Propeller Performance at Low Acoustic Values

Greased Lightning

Propeller Performance at Low Acoustic Values

March 2nd, 2016

Joshua Cruz

Alfredo Deschamps

William Foronda

Raul Gomez

Robert Hughes

Larry Johnson

Mihir Patel

Arthur Terlecki

Table of Contents

List of Figures______2
List of Tables______3
Abstract______4
Introduction______5
Methods______6

Thrust Stand______9
Microphone Stand______11
Testing______12

Results______14
Discussion______26
Budget______28

References______29

List of Figures

Figure 1: ODU Wind Tunnel (Test Section)______7

Figure 2: Thrust Stand Design______9

Figure 3: Stress Analysis with Displacement of Thrust Stand______11

Figure 4: Stationary Microphone Stand (in Wind Tunnel)______11

Figure 5: Mobile Microphone Stand (in Wind Tunnel)______12

Figure 6: Sound Pressure Level Equation______13

Figure 7: Forward Flight Efficiency Equation______15

Figure 8: Hover Flight Efficiency Equation______15

Figure 9: Advance Ratio Equation______15

Figure 10: Efficiency and Advance Ratio of Pitch Constant Propellers______15

Figure 11: Coefficient of Power Equation______16

Figure 12: Coefficient of Power and Advance Ratio of Pitch Constant Propellers______17

Figure 13: Coefficient of Thrust Equation______17

Figure 14: Coefficient of Thrust and Advance Ratio of Pitch Constant Propellers______17

Figure 15: Acoustic Values of Pitch Constant Propellers______18

Figure 16: Efficiency and Advance Ratio of Pitch Variable Propellers______19

Figure 17: Coefficient of Power and Advance Ratio of Pitch Variable Propellers______20

Figure 18: Coefficient of Thrust and Advance Ratio of Pitch Variable Propellers______21

Figure 19: Blade Passing Frequencies of Hover Mode______22

Figure 20: Blade Passing Frequencies of Flight Mode______23

Figure 21: Blade Passing Frequencies inside the Sound Module Room______23

Figure 22: Hover Efficiency and Coefficient of Thrust of Pitch Variable Propellers______24

Figure 23: Hover Efficiency and Thrust of Variable Pitch Propellers______25

List of Tables

Table 1: Budget Analysis______28

Abstract

The acoustic and aerodynamic properties of a flight propeller exist as interdependent properties of the mechanism, causing changes in one from alterations made to the other. NASA Langley Research Center recently tasked Old Dominion University engineering students with studying and analyzing the acoustic properties of propellers, and using the data to corroborate a propeller design, proposed by Georgia Institute of Technology engineers, with improved acoustic properties without sacrificing aerodynamic requirements for their GL-10 “Greased Lightning” prototype drone. The ODU team designed a thrust stand to support a propeller system in order to collect data on the sound generation, thrust force development, and efficiency of the propeller in a wind tunnel. Analyzing the data collected from five propellers, the team determined the most efficient, in terms of thrust force developed to noise generated, and identified some overall trends in propeller that can guide the design process. Utilizing the data analysis, Georgia Institute of Technology team will design a modified propeller to perform with the thrust force requirements of flight while reducing the noise generated. Preliminary results displayed a favorable design trend towards low tip speeds and square-tip propeller blades showing a reduction in sound generation.

Introduction

Recent advancements in aerodynamic vehicles have focused on the concept of creating smaller aircraft that do not require being manned by pilots (i.e. “drones”). While this has been accomplished in a broad spectrum of operations, a common objection has been the volume of sound generation from propellers for these aircraft. The benefits of a quiet drone are easily understood, whether in the civilian or military field; less noise disturbance, increased stealth performance, and other factors. Any of these benefits affirm the purpose of these vehicles, utilized for package delivery and data acquisition1.

In 2014, NASA introduced the GL-10 “Greased Lightning” drone. This aircraft utilizes ten propellers powered by electric motors, as opposed to the traditional propeller design for flight vehicles. This replacement allows for energy conservation, more reliable flight systems, and higher performance in aerodynamic operations2. While traditional propeller systems utilize fossil fuels and various petroleum based means of energy generation, the electric motors of the GL-10 drone require less weight, less heat, and less maintenance on behalf of flight crews. Meanwhile the ten propellers mounted along the wings of the drone serve to generate the required amount of force for a take-off and sustained flight; including the vertical take-off and mid-air transition to normal flight procedure. In the military sector, this design alteration will cost less for missions with a higher rate of mission success. To the civilian sector, the design changes can allow for higher usability in the personal field, for example, it could be used as transportation for small numbers of people, and all this would lead to increasing productivity for companies and reducing the time of orders to reach consumers.

Noise generation and acoustics are a very difficult field of study, with little control variability recognized in the production aspects of sound energy. The conventional approach to reducing noise generation has been to utilize absorbers or reflectors. Absorbers convert the sound energy to thermal energy to allow for dissipation, and reflectors prevent sound energy from entering previously occupied spaces by reflecting the incident wave field to block the vector travel3. The primary step in identifying steps to reduce noise generation is focusing on the areas of propagation, creating systems to counter propagation effects. Noise is generated by the rotational effects of the propeller as well as the vortices trailing off of the tip of the propeller blades. Depending on the tip speeds of the propeller, both vortex noise and rotational noise due to blade thickness are lower than the rotational noise caused by thrust and torque4. With the focus on propeller design, studies have been conducted to conclude the effect of change from propeller angle5 and blade deformations or shape6. Another addition for noise comes from the number of active propeller systems for the plane, corresponding with the number of propellers linked to each engine; noting that a 1:1 ratio for the two is more efficient5.

Methods

The entirety of the Greased Lightning acoustic performance testing is under the supervision of NASA, established by the points of contact (initially Bill Fredericks, subsequently Mark Moore, and more recently Ran Cabell). The overall task of acoustic examination was tasked to the National Institute of Aerospace, which subcontracted the experimentation procedure to two of its member institutions, Georgia Institute of Technology and Old Dominion University. The Old Dominion University testing was subdivided into a senior design project of undergraduates, and a graduate level project.

Coupled aerodynamics and acoustic data acquisition was recognized as the key factor for the experiment, and attention to details affecting its acoustic properties were given priority. Initial work began with designing aerodynamic stands to conduct testing, which will be explained in more detail in the testing procedure section. The ODU Wind Tunnel was selected as the primary aerodynamic testing simulation site, with an anechoic chamber chosen primarily for acoustic testing.

Interface software for the wind tunnel allowed control of notable and important factors to acoustic mechanisms of the test propellers, and reducing initial margin of error for data. All propellers chosen were fixed to a 16” diameter, and then selected for variance in pitch, blade number, blade-tip design, and material composition. Blade number variances were selected as two and three blade for analysis of different passing frequencies. Blade-tip designs chosen were scimitar, displaying a swept back rounded edge, square, designed with a flattened edge, and classical, with rounded edges of original propeller design. Material composition selected were wooden and glass-filled nylon to analyze the effect of density and surface friction on sound generation. Pitch angle chosen to test was 6°, 8°, and 10° for analysis of rotational speed of propeller on sound generation.

Alongside mechanical testing performed, Georgia Institute of Technology has a team of graduate students performing theoretical analysis via computer simulation software, determining an optimized propeller design that will be benchmarked against the analyzed data from this test. The mechanical testing was conducted through three different divisions of responsibility; mechanical force analysis, acoustic energy analysis, and performance analysis. The three divisions acted in concert to record data from shared testing environments and analyzing data for a full spectrum diagnostic on the propeller design to its effect on acoustic generation.

Data acquisition in reference to acoustics can be a demanding process, requiring multiple points of analysis and structural formations to contend with the needed factors to acquire data, but without interrupting flow and causing data to be contaminated. For purposes of this testing, a thrust stand was design to hold the test propeller, two microphone stands were designed to capture acoustic data, and the wind tunnel was calibrated and utilized to serve as the testing environment. The wind tunnel test section is highly reverberant, but separate anechoic chamber tests were expected to provide some correction factors.

Thrust Stand Design

Initial thrust stand design focused on a curved airfoil strut but moved to a straight airfoil with supporting top plate to hold the required pieces for testing in an L-shape design, fabricated from steel. In order to capture the mechanical data of the propellers, and later be able to analyze effects on acoustics, multiple parts were partitioned into the design. A magnetic inductance sensor, capable of sensing objects passing within 3mm of the receptor head, served as an RPM sensor to the propeller, allowing for the acquisition of a rotational speed for each propeller. An ATI Gamma 6-axis load cell was incorporated into the design to allow for thrust data to be recorded, to ensure that tests are performed only under flight permissible conditions; accompanied by an adapter and mount to allow for rigid connection to the thrust stand. A 380kV motor (Scorpion SII-4020, Scorpion Power Systems LTD) was selected to drive the propeller at a range of RPMs. This is similar to the motor used by NASA in the actual GL-10. The whole of the companion pieces were covered by a 3D printed polymer material nacelle cover, allowing for each piece to maintain performance efficiently, without the obtrusive bodies becoming introduced to the flow. The nacelle design mimicked a generic aircraft design, with a streamline body that tapers to a smaller focal point, allowing for airflow to remain unobstructed and larger boundary layers across its surface to reduce wake generation. The design allows the propeller tobe the only component exposed to the flow unobstructed, and places greater focus on the acoustic generation from it.

After the preliminary design of the thrust stand was finalized, the ODU team was able to run both a stress analysis and a CFD simulation of the model. The stress analysis was based on loads more than what the actual thrust stand would encounter. The stress analysis module was included in Autodesk Inventor to run the simulation, three types of loads were applied, axial, thrust, and a moment. The axial force was based on an approximate total weight of the force balance, the load cell adapter, the motor mounting plate and the motor, estimated as 3 lbs of axial force downwards on the force balance adapter as all the loads will be generated on that piece which is a coupler for all the components to the stand. When the team ran the first test on the stand, it was getting about 7 lbs of thrust, this thrust would tend to pull everything outwards, so a thrust load of 9 lbs pulling out was imparted, the team also put a counter force of 2 lbs that would be generated from the velocity of the wind, hitting the nacelle, most of it will be redirected by the aerodynamic shape of the nacelle, but some would act inwards onto the adapter. The distance from the end of the adapter to the tip of the motor was about 6.14 inches, and the team applied a moment of 18.42 lbs-in, with 3 lbs being the load. The analysis generated a set of results, such as stresses and deflections, the maximum deflection that the stand encountered was about .0015685 in, and the smallest was about 5.5×10-5 in which is not very effective in bending the stand, therefore it was concluded that the design was sturdy enough to withstand the forces that would be generated during testing. It was considered to model a dummy stand in inventor to study how flow affects the stand with specified static pressure inside a wind tunnel, taking theresults into account the team designed the nacelle that would act as an aerodynamic cover for the stand to redirect flow in a streamline.

Microphone Stand Design

To capture the acoustic data, as is the focus of the entire testing, two microphone stands were designed, one stationary and one mobile. The two stand design allows for analysis of upstream and downstream flow mechanics, giving opportunity to recognize acoustical components not present in the entire flight, and most likely to be subject to a flaw in the testing environment or equipment. The stationary Microphone stand was constructed using steel and PVC pipe, serving to hold the microphone (BrüelKjær Microphone Unit Type 4910-L-001) at a constant angle, chosen as 9°, and location in relation to the thrust stand, set far enough away to not be effected by pressure interferences and in line with the propeller direction to capture maximum acoustic information in downstream flow.

The mobile Microphone stand was constructed using an aluminum airfoil and rubber friction components, designed to stand perpendicular to the testing environment ground, with the microphone held parallel to the testing environment ground. The holding cone compliments the design by allowing the microphone to have translation in four degrees of motion: x, y, z, and Ѳx planes. This range of motion allows for capture at all necessary locations of the testing environment. With both Microphone stands in the testing environment, acoustic data can be captured at different angle reference frames, and monitored for disturbances and irregularities in the testing flow.

Testing Procedure

The ODU Wind Tunnel testing procedure is regulated in multiple processes performed in a distinct order to allow for efficient and manageable data acquisition and processing. The test propeller is attached to the thrust stand, which already sits in the middle of the testing environment of the wind tunnel. The stationary Microphone stand remains behind the thrust stand in direct line with the propeller, attached to pulse software (BrüelKjær PULSE 13.1). The wind tunnel is sealed to prevent exterior air flow entrance. A preliminary acoustic measurement of the unmoving propeller is recorded, to account for ambient sound in the testing environment. The wind tunnel is controlled through the labview software, with signals sent to the motor to induce torque. The propeller is moved at increasing RPS until the required thrust for hover flight is reached, and acoustic measurements are recorded. The propeller RPS is reduced to zero, and the wind tunnel is turned on to match flight velocity characteristics. Propeller RPS is increased until forward flight thrust is reached, and acoustic measurements are recorded. Acoustic data is analyzed through MATLAB software into an overall SPL, calculated with an equation, and power spectrum, allowing for recognition of focus frequencies within the human hearing range. Acoustic data is plotted against the performance data of the propellers in regards to efficiency and advanced ratio. Comparisons are analyzed to determine the performance of the propeller in relation to its acoustic properties. Testing is performed for each propeller, with all data recorded compared to evaluate the most acoustically efficient propeller within the testing parameters.

Inside the ODU Wind Tunnel the blade passing frequencies of the propellers become distorted by the blade passing frequencies of the turbines used to generate wind speed. To evaluate the realistic blade passing frequencies of the propellers a sound module room was utilized. A sound module room is used in music production to create an acoustically sealed room, which for purposes of this testing allows for the evaluation of a propeller without external mechanism frequency distortion. Using this room, hover mode testing is performed on a propeller to measure the realistic blade passing frequencies, which can be compared to the blade passing frequencies measured in the ODU Wind Tunnel; SPL levels will be higher in the sound module room due to design, but are not consequential to testing.

Results

Testing has unveiled specific trends from the data analysis, which will guide future testing in selection of propeller styles to focus on. Data analysis was performed on propellers with variable blade number, blade-tip, and material composition, with pitch angles of 8° and 10°, and then performed on two and three blade, scimitar-tip, glass-filled nylon propellers with variable pitch angles. All data analysis is focused on performance and acoustic properties of the propellers.

In terms of Efficiency, seen in equations below for forward flight mode and for hover mode, and Advance Ratio, the propellers acted in an average until the Advance Ratio reached .4, upon which the wooden 16x10 two-blade square-tip propeller performed best by maintaining an Efficiency of greater than .5 by an Advance Ratio of .6, while the remaining propeller styles reached zero efficiency rating in data.