2 2 0 1 0 1 1 B B d d dB 0 6 0 6 10 1 B 0 d 1 0.5 2 4 0 dB 1 0 B 1 d 1 08 0 3 1 X(m) X(m) B 0 2d 11 dB 4 4 11
-2 dB 04 5 dB 1 -0.5 16 dB 1 2 10
100 dB
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.29: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.165
Analysis of the BVISPL maps shows that once again the failure of the skid microphones to capture noise trends is a result of directivity changes. Each microphone suggests that the noise at an advance ratio of 0.13 is significantly higher than that at 0.14, though both the peak overall noise levels and the advancing side noise levels in the observer plane are very close in value. Comparison of Figures 4.22 and 4.24 show that while the magnitude of the advancing side lobe is little changed, its directivity shifts away from the advancing side skid, thus decreasing the noise level at the skid microphones.
A failing of the microphones in this case is their inability to track the increase in noise between the 0.155 and 0.16 advance ratio cases. Comparing Figures 4.26 and 4.27 it is seen that though the peak noise increases between 0.155 and 0.16, the directivity 172 moves slightly away from the microphones, resulting in a lower recorded noise level from the microphones.
4.1.3 Variable Shaft Tilt, 0.135 Advance Ratio
Previous sections analyzed the ability of skid microphones to track with changes in noise level for changing advance ratio. In this section the advance ratio is held fixed at
0.135 while the nose down shaft tilt is varied. Figure 4.30 compares all of the skid microphones with the peak advancing side noise and overall peak noise on the observer plane while Figures 4.31 through 4.35 make the comparison with individual microphones.
110
108
106
104 Max. Max. Adv. 102 Mic 1 100 Mic 2 Mic 3 98 BVISPL (dB) Mic 4 96 Mic 5
94
92
90 3 3.5 4 4.5 5 5.5 6 Shaft Tilt (deg)
Figure 4.30: BVISPL at observer microphones for an advance ratio of 0.135 and variable shaft tilt.
173
111 106
110 104
109 102
108 100 Max. 107 98 Adv. Mic 1 106 96
105 94 Skid Mic BVISPL (dB)
WOPWOP BVISPL (dB) 104 92
103 90
102 88 33.544.555.56 Shaft Tilt (deg)
Figure 4.31: BVISPL at observer microphone 1 for an advance ratio of 0.135 and variable shaft tilt.
111 106
110 104
109 102
108 100 Max. 107 Adv. 98 Mic 2 106 96 105 Skid Mic BVISPL (dB) BVISPL Mic Skid
WOPWOP BVISPL (dB) 94 104
103 92
102 90 3 3.5 4 4.5 5 5.5 6 Shaft Tilt (deg)
Figure 4.32: BVISPL at observer microphone 2 for an advance ratio of 0.135 and variable shaft tilt.
174
111 108
110 106
109 104 108 102 Max. 107 Adv. 100 Mic 3 106 98 105 SkidMic BVISPL (dB) 96 WOPWOP BVISPL (dB) 104
103 94
102 92 33.544.555.56 Shaft Tilt (deg)
Figure 4.33: BVISPL at observer microphone 3 for an advance ratio of 0.135 and variable shaft tilt.
111 108
110 106 109
108 104 Max. 107 Adv. 102 Mic 4 106
105 100 Skid Mic BVISPL (dB)
WOPWOP BVISPL (dB) 104 98 103
102 96 3 3.5 4 4.5 5 5.5 6 Shaft Tilt (deg)
Figure 4.34: BVISPL at observer microphone 4 for an advance ratio of 0.135 and variable shaft tilt.
175
111 108
110 107
106 109 105 108 104 Max. 107 Adv. 103 Mic 5 106 102 105 101 (dB) BVISPL Mic Skid
WOPWOP BVISPL (dB) 104 100
103 99
102 98 3 3.5 4 4.5 5 5.5 6 Shaft Tilt (deg)
Figure 4.35: BVISPL at observer microphone 5 for an advance ratio of 0.135 and variable shaft tilt.
Once again, correlation between the skid microphones and both the overall peak observer plane noise levels and the advancing side observer plane noise levels is generally poor. Peak observer plane noise is seen to increase continuously from 3.5 to 5 degrees shaft tilt. Not one of the microphones is able to capture the increase in peak
BVISPL observed when increasing shaft tilt from 3.5 to 4 to 4.5 degrees. Only microphones 4 and 5 show an increase in noise level between 4.5 and 5 degrees shaft tilt.
To identify the causes for this, the BVISPL plots in the microphone plane and the observer plane are again compared. 176
1.5 4 B
B d B d d 4 4 2 0 6 dB 9 B 0 1 10 d B 0 B 1 10 d d
0 2 B 0 9 B d 1 d B 04 8 d 1 1 9 9 2 dB 6 dB 0 Skid Mics 08 1 1 2 1 B dB 0 d 6 11 10 B 0.5 2 d 2 11
dB 0 3 114 X(m) X(m) dB 04 0 1 B d 4 2 0 B 1 B d d 8 0 9 0 1 -2 B 5 d -0.5 6 9
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.36: BVISPL in a) Skid microphone plane and b) Observer plane for 3 degree backward shaft tilt and advance ratio of 0.135
1.5 4 B B d d B B 8 d 2 d 9 0 10 9 0 2 dB 8 1 dB 2 9 9 B B d d 4 94 9 dB 1 B 02 d 1 Skid Mics 6 2 9 B d 1 0 0 1 B B d d 4 4 0 0.5 10 1 2 B d 8 B 0 d 1 6 0 0 3 1 B X(m) dB X(m) d 12 B 1 4 d 0 0 0 1 1 1 B 4 d 2 0 1 B B d B 4 d 1 -2 4 8d 1 0 9 5 1 B -0.5 d 0 0 1
B
d
6
9
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.37: BVISPL in a) Skid microphone plane and b) Observer plane for 3.5 degree backward shaft tilt and advance ratio of 0.135
177
1.5 4 dB B 0 B dB d 9 d 6 8 0 9 B B 0 9 d 6 d 1 2 9 9 B B 1 d d 00 94 8 1 dB 9 dB Skid Mics 104 2 B 2 d 1 10 102 dB B d 6 106 dB 0 0.5 B 1 2 d 4 B 0 d 1 0 B 1 d 1 8 0 0 3 1 B X(m) X(m) d B 6 d 0 2 1 B 0 1 d 1 2 B 0 1 4 4d 11
B B
d d -2 B 4 0 d
5 0 0 8 1 -0.5 1 9
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.38: BVISPL in a) Skid microphone plane and b) Observer plane for 4 degree backward shaft tilt and advance ratio of 0.135
1.5 4 8 6 dB B d B B d B 94 d d 8 6 B 9 0 8 0 B 1 d 2d 4 9 0 96 dB B 1 d dB B 0 8 d 1 9 9 2 0 Skid Mics dB 1 B 2 4 d 9 0 B 1 0 d B 1 d 4 B 0 d 8 1 9 4 0 B 1 d 0.5 B B 2 2 d d 0 8 8 1 0 0 1 1 0 B 3 d X(m) 6 X(m) 0 B 1 d 2 0 1 B 1 d B d 8 6 B 0 4 B 0 d 1 d 1 B 0 6 d 1 1 2 1 1 0 1 -2 5 B d B B d d -0.5 4 B 1 4 0 d 1 0 0 8 1 1 9
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.39: BVISPL in a) Skid microphone plane and b) Observer plane for 4.5 degree backward shaft tilt and advance ratio of 0.135
178
1.5 4 dB B B 88 B d d d 2 B 0 8 9 d 9 B 9 B 6 d d 9 2 2 B 9 0 d 1 4 dB dB 9 B 94 0 B d 10 d 1 0 06 0 B 1 Skid Mics 1 96 d B d B 2 dB 4 d 0 1 8 4 1 9 0 1 B d B 2 d 0 8 B 1 0 d 0.5 1 2 8 B 0 B d 1 d 2 6 1 0 1 1 B 0 3 0d X(m) 1 X(m) 1 B d dB 0 B 8 6 B d 0 0 d 6 1 1 4 1 1 1 4 1
-2 B 5 d 4 -0.5 0 1 B
B d
d 0
2 0
0 1
1
-1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.40: BVISPL in a) Skid microphone plane and b) Observer plane for 5 degree backward shaft tilt and advance ratio of 0.135
1.5 4 9 B 2 d d B B d 2 9 9 4 0 dB 9 B 9 d 6 6 dB B 9 d 4 1 Skid Mics 9 98 dB 2 B dB B d B 98 1 d 00 00 d 8 1 1 9 102 dB 100 0.5 dB 2 dB 4 10 B d B 0 106 2 d 1 3 10 4 dB 0 2 X(m) 10 X(m) d 0 B 1 4 08 dB
dB 6 0 dB 1 8 -2 5 10 -0.5
B d 0 1 1 -1 -4 0 -0.5 -1 -1.5 -2 2 1 0 -1 -2 -3 -4 Y(m) Y(m)
Figure 4.41: BVISPL in a) Skid microphone plane and b) Observer plane for 6 degree backward shaft tilt and advance ratio of 0.135 179 Comparing Figures 4.36a through 4.39a it can be seen that there is a an excursion from the primary advancing side lobe that overlaps the advancing side skid. As the nose up shaft tilt increases from 3 to 5 degrees, (Figures 4.36 to 4.39) this secondary lobe weakens until it is no longer visible. As the lobe directly overlaps the advancing side skid, it is the strength of this secondary lobe that the skid microphones are measuring, not the primary lobe. Thus, the observed noise at the microphones decreases over this entire range, regardless of what happens to the primary lobe. It is only when the secondary lobe is eliminated that the microphones again begin to track with the primary lobe, and thus reasonably correlate with observer plane data.
4.1.4 Constant Flight Condition (4 deg backward Shaft Tilt, 0.135 Advance Ratio), Variable IBC Input
While earlier sections of this chapter examined the ability of skid microphone to track with observer plane BVI noise for changes in flight condition, the present section maintains a constant flight condition and varies the strength of an IBC input. For this scenario the flight condition was held at a 4 degree nose up shaft tilt and 0.135 advance ratio while a localized IBC input (as examined in Chapter 3) of varying magnitude centered at 140 deg azimuth and extending for 40 deg azimuth was added. The input holds to its max value for half of its duration. The first and last quarters of its duration are spent ramping up and down to and from the peak value, as discussed in section 3.3. As seen in Chapter 3, this input reduces the strength of the interacting vortex, thus reducing
BVI noise. 180 Figure 4.42 compares the maximum advancing side noise and overall maximum noise in the observer plane with the BVISPL at each of the skid microphones while the strength of the IBC input varies from 0 to -3 degrees. Individual microphones are examined in Figures 4.43 through 4.47.
110
108
106 Max. 104 Max. Adv. Mic 1 102 Mic 2 Mic 3
BVISPL (dB) 100 Mic 4 Mic 5 98
96
94 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.42: BVISPL at observer microphones for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input.
181
107 99.5
106.5 99 106
105.5 98.5
105 98 Max. Adv. 104.5 Mic 1 97.5 104
103.5 97 (dB) BVISPL Mic Skid WOPWOP BVISPL (dB) 103 96.5 102.5
102 96 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.43: BVISPL at observer microphone 1 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input.
107 99
106.5 98.5 106
105.5 98
105 97.5 Max. Adv. 104.5 Mic 2 97 104
103.5 96.5 (dB) BVISPL Mic Skid WOPWOP BVISPL (dB) 103 96 102.5
102 95.5 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.44: BVISPL at observer microphone 2 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input.
182
107 100
106.5 99.5
106 99 105.5 98.5 105 Max. 98 Adv. 104.5 Mic 3 97.5 104 97 103.5 (dB) BVISPL Mic Skid
WOPWOP BVISPL (dB) 96.5 103
102.5 96
102 95.5 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.45: BVISPL at observer microphone 3 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input.
107 99.5
106.5 99
106 98.5 105.5 98 105 Max. 97.5 Adv. 104.5 Mic 4 97 104 96.5 103.5 Skid Mic BVISPL (dB) WOPWOP BVISPL (dB) 103 96
102.5 95.5
102 95 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.46: BVISPL at observer microphone 4 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input.
183
107 100.5
106.5 100 106
105.5 99.5 105 Max. Adv. 104.5 99 Mic 5
104 98.5 103.5 Skid Mic BVISPL (dB) WOPWOP BVISPL (dB) 103 98 102.5
102 97.5 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.47: BVISPL at observer microphone 5 for an advance ratio of 0.135, 4 degree backward shaft tilt, and variable IBC input.
As the magnitude of the IBC input increases, the maximum advancing side noise in the observer plane steadily decreases. Microphones 1-4 generally show a much sharper decrease in noise for low to moderate input magnitudes, but actually increase in noise going from an IBC input of -2 to -3 degrees. Only microphone 5 appears to track with the maximum advancing side noise in the observer plane with increasing IBC input magnitude.
4.1.5 Constant Flight Condition (3 deg backward Shaft Tilt, 0.17 Advance Ratio), Variable IBC Input
A second case of varying the IBC input while holding a constant flight condition was considered. The input profile is once again centered at 140 degrees azimuth, runs over a 40 degree range, and is of the ramped input type described in Chapter 3. In this 184 case the shaft tilt is 3 degrees nose up and the advance ratio is 0.17. Figure 4.48 compares all of the skid microphones with the maximum advancing side and overall maximum observer plane noise, while Figures 4.49 through 4.53 break down the comparison to individual microphones.
112
110
108
106 Max. Max. Adv. 104 Mic 1 102 Mic 2 Mic 3 100 BVISPL (dB) Mic 4 98 Mic 5
96
94
92 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.48: BVISPL at observer microphones for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input.
185
112 100.5
111 100 110 99.5 109
99 Max. 108 Adv. Mic 1 107 98.5
106 98
WOPWOP BVISPL (dB) 105 97.5 104
103 97 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.49: BVISPL at observer microphone 1 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input.
112 101
111 100 110 99 109
98 Max. 108 Adv. Mic 2 107 97
106 96 (dB) BVISPL Mic Skid
WOPWOP BVISPL (dB) 105 95 104
103 94 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.50: BVISPL at observer microphone 2 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input.
186
112 103.5
111 103
110 102.5 109 102 Max. 108 Adv. 101.5 Mic 3 107 101 106 Skid Mic BVISPL (dB) 100.5 WOPWOP BVISPL(dB) 105
104 100
103 99.5 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.51: BVISPL at observer microphone 3 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input.
112 104.5
111 104
110 103.5 109 103 Max. 108 Adv. 102.5 Mic 4 107 102 106 Skid Mic BVISPL (dB) BVISPL Mic Skid
WOPWOP BVISPL (dB) 101.5 105
104 101
103 100.5 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.52: BVISPL at observer microphone 4 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input.
187
112 104
103.5 111 103 110 102.5 109 102 Max. 108 101.5 Adv. Mic 5 107 101 100.5 106 Skid Mic BVISPL (dB) BVISPL Mic Skid 100 WOPWOP BVISPL (dB) 105 99.5
104 99
103 98.5 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC Input (deg)
Figure 4.53: BVISPL at observer microphone 5 for an advance ratio of 0.170, 3 degree backward shaft tilt, and variable IBC input.
Again, increasing the magnitude of the IBC input from 0 to -3 degrees is seen to steadily decrease the peak advancing side noise. Each skid microphone is able to follow this trend from 0 to -2 degree input, but microphones 1 and 2 predict an increase in noise in the -2 to -3 degree input range, while microphone shows little change. The skid microphones do a better job of tracking observer plane noise in this case. This is the only case considered where some of the microphones are able to follow the observer plane noise trends. This is somewhat expected as the IBC input results in a large change in the magnitude of the BVISPLs, but without the significant change in directivity that appears to have prevented the microphones from tracking well in earlier cases. 188 4.1.6 Summary of Skid Microphone Feedback
From the results in sections 4.1.1 through 4.1.5, it is seen that skid microphones tend to track quite poorly with changes in the noise level in an observer plane 4 meters below the disk. This was found to be due to three major sources:
1. Directivity changes. Subtle differences in the BVI event can result in changes in
the directivity of the advancing side lobe. When this lobe moves away from the
microphones, they detect a drop in BVISPL, even if the strength of the lobe, and
thus the noise in the observer plane, increased.
2. Secondary lobes. On occasion the BVI event can produce weaker secondary lobes
with a different directivity than the primary lobe. If one of these happens to fall
upon the skid microphones, they will be more heavily influenced by this lobe than
the primary one. Since the two lobes do not necessarily grow or weaken together,
the skid microphones may produce erroneous results.
3. Interference. The multiple BVI events happening on the rotor can interfere
constructively or destructively with each other. The magnitude of these effects
can be large enough that they overwhelm the overall changes in BVI noise. When
the interference falls on the skid microphone locations it can cause them to track
poorly with observer plane data.
It must be noted that the above observations hold primarily for microphones on the skids of the rotorcraft. If they were positioned on masts in the path of the primary 189 BVI noise lobe it is likely that their ability to track with observer plane data would improve dramatically. Such microphones would be of little practical use though. The present study also did not examine the effect that the fuselage itself would have on the ability of skid microphones to track with observer plane noise.
4.2 Blade Pressures as a Feedback Metric
A second potential feedback mechanism for active BVI noise control is blade pressure measurements [6, 8]. Here, dynamic pressure sensors mounted on the blade surface are used to predict the resulting BVI noise in the observer plane. Existing methods [6] require a priori knowledge of the wake geometry in the disk plane and using empirical weighting factors (to emphasize parallel events and marginalize oblique events), in conjunction with blade pressure data. Such a method does not account for possible changes in the wake geometry due to active control input.
4.2.1 First Quadrant RMS
The goal of the feedback metric is to be able to correlate a measured quantity with observed BVI noise in real time. Thus, a simple method was desired to minimize computing time. An initial choice was to simply calculate the RMS of the 6th – 40th harmonics of the blade passage frequency of the blade loading time history
(representative of blade pressures) at a given radial blade station within a region of interest (the first quadrant for advancing side interactions). Though this produced some 190 positive initial results as seen for the 3 degree nose down shaft tilt case in Figure 4.54 for radial stations of 0.7 R, 0.8 R, and 0.9 R, no single station was capable of tracking with the resulting noise data through all conditions. This is expected, as the radial position of the primary interaction changes based on flight condition. Combinations of RMS data at different radial stations with different weighting factors also failed to track well with observer plane data. It was determined that for a simple analysis of the blade loading time history to yield accurate results it would need to include external information as to the general wake structure in order to be able to properly weight the influence of each radial station. This was determined to make these methods unfit as a feedback metric for IBC.
Figure 4.54: First quadrant RMS tracking with maximum advancing side noise.
191 4.2.2 BPAP
The attempt to use a simple RMS of the blade loading time history fails to accurately track with the maximum BVISPL of the advancing side because, though it takes into account the impulsivity of the blade-vortex interaction, it does not take into account the nature of the interaction. It does not take into account how parallel the interaction is, whether the interaction begins inboard and progresses outboard or begins outboard and progresses inboard, or how the several advancing side interactions combine to produce the resulting observer-plane noise. Though earlier work has taken these parameters into account when using blade pressures as a feedback mechanism [6], they utilized a computed wake structure to gauge their effect. As the shape of the wake is not information available in a rotorcraft in flight, another method was sought for this study.
Rather than developing a new noise metric from simple terms, it was determined that a more direct technique would be to take the existing method for translating blade pressure measurements into predicted BVISPLs, WOPWOP, and simplify it until it was fast enough to run in real-time and required as inputs only that information available in the rotorcraft in flight. Thus, the peak noise level could be extracted from a calculated noise map based on measured blade pressures, eliminating the potential of missing the peak location as it changed due to changing flight condition. 192 4.2.2.1 BPAP Model Development
The development of the Blade-Pressure Acoustic Prediction (BPAP) code begins with Farassat’s formulation 1A of the Ffowcs Williams and Hawkings equation [12], the basis for WOPWOP [13], as seen in Equation 4.1:
pX ,t pT X ,t pL X ,t pQ X ,t 4.1
Here p is the acoustic pressure at the observer location, X is the observer
location, and t is the time of observation. The acoustic pressure is broken up into the
thickness source, pT , loading source, pL , and quadrupole source, pQ . As BVI noise is
an effect of impulsive loading on the blade, only the loading source is of concern, and the
other two terms are ignored. This reduces Equation 4.1 to Equation 4.2
pX ,t pL X ,t 4.2
The loading noise term, shown in Equation 4.3 [13], itself, can be significantly
simplified for the current investigation.
1 L L L 4 p X ,t r dS r M dS L 2 2 2 c f 0 r1 M r f 0 r 1 M r ret ret 4.3 1 L dM cM M 2 r r r dS c r 2 1 M 3 f 0 r ret
In Equation 4.3, c is the freestream speed of sound, Lr is the component of loading
due to pressure in the direction of sound radiation (from the source to the observer), Lr is the time derivative of Lr, r is the distance from the source to the observer, M is the local
Mach number at the source (not the local fluid Mach number), Mr is the component of the 193 local Mach number at the source in the direction of sound radiation, M r is the time derivative of Mr, and LM is the dot product of the loading vector and Mach number vectors at the source. The term f=0 indicates that the integral is to be carried out over the surface of the blade and ‘ret’ indicates that the terms inside the brackets are to be calculated at the retarded time when the sound waves leave the source rather than the observer time, t.
As BVI noise is a function of the impulsive loading of a blade-vortex interaction, it is only those terms related to this impulsivity, those that are a function of the rate of change of blade loading, that are of concern. Thus, those terms related to the loading on the blade itself, Lr, are not critical to characterizing BVI noise, while those that are
dependent on the rate of change of the blades loading, Lr , are. Eliminating the unimportant terms leaves the much simplified Equation 4.4.
1 Lr 4 pX ,t 4 pL X ,t dS 4.4 c r 1 M 2 f 0 r ret
To further simplify the equations, it is noted that at any radial, i, the changes in r and Mr around the blade are small with respect to their base values. Thus, they may be considered constants for a given time and radial station and removed from the integral.Note that both of these variables still remain functions of time. If only a single radial station is considered, then the integral is no longer carried out over the surface of the blade (dS), but around the blade in the chordwise direction (dc) and the result multiplied by the radial length of the considered section of the blade (di). This results in 194 Equation 4.5, which gives the influence of a single radial station of the blade on the final observer acoustic pressure.
d 4 pX ,t i L dc 4.5 i cr 1 M 2 r i i r i ret f 0 ret
The remaining integral calculates the portion of the time derivative of loading at the considered blade station in the radiation direction. This, multiplied by the radial length of the considered section of the blade (di) results in the time derivative of the net
force on the blade section in that direction, Fr i . The summation of these elements over the length of the blade results in the overall acoustic pressure due to that blade as shown in Equation 4.6.
n F r i 4 pi X ,t 2 4.6 i1 cr 1 M i r i ret
The noise propagation equation has now been reduced to a form only requiring knowledge of the relative positions and motions of the blade and observer, and the time derivative of the loading on the blade element. For a rotorcraft in flight, the loading may be estimated from pressure measurements on the blade’s surface, while in the present study it is supplied by the comprehensive code used, as described in section 3.1. An observer plane is chosen that moves with the rotorcraft rather than fixed to the ground, one fixed in the hub coordinate system. This choice allows for a single projection of the
BVI noise results over the observer plane rather than one that changes as the rotorcraft passes. With the observer fixed, all that remains is to determine the blade position and motion relative to the observer. 195 The blade position and motion depend on not only the forward speed of the rotorcraft and rotational speed of the rotor, but also the flap, lag, and pitch motions and the blade bending and torsion. As the effect of flap, lag, pitch, bending, and torsion on blade position and velocity is both small relative to the effects of rotor rotation and forward motion of the rotorcraft, and not easily identifiable from within the flying rotorcraft, they are ignored in the creation of the present metric. Thus, the position and velocity of the blade are now dependent purely upon the forward speed of the rotorcraft,
V, the rotational speed of the rotor, ω, and the azimuthal position of the blade, ψ. The only additional piece of instrumentation necessary for the rotorcraft, beyond the blade pressure sensors, is an encoder to correlate measured blade pressures with azimuthal angle.
Still remaining is the calculation of retarded time. Note that Equation 4.6 still requires that the math be carried out at the time of the source, not the time of the observer. As the blades move relative to the observer, the time it takes for the sound to reach the observer, and thus amount of time the calculation should be retarded from the observer time, changes. WOPWOP starts with the observer and then calculates back in time to find the position and time delay for the source [13]. The present analysis uses a forward-in-time method similar to that found in codes developed by Xue [14], Leishman
[15], and Brentner [16].
The BPAP methodology follows a three step process. As an input to the code, a comprehensive code is used to march a reference blade around the azimuth and calculate the time derivative of loading at each time step for each radial station. (This information would be gathered from blade pressures and an encoder signal on a real rotorcraft.) As 196 the advancing side interactions are limited to the first quadrant, only this portion of the azimuth is critical to identifying advancing side BVI noise. Thus, BPAP only uses first quadrant loading data when calculating BVISPLs. First, Equation 4.6 is evaluated at each of these positions and times in source time. At each of these times and postitions, a delay, the time it would take for the pressure wave to travel from this point to the observer, is also calculated. This delay is a function of the relative positions of the source and observer and the forward speed of the aircraft. Second, for each of these points the time delay is added to the time index, thus moving the signal from source time to observer time. Note that the time index is no longer evenly sampled in time, and the signal from each radial station will arrive at the observer location over a different range of time. This necessitates step three, linearly interpolating the data onto an evenly discretized time signal. This process results in a time history of acoustic pressure at the observer location for each azimuthal station, as seen in Figure Figure 4.55. Here, the horizontal lines represent the time over which a signal from a given radial station arrives at the observer location. The start of observer time corresponds to the point at which the signal from the first radial station when the blade was 0 degree azimuth arrives at the observer location.
Though the signal is generated over a full quarter-period (when the blade moves from 0 to 90 degrees azimuth), it is shown that the signal takes up less than a full quarter-period in observer time. This is due to the blade moving towards the observer location. Were an observer position chosen such that the blade was moving away from it, the range over which the signal arrived at the observer location would extend longer than a quarter- period. Only one quarter-period of observer time is considered as a 4-bladed rotor will repeat this signal 4 times per revolution. As seen in the radial station 2 line, the signal 197 arrives over a different range of observer time than that from station 1. As the distance from station 1 to the observer and station 2 to the observer are different, the resulting time delay is different, and the two signals are offset in observer time. The radial station 3 signal is offset enough that the data continues out past a quarter-period in observer time.
One quarter-period is subtracted from the time index of this remaining data, causing it to wrap back around to 0 observer time. This may be thought of as thought of as section A coming from one blade in the latter portion of the first quadrant with the signal cutting off when that blade reaches 90 degree azimuth. Section B would then originate with the next blade as it passes through 0 degrees azimuth and marches through most of the first quadrant before the end of recorded observer time. The relative observer times over which signals from various radial stations arrive is highly dependent on the observer location itself, with the signal from one station arriving first at one observer location, while the signal from a different observer location may arrive first at another observer location. 198
Source Time ψ = 0° 1
2
Radial Station A B 3 Source Time ψ = 90°
0 ¼ Period Observer Time
Figure 4.55: Relative arrival times at observer location for different radial sources.
These time histories are summed based on their time index and run through an
FFT to convert them to the frequency domain. As with the analyses in earlier chapters that utilized the full version of WOPWOP, the energies within the sixth through fortieth harmonics of blade passage are summed and then converted to a BVISPL at the observer location. This process is repeated for each observer location on a defined observer plane, giving a map of the resulting BVI noise. A peak BVISPL is then extracted from this map.
For the present study, BPAP analyzes only the portion of the observer plane that has been seen to include the primary peak advancing side noise. This area ranges from -1 to 2 meters from the hub in the x direction and -1 to -3 meters from the hub in the y direction with a resolution of 0.25 meters.
This method can be sped up further by reducing the number of radial stations and the portion of the azimuth examined. Each radial station in the summation in 199 Equation 4.6 represents another set of pressure sensors which add complexity and weight to the blade. It is thus desired to minimize them while still obtaining enough data to accurately correlate blade pressures to observer-plane noise. While the comprehensive code (and thus WOPWOP) uses fifteen radial stations, only four are used for BPAP, those at the 50%, 63%, 77%, and 90% blade stations. This covers the critical range over which BVI events have been shown to occur (Chapter 2). One result of the elimination of so many terms in the developed equations, is that the BPAP predictions fall significantly below those of WOPWOP in absolute value. As this metric is not designed to predict exact noise levels, but rather to accurately relate the changes in those levels due to changing flight conditions, IBC inputs, etc, this is not seen as a problem. The values predicted by BPAP could be related to a second Y axis in the following graphs as was done for skid microphone predictions, but for ease reading a value of 7.5 dB was simply added to all BPAP predictions. This puts the resulting BVISPLs from BPAP of a similar magnitude to those from WOPWOP itself, thus allowing for a single Y axis.
4.2.2.2 BPAP Results
The ability of the BPAP metric to track with observer plane noise was examined for a variety of flight conditions and IBC inputs. As in earlier cases, the observer plane is located 4 meters below the disk plane. The same conditions examined in section 4.1 with skid microphones were reconsidered with blade pressure data utilizing BPAP:
1) Constant Shaft angle (3 deg), changing advance ratio. 200 2) Constant advance ratio (0.135), changing shaft tilt.
3) Constant flight condition (4 deg shaft angle, 0.135 advance ratio), variable IBC
input.
4) Constant flight condition (3 deg shaft angle, 0.17 advance ratio), variable IBC
input.
Figure 4.56 shows the results for a constant 3 degree shaft tilt while varying the advance ratio between 0.13 and 0.17. The metric is seen to follow well with peak advancing side noise. As the metric only utilizes data from the first quadrant it is expected that it would track far better with the peak advancing side noise than with the overall peak BVISPL as is seen. Comparing Figure 4.56 with Figures 4.2 through 4.7 shows that the BPAP data tracks dramatically better with the maximum advancing side noise than skid microphones.
113
112
111
110
109 Metric
108 Max Adv.
107 BPeak VISPL (dB)
106
105
104 0.13 0.135 0.14 0.145 0.15 0.155 0.16 0.165 0.17 Advance Ratio
Figure 4.56: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 3 backward degree shaft tilt and variable advance ratios. 201 Next, the shaft tilt was varied while holding the advance ratio constant at 0.135.
Figure 4.57 shows that once again the metric tracks very closely with the peak advancing side noise. Between shaft angles of 3.0 and 3.5 degrees it is seen that the metric does increase in value while the peak advancing side decreases. Despite this, BPAP is seen to track with the observer-plane noise far better than the skid microphones seen in Figures
4.30 through 4.35.
110
109
108
107 Metric
Max Adv. 106
105 Peak BVISPL (dB)
104
103 33.544.555.56 Shaft Tilt (deg)
Figure 4.57: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 0.135 advance ratios and variable shaft tilt.
The metric was next examined with an IBC input. As in sections 4.1.4 and 4.1.5, this input consists of a 50% peak value ramp up/ramp down signal centered at 140 degrees azimuth varying from zero to three degrees in amplitude. This flight condition is an advance ratio of 0.135 and a shaft tilt of four degrees. Figure 4.58 shows that the metric again tracks very well with the advancing side noise.
Comparing Figure 4.58 to Figures 4.42 through 4.47 it is seen that BPAP once again does a superior job of tracking with observer plane noise. Of all the skid 202 microphones, only microphone 5, shown in Figure 4.47, does not correctly predict the peak observer plane noise increasing when the amplitude of the IBC input goes from -2 to -3 degrees, yet BPAP compares favorably even to microphone 5.
108
107
106
105 Metric
Max Adv. 104 Peak BVISPL (dB)
103
102
101 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC input (deg)
Figure 4.58: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 0.135 advance ratios, 4 degree backward shaft tilt, and variable IBC input.
Next, the effect changing the magnitude of the same IBC input is considered for a flight condition of 0.170 advance ratio and 3 degree shaft tilt. Figure 4.59 shows that
BPAP does an even better job of tracking the peak observer-plane advancing side noise in this condition than it did in Figure 4.58. Comparing these results with Figures 4.48 through 4.53 it is seen that only microphone four managed to track with peak advancing- side observer plane noise as well as BPAP. But, when Figure 4.46 is reconsidered it is seen that the ability of this microphone to follow BVI noise levels is highly dependent upon the flight condition whereas BPAP is able to accurately track peak advancing side
BVISPLs due to changes in IBC input for all considered flight conditions. 203
113
112
111
110
109 Metric 108 Max Adv. 107 Peak BVISPL (dB) BVISPL Peak 106
105
104
103 0 -0.5 -1 -1.5 -2 -2.5 -3 IBC input (deg)
Figure 4.59: BPAP tracking with peak BVISPL and peak advancing side BVISPL for 0.170 advance ratios, 3 degree backward shaft tilt and variable IBC input.
As the BPAP metric does an excellent job tracking with peak advancing side observer plane noise due to changes in IBC, it would likely be an excellent choice to act as a feedback method for closed loop IBC noise reduction.
4.2.3 Summary Blade Pressure Feedback
Simple metrics based on blade pressure feedback were unable to follow the trends in data of observer plane peak advancing side BVI noise levels. A new metric, BPAP, based on the WOPWOP code was shown to track well with peak advancing side noise in a variety of flight conditions and IBC inputs. The relatively simple calculations necessary for BPAP allow for a fast computation time, making it feasible as a real-time IBC feedback metric. Though BPAP was developed with the intent of utilizing it as a metric for BVI noise in closed-loop IBC applications, its ability to track with noise in a range of 204 flight conditions suggests that it would serve equally well in aiding in the flying of noise- abatement flight paths. Also, as the metric is based on the actual pressures seen on the blade, this metric should be capable of taking into account variations in the aerodynamic environment such as wind gusts that other methods might not.
References
1. Kube, R., Achache, M., Niesl, G., and Splettstoesser, W. R., “A Closed Loop Controller for BVI Impulsive Noise Reduction by Higher Harmonic Control,” Proceedings of the 48th Annual Forum of the American Helicopter Society, Washington DC, June 3-5, 1992, pp. 819-842.
2. Nguyen, K., Betzina, M., Kitaplioglu, C., “Full-Scale Demonstration of Higher Harmonic Control for Noise and Vibration Reduction on the XV-15 Rotor,” Proceedings of the 56th Annual Forum of the American Helicopter Society, Virginia Beach, Virginia, May 2-4, 2000.
3. Zhang, M. M., Cheng, L., Zhou, Y., “Closed-Loop Controlled Vortex-Airfoil Interactions,” Physics of Fluids, v 18, n 4, April 2006.
4. Bebesel, M., Roth, D., Pongratz, R., “Reduction of BVI Noise on Ground – In- Flight Evaluation of Closed-Loop Controller,” Proceedings of the 28th European Rotorcraft Forum, Bristol, United Kingdom, Sept. 2002, pp. 19.1-19.9.
5. Patt, D., Liu, L., Friedmann, P. P., “Helicopter Noise Reduction by Actively Controlled Flaps,” Proceedings of the 26th AIAA Aeroacoustics Conference, Monterey, California, May, 2005.
6. Honert, H., van der Wall, B. G., Fritzsche, M., Niesl, G., “Realtime BVI Noise Identification from Blade Pressure Data,” Proceedings of the 24th European Rotorcraft Forum, Marseilles, France, 15-17 September, 1998, Paper No. AC08.
7. You, J. Y., Kwon, O. J., and Han, Y. O., “Viscous Flow Simulation of Rotor Blades with Tip Slots in Hover,” Journal of the American Helicopter Society, Vol. 54, No. 1, Jan. 2009.
8. Ishii, H., Gami, H., and Okuno, Y., “Helicopter flight tests for BVI noise measurement using an onboard external microphone,” AIAA Atmospheric Flight Mechanics Conference, San Francisco, CA, Aug. 2005. 205 9. Patt, D., Liu, L., and Friedmann, P. P., “Active Flaps for Noise Reduction: A Computational Study,” Proceedings of the 61rst Annual Forum of the American Helicopter Society, Grapevine, TX, June, 2007.
10. Patt, D., Liu, L., and Friedmann, P. P., “Active Flaps for Noise Reduction: A Computational Study,” Journal of the American Helicopter Society, Vol. 51, pp. 127-140, Apr. 2006.
11. Blacodan, D., Elias, G., Prieur, J., and Papillier, D., “Noise Source Localization on a Dauphin Helicopter in Flight,” Journal of the American Helicopter Society, Vol. 49, pp. 425-435, Aug. 2004.
12. Farassat, F., and Succi, G., “The Prediction of Helicopter Rotor Discrete Frequency Noise,” Vertica, Vol. 7, (4), 1983, pp. 309-320.
13. Brentner, K. S., “Prediction of Helicopter Rotor Discrete Frequency Noise - A Computer Program Incorporating realistic Blade Motions and Advanced Acoustic Formulation,” NASA TM 87721, 1986.
14. Advanced Rotorcraft Technology, Inc., “Kirchhoff Code-a Versatile CAA Tool,” NASA SBIR Phase I Final Report, contract NASI-20366, June 1995.
15. Leishman, J. G., “Aeroacoustics of 2D and 3D Blade-Vortex Interaction using the Indicial Method,” Proceedings of the 52nd Annual Forum of the American Helicopter Society, Washington, DC, June, 1996.
16. Brentner, K. S., Perez, G., Bres, G. A., Jones, H. E., “Maneuvering Rotorcraft Noise Prediction,” Journal of Sound and Vibration, Vol 39, pp 719-738, 2003.
206
Chapter 5
Conclusions and Recommendations for Future Work
While IBC has previously been examined for BVI reduction, this dissertation approached the problem from a new direction. Rather than parametrically varying a few parameters of a harmonic IBC input and then attempting to explain the reasons behind any noise reduction, it was first sought to better understand the BVI phenomenon and then develop localized IBC inputs to specifically target aspects of the interaction. This led to the development of robust inputs capable of reducing BVI noise in a variety of flight conditions. A new feedback metric utilizing blade pressures was then developed that was shown to follow the trends in peak observer plane BVISPL better than skid microphones, thus allowing for closed-loop control of IBC inputs.
5.1 Sensitivity of BVI-Induced Noise and Vibration to Variations in Individual Interaction Parameters
The influence of various blade-vortex interaction parameters on BVI noise and
BVI-induced vibratory hub loading was examined. An externally imposed vortex was allowed to interact with a rotor and key interaction parameters such as vortex strength, core-radius, the blade-vortex miss-distance, the spanwise location and extent of the interaction, the angles between the vortex and the blade in the blade-shaft plane and the disk plane, and finally the blade lift at the time of interaction were varied.
207 1. Reduction in vortex strength, Γ, reduces the peak BVI noise. Larger noise
reductions (for a given reduction in Γ) are obtained when the baseline vortex
strength is moderate to low. Reduction in BVI-induced vibratory hub loading is
directly proportional to reduction in vortex strength.
2. Increasing the core radius reduces BVI noise provided the miss-distance is small
(less than the core radius, so the blade is intersecting the vortex core). For larger
miss-distances (that place the blade outside the vortex core) change in core radius
has little effect on the peak BVI noise. For moderate baseline values of vortex
core radius, reduction in core radius has little effect in reducing BVI-induced
vibratory hub loading.
3. As long as the vortex core is not passing through the blade, increase in miss-
distance has a dramatic effect on reducing BVI noise. The largest reductions in
BVI noise are observed when the initial miss-distance is small, whereas the
reductions for comparable increases in miss-distance are smaller than when the
initial miss-distance is larger (beyond one chord length). Increased miss-distance
reduces both the overall amplitude of the BVI-induced vibratory hub loading as
well as the higher harmonic content.
4. Modest reductions in BVI noise can be obtained as the event moves to a more
inboard location and decreases in length. A more inboard interaction can actually
increase the amplitude of the 4/rev BVI-induced vibratory hub loading slightly,
although the percentage of vibration energy in the higher harmonics reduces.
5. Even a modest inclination (10-20°) of the vortex relative to the blade, in the
blade-shaft plane, can produce large reductions in BVI noise, as well as BVI- 208 induced vibratory hub loading. Inclination in the blade-shaft plane also reduces
the percentage of vibration energy contained in the higher harmonics. This
suggests that introducing modest amounts of anhedral or dihedral at the blade tips
may be attractive.
6. Inclination of the vortex in the disk plane, from a more parallel to a more oblique
orientation, also reduced the BVI noise. However, only very modest noise
reductions (~ 2-3 dB) are obtained for small inclination angles (of up to 10°).
Even for larger inclination angles, the noise reductions are smaller than those due
to comparable inclinations between the blade and vortex in the blade-shaft plane.
This suggests that forward or backward sweep of the outboard regions may be
less effective than blade anhedral/dihedral. The noise reductions are dependent on
the sign of the inclination angle. Reductions in BVI-induced vibratory hub
loading are also smaller than those due to inclination in the blade-shaft plane.
7. When the blade pitch is changed but the interacting vortex is held fixed relative to
the blade, there is only a change in the steady component of lift, but no change in
the impulsive lift. However, when the interacting vortex is free to move, the
bound vorticity of the blade causes it to convect upward. By increasing the blade
pitch, the magnitude of the bound vorticity increases, so a vortex that is passing
below the blade gets pulled toward it by the bound vorticity and a vortex in the
disk plane or above the blade gets pushed further upward, and away from the
blade. In practice, however, it is difficult to achieve noise reductions by changing
the blade lift around the time of interaction as the effect of the change in bound 209 vorticty is negated by other influences such as the change in tip vorticity of the
blade.
5.2 Localized Individual Blade Root Pitch Control for BVI Noise Reduction
The influence of Localized Individual Blade Control root pitch actuation inputs on Blade-Vortex Interaction noise was examined. IBC inputs (pitch reductions) were specifically considered over portions of the second quadrant of the rotor disk to reduce the strength of the vortex elements that convect downstream and produce advancing side
BVI in the first quadrant. Similarly, pitch reductions over portions of the third quadrant were designed to reduce the strength of the vortex elements that produced retreating side
BVI in the fourth quadrant of the rotor disk. The localized IBC pitch inputs were introduced over a limited 40˚ azimuthal range, and the second and third quadrant inputs were considered both individually, as well as in combination. Different IBC pitch input amplitudes and profiles – such as a truncated step, a ramped input, a half-period sine pulse, and a full-period cosine pulse – were considered. Simulations were conducted for a
4 meter diameter model rotor similar to that used in the HART test, for a condition representative of a low-speed descent characterized by high BVI noise.
1. The ramp-down-hold-ramp-up IBC input profile was found to be the most
effective of those considered in reducing BVI noise. For a given IBC pitch input
amplitude, , this profile resulted in the pitch being reduced to the maximum
value (of − ) over a substantial azimuthal range, compared to the half-period 210 sine pulse or the full-period cosine pulse. However, since there is no sharp or
sudden change in pitch (as with the truncated step input), it is implementable, and
since there is no instantaneous change in lift, the IBC input does not itself become
a significant source of noise (IBC noise).
2. Ramped IBC inputs in the second quadrant were highly effective in reducing the
advancing side BVI noise (from first quadrant BVI events). Input amplitudes of
= 1˚, 2˚, and 3˚ were considered. For the 1˚ input, the advancing side noise was
substantially reduced (up to 5.5 dB for the primary (k−3) interaction), and for the
2˚ and 3˚ inputs, advancing side noise was reduced to the background level.
However, the second quadrant inputs resulted in changes in wake and retreating
side interaction geometry so as to exacerbate the noise associated with the
dominant (k−0) retreating side interaction. The overall effect of the 3˚ second
quadrant input was to eliminate the advancing side noise lobe (from a baseline
level of 109.2 dB) and produce a peak retreating side noise level of 107.0 dB
(compared to 107.9 dB for the baseline). Thus, it is seen that when a localized
input (second quadrant input, in this case) is effective in its objective (reducing
advancing side noise), BVI noise levels due to sources that this input did not
address (retreating side BVI) gain prominence. This limits the total noise
reductions (from 109.2 dB for the baseline on the advancing side, to 107.0 dB
with the second quadrant IBC input, on the retreating side).
211 3. Ramped IBC inputs in the third quadrant have a more limited effect in reducing
retreating side BVI noise than the second quadrant inputs have on the advancing
side noise. Although the actual reductions in the strength of the interacting
vortices are comparable for second and third quadrant inputs of similar
amplitudes, the percentage reduction in strength is much smaller with the third
quadrant inputs. This is due to the larger circulation on the retreating side of the
disk (to compensate for lower dynamic pressures). In addition, the third quadrant
IBC inputs that reduce the strength of the interacting vortices appear to produce
detrimental changes in BVI geometry and miss-distance that partially negate the
benefits associated with the reductions in vortex strength. For a = 1˚ third
quadrant IBC input, the peak retreating side BVI noise level reduced by 1.1 dB,
and for = 2˚ the reduction was 2.8 dB.
4. When a ramped third quadrant IBC input is added to a ramped second quadrant
input, the largest overall reductions in BVI sound pressure levels are obtained.
When a = 3˚ amplitude is used for both the second and third quadrant inputs,
the advancing side BVI noise lobe is eliminated and the retreating side peak BVI
noise is reduced to 104.6 dB. Comparing to the baseline (109.2 dB peak
advancing side, and 107.9 dB peak retreating side), this represents a 4.6 dB net
reduction in peak BVI noise.
5. Localized IBC pitch reductions would require retrimming of the rotor.
Retrimmining the rotor in the presence of a second quadrant IBC input results in 212 an overall increase in lift over the advancing side and the front of the rotor disk to
compensate for the reduction in lift due to the localized IBC input. Although this
increase partially negates the localized lift reduction associated with the IBC
input, the overall mechanics and the effects of the IBC input are preserved. For a
3˚ amplitude second quadrant input, an advancing side peak BVI noise reduction
of 5.4 dB was still obtained, although the BVI noise was not completely
eliminated.
5.3 Metrics for BVI noise
Both skid microphones and blade-pressure measurements were examined for their ability to track with the peak advancing-side BVISPL on an observer plane below the rotor. Five skid microphone locations were considered for a BO-105. It was found that these did a poor job in tracking Peak advancing side BVISPL. Blade pressure measurements were analyzed with a simple RMS method, and then a simplified version of the WOPWOP acoustic code, dubbed BPAP. It was found that the RMS methods were not able to follow well with the peak BVI noise, while the BPAP code tracked very well with peak advancing side BVISPL.
1. Subtle differences in the BVI event can result in changes in the directivity of the
advancing side lobe. When this lobe moves away from the microphones, they
detect a drop in BVISPL, even if the strength of the lobe, and thus the noise in the
observer plane, increased. 213 2. On occasion the BVI event can produce weaker secondary lobes with a different
directivity than the primary lobe. If one of these happens to fall upon the skid
microphones, they will be more heavily influenced by this lobe than the primary
one. Since the two lobes do not necessarily grow or weaken together, the skid
microphones may produce erroneous results.
3. The multiple BVI events happening on the rotor can interfere constructively or
destructively with each other. The magnitude of these effects can be large enough
that they overwhelm the overall changes in BVI noise. When the interference falls
on the skid microphone locations it can cause them to track poorly with observer
plane data.
4. Simple RMS analysis of the blade pressure time history is incapable of following
observer plane noise as it fails to take into account how parallel the interaction is,
one of the prime considerations in the magnitude of the resulting noise.
5. The BPAP metric is capable of taking blade-pressure measurements and tracking
very well with observer plane noise through a variety of flight conditions and IBC
inputs. The reduction in complexity from the full version of WOPWOP allows for
an increase in run-speed that makes it feasible for real-time calculation of BVI
noise, and hence for feedback for a closed-loop controller.
5.4 Recommendations for Future Work
This dissertation developed new localized inputs to reduce IBC noise following a different approach than is commonly used, that is tailoring non-harmonic IBC inputs to 214 target specific interaction parameters rather than parametrically vary harmonic IBC inputs. This method may be utilized to potentially develop many other inputs capable of modifying other aspects of the BVI. For example, inputs targeting the time of interaction could change the bound vorticity in order to affect miss-distance. Or, inputs slightly ahead of the interaction could modify inflow in such a way as to change the path of interacting vortex, either by changing miss-distance or by causing the interaction to occur at an azimuthal location that will produce steeper angles in the disk plane. If multiple trailing edge flaps or blade twist are utilized, it may even be possible to vary the inflow in the radial direction in such a way as to modify the blade-shaft plane angle of the BVI.
Further investigation needs to be done on the feasibility of actuating the rotor blade in the method prescribed in this dissertation. It needs to be shown that actuators are capable of producing the level of actuation required for significant BVI noise reduction.
If it is determined that actuators are not currently capable of this, another method of implementing these physically motivated inputs must be explored, such as trailing-edge flaps.
Earlier investigations into HHC and harmonic IBC for the reduction of BVI noise reported that the best inputs for reduction of BVI noise were generally accompanied by dramatic increases in vibratory hub loads. The effects of the non-harmonic inputs presented in this dissertation are not known and must be investigated. It is likely that any closed-loop control of BVI noise utilizing non-harmonic IBC inputs will also need to consider vibration in the target function.
While the methodology used in this dissertation has the potential to develop new inputs, both these and the inputs developed in this dissertation need to be tested to 215 demonstrate their effectiveness. Such a test will necessarily have a traverse of microphones in the observer plane, but should also include microphones representing skid positions to test the prediction that these will track poorly with observer plane noise.
Ideally, such a test could also examine the effectiveness of the proposed feedback metric for identifying BVI noise, though that would necessitate extra instrumentation on the blade. Vita Brendon Daniel Malovrh EDUCATION: Doctor of Philosophy in Aerospace Engineering, The Pennsylvania State University, December 2012. Master of Science in Aerospace Engineering, The Pennsylvania State University, May 2000. Bachelor of Science in Aerospace Engineering, The Pennsylvania State University, May 1997. PROFESSIONAL EXPERIENCE: Aerospace Engineer, US Army Joint Research Program Office, NASA Langley Research Center, August 2004-Present. PUBLICATIONS: Malovrh, B., and Gandhi, F; “Localized Individual Blade Root Pitch Control for Helicopter Blade-Vortex Interaction Noise Reduction” Journal of the American Helicopter Society, Vol. 55, no. 3, pp. 032007-1-12, Jul 2010. Allan, B., Jenkins, L., Yao, C., Bartram, S., Hallisssy, J., Harris, J., Noonan, K., Wong, O., Jones, H., Malovrh, B., Reis, D., and Mace, W.; “Navier-Stokes Simulation of a Heavy Lift Slowed-Rotor Compound Helicopter Configuration,” AHS 65th Annual Forum, Grapevine, TX, May 2009. Jones, H., Wong, O., Watkins, A., Noonan, K., Reis, D., Malovrh, B., and Ingram, J.; “Initial Assessment of Surface Pressure Characteristics of Two Rotary Wing UAV Designs,” AHS 62th Annual Forum, Phoenix, AZ, May 2006. Jones, H., Wong, O., Noonan, K., Reis, D., and Malovrh, B.; “Aerodynamic Characteristics of Two Rotary Wing UAV Designs,” AHS Vertical Lift Aircraft Design Conference, San Francisco, CA, Jan. 2006. Malovrh, B., and Gandhi, F.; "Sensitivity of Helicopter Blade-Vortex-Interaction Noise and Vibration to Interaction Parameters", Journal of Aircraft, Vol. 42, No. 3, 2005, pp. 685-697. Malovrh, B., and Gandhi, F., “Mechanism Based Phenomenological Models for the Pseudoelastic Hysteresis Behavior of Shape Memory Alloys,” Journal of Intelligent Material Systems and Structures, Vol. 12, No. 1, Jan. 2001, pp. 21-30. Malovrh, B., and Gandhi, F., “Time Domain Mechanical Models for SMA Pseudoelastic Damping Behavior,” Proceedings of the 2000 SPIE Conference on Smart Structures and Materials, March 2000, SPIE Vol. 3989, pp. 324-335. Malovrh, B., and Gandhi, F.; "Influence of Balanced Rotor Anisotropy on Helicopter Aeromechanical Stability", AIAA Journal, Vol. 37, No. 10, 1999, pp. 1152-1160. Malovrh, B., and Gandhi, F., “Mechanism Based Phenomenological Models for Pseudoelastic Damping Behavior of Shape Memory Alloys,” Proceedings of the 40th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, St. Louis, Missouri, Apr. 1999, pp. 2723-2733. Wolons, D., Gandhi, F., and Malovrh, B., “An Experimental Investigation of the Pseudoelastic Hysteresis Damping Characteristics of Nickel-Titanium Shape Memory Alloy Wires,” Proceedings of the 39th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, Long Beach, California, Apr. 1998, pp. 2821 – 2833. Wolons, D., Gandhi, F., and Malovrh, B., “Experimental Investigation of the Pseudoelastic Hysteresis Damping Characteristics of Shape Memory Alloy Wires,” Journal of Intelligent Material Systems and Structures, Vol. 9, No. 2, Feb. 1998, pp. 116-126. Gandhi, F., and Malovrh, B., “Influence of Balanced Rotor Anisotropy in the Design of Aeromechanically Stable Helicopters,” Proceedings of the 53rd Annual Forum of the American Helicopter Society, Virginia Beach, Virginia, Apr. 1997, pp. 783 - 796.