Multi-degree-of-freedom liner development: Concept to flight test

Abstract

The emphasis on increased turbofan fuel efficiency requires advanced turbofan designs that will integrate higher engine bypass ratios and shorter nacelles. The resulting acoustic signature of these designs will have a more broadband character as well as a smaller available area for liner installation. This two-fold impact compels a need for an improvement in the state of the art in liner technology. Increasing the acoustic absorption efficacy over a broader frequency range is a means to address this need. An acoustic liner development and optimization process was conceived and employed to achieve and demonstrate an improved broadband liner design concept. A series of increasing technology readiness level liner studies were conducted to enhance the optimization methodology while validating the concept. This progression spanned several NASA Aeronautics Research Mission Directorate programs/projects due to its relevance. This article reviews the development and evaluation process of the multi-degree-of-freedom liner technology concept from formation through simple experimental models to a flight test over an approximate 10-year period, focusing on the discrete tests comprising the development.

Keywords

Turbofan noise reduction acoustic liners experimental testing flight test

Introduction

SIGNIFICANT reduction in aircraft noise is required to meet ongoing noise regulation in the USA and Europe. Furthermore, increased noise levels resulting from turbofan engine design advancements necessitated by higher performance and reduced emission goals will need to be mitigated. Since the turbofan engine is a large contributor to aircraft noise, any overall reduction in aircraft noise must address engine noise reduction.¹ The introduction and utilization of higher bypass ratios and integration of shorter engine nacelles in advanced turbofan designs have created a need for increased fan noise reduction over a broad frequency range. Next-generation broadband liner designs must take advantage of novel liner configurations to account for the changing acoustic signature and geometric constraints. The development of such a novel liner configuration and assessment of a broadband acoustic liner optimization process to address this need was pursued through a series of recent design and experimental studies.

Aircraft noise² can be separated into two general sources: (i) propulsion noise and (ii) airframe noise, pictorially identified in Figure 1. Depending on the aircraft and flight condition the relative levels of these sources vary; generally, the turbofan engine noise dominates for the current conventional tube and wing configuration. Turbofan engine noise sources can be further decomposed into internal and external sources. Figure 2 presents notional relative strengths of these sources. Turbofan internal noise results from a variety of sources within the engine as shown in Figure 3. The fan, comprised of the rotor and stator, is the main source within the bypass duct. The rotor noise sources arise from inflow and boundary layer distortions, and rotor wakes, vortices, and self-turbulence. The stator has a similar set of noise sources, and rotor-stator interaction adds additional noise sources. In general, within the turbofan, each of these sources creates a combination of broadband and tonal components. Over the last few years, the relative contribution of each of these components has changed with the trend toward increasing fan bypass ratio – from tone to broadband dominated as the bypass ratio increased (Figure 4).

Figure 1.

Sources of in-flight aircraft noise (from Boeing Aircraft Company).

Figure 2.

Representative levels of aircraft noise (from reference²).

Figure 3.

Sources of turbofan engine noise.

Figure 4.

Turbofan engine design trend. (a) 1960’s – 1970’s: BPR ∼ 4. (b) 1990’s – 2010’s: BPR > 9.

Currently, there are two primary means to reduce propulsion noise: (i) through careful design of the source (though aeroacoustics takes a lower priority relative to other concerns, most notably performance, operability, reliability, maintenance, and cost) or (ii) attenuation of the noise after it is generated.³ A few examples of source design include stator vane count change, increased blade row spacing, and blade and/or vane shaping.⁴ The main method currently employed for noise attenuation after it is generated is through the installation of passively absorptive liners (honeycomb/resistance sheet) in locations along the nacelle internal walls. Currently, these liners are installed in locations where the aerodynamic and thermodynamic conditions are relatively moderate (in the bypass area as compared to the core, for example) as Figure 5 illustrates.

Figure 5.

Turbofan engine acoustic liner locations.

This article presents the development of the Multi-Degree of Freedom (MDOF) liner concept which was designed to have a greater attenuation over a wider bandwidth to accommodate the proposed turbofans with shorter nacelles and more broadband contribution to Effective Perceived Noise Level (EPNL). The development of this advanced liner concept followed a deliberate path of extending the liner design and evaluation tools coupled with a series of tests to evaluate the efficacy of both the tools and the liner concept. A series of MDOF liner designs and studies at increasing Technology Readiness Level (TRL) was conducted. The tests ranged from very-low TRL (impedance tubes) culminating in a high TRL flight test. During the progression, a liner optimization process was developed and validated. To increase confidence in the tools used for the broadband liner design process, the overall optimization methodology was enhanced in conjunction with these studies.

Initial MDOF liner designs targeted the NASA Langley Liner Technology Facility (LTF)⁵ to validate the concept. Subsequent design and testing in the NASA Glenn Advanced Noise Control Fan (ANCF) Rig^6–8 and 9 ft × 15 ft Low-Speed Wind Tunnel (LSWT)^9–12 provided further confidence and enhancements in the design process. Based on these results, further validation was pursued through the fabrication and testing of liner designs for the NASA Glenn DGEN380 Aero-propulsion Research Turbofan (DART).^13,14 The demonstration of the MDOF liner acoustic efficacy and manufacturability, and the detailed validation of the design/prediction process, led to a flight test on a Boeing 737–7 MAX¹⁵ narrow-body airliner flight demonstrator.

The structure of the paper is as follows: After this introduction, in the next section, a brief review of liner theory is presented with a follow-on specifically addressing the MDOF concept and analytical model (detailed descriptions of the models can be found in the references). In the Technology progression of the Multi-Degree of freedom liner concept development section, a description of the series of increasing TRL test projects is presented. For each test (presented in temporal order in separate subsection) a description of the test article is provided, followed by the design of the liner unique to that test article, and then a summary of the particular results. This order is chosen to illustrate the sequential, developmental character of the technology progression. The reader will be directed to specific references for detailed descriptions of the methodologies and analysis. Finally, an overall conclusion is presented in the Summary section.

It is noted that significant work occurred outside of this progression of tests, as addressed in^16–18 and other advanced liner concepts are following a similar development path: Over-the-Rotor treatment^19–22 and potentially Bio-liner^23,24 are two examples.

Concept background

Acoustic liners have been studied since the early 1950s for aircraft noise reduction.²⁵ In general practice, passive panels are mounted in the walls of an aircraft nacelle inlet and/or aft-bypass duct to absorb noise generated by the fan. Liners mounted in current commercial aircraft nacelles are almost exclusively perforate-over-honeycomb structures. The combination of the porous face sheet and underlying resonant cavity with a sealed backing acts as a Helmholtz resonator for the dissipation of incident acoustic energy propagating over the liner (see Figure 6).

Figure 6.

Traditional liner components terminology.

Passive liners currently in service are generally Single-Degree-of-Freedom (SDOF) or Double-Degree-of-Freedom/Two-Degree-of-Freedom (DDOF/2DOF – used interchangeably) structures. SDOF liners have a single resonant cavity, and couple to a single design frequency (typically a fan harmonic) and its odd harmonics. They are quite narrow-band in their attenuation. Depicted in Figure 7, a DDOF has an embedded porous septum of uniform character creating two separate resonant cavities. The height of the septum is carefully chosen such that the upper and lower cavities are tuned to two distinct frequencies, thereby resulting in improved broadband attenuation (also Figure 7). The focus of the current article is the evaluation of MDOF liners designed as an extension of DDOF liners, whereby the cavity heights and septum resistances are judiciously chosen to extend the acoustic attenuation toward more broadband character.

Figure 7.

Acoustic liner sections with generalized absorption spectra. (a) SDOF (b) DDOF.

Liner theory and modeling summary

As mentioned previously the current method to attenuate turbofan noise is the installation of passive liners in the duct nacelle walls. One of the most important parameters for acoustic liner characterization is the acoustic impedance. This parameter defines how well the liner absorbs sound under different conditions. For a traditional local-reacting acoustic liner, it is defined as the ratio of the acoustic pressure and the normal component of the acoustic velocity at the surface of the liner.²⁶ The acoustic impedance is normalized by the characteristic impedance of air. It is a function of frequency, liner design and geometry, and the aeroacoustic environment in which the liner is installed. The impedance is then used to represent acoustic liners via impedance boundary conditions in propagation codes to predict the resultant acoustic absorption. Sound absorption is directly affected by the duct geometry. In contrast, the acoustic impedance is an intrinsic property of an acoustic liner, which implies it is independent of duct geometry. For these reasons, liner samples are generally tested in impedance tubes to obtain the acoustic impedance prior to evaluation in test rigs that incorporate geometric and aeroacoustic environments of varying relevance.

The liner design optimization process uses in-duct propagation results to provide predicted optimum impedance values at specific flight conditions and frequencies. The liner modeling tools provide design impedance values that best approximate the optimum impedance. Based on that, the desired liner configuration is specified and manufactured, and the as manufactured dimensions are utilized to obtain the predicted liner performance. The manufactured liner is tested and comparisons between the predicted and measured liner attenuation analyzed. This liner design process is illustrated in Figure 8.

Figure 8.

Liner design process.

The impedance prediction model used in these studies combines two models. The first is a transmission line model²⁷ that assumes acoustic wave propagation through each layer of the liner and the second is a lumped element model^28–30 used to compute the impedance change across perforates. The normalized surface impedance spectra presented by each chamber of the liner are computed separately and are then combined to determine an effective surface impedance spectrum that is assumed uniform across the liner surface.

For the investigations described herein, the majority of the duct propagation and radiation predictions were conducted using the CDUCT-LaRC (CDL) code.³¹ This code calculates the propagation of a given acoustic source ahead of the fan face or aft of the exhaust guide vanes in the inlet or exhaust ducts, respectively. The code has the capability of computing the noise radiation field outside the duct from the propagation calculations. The three-dimensional duct may include acoustic treatment and incorporate struts/bifurcations. The duct propagation module is based on the CDUCT code described in references.^32–34

Manufacturing constraints (e.g., liner geometric parameters such as porosity and core depth) are key ingredients in this modeling phase and must be taken into account to design liners that can be realistically achieved. In this study, the honeycomb cells were restricted to contain one mesh-cap in a MDOF liner chamber. The mesh-cap depth, as well as its DC flow resistance, were allowed to vary from cell to cell. Finally, based on currently available manufacturing techniques, the pattern for the MDOF liner was limited to a configuration of four cells.

Multi-degree of freedom liner concept

SDOF and DDOF liner configurations are currently employed in aircraft engine nacelles. Additional degrees of freedom (and thus, increased absorption over the desired frequency range) can be achieved by replacing the uniform embedded septum layer with distinct mesh-caps that can be mounted at selected heights and with prescribed resistances in each core chamber - the variable-depth liner illustrated in Figure 9. The technique of embedded septum layers is becoming increasingly viable for turbofan engine applications due to recent advances in manufacturing techniques. Individual core chambers of this liner have different depths and each chamber functions as a quarter-wavelength resonator tuned to a different frequency. The liner can be designed to absorb sound over a wide frequency range by careful selection of the different chamber depths (to tune to different frequencies as illustrated in Figure 10).

Figure 9.

MDOF liner cell illustration.

Figure 10.

DDOF vs MDOF liner absorption spectra comparison.

A common characteristic of most of these liner configurations is that they are designed to achieve increased absorption bandwidth with the same volume. Also, these configurations are distinct from those that are currently used in aircraft applications. As a result, prediction models were required to be extended to account for the distinct features of the MDOF concept. This required that the experimental methods used to evaluate these liners be continuously improved such that the results are sufficiently accurate to enable these modeling improvements. Given the nature of these liners, there is also a need for improved experimental and analysis tools to allow more data (due to higher frequency resolution) to be processed quickly.

Technology progression of the multi-degree of freedom liner concept development

A progression of experimental investigations centered about tests on specific test rigs was conducted over the period 2012–2018 (approximately). For each investigation, a version of the computer model described in the previous section was used to specify the treatment impedance for the expected acoustic signature, along with obtaining the predicted attenuation. A liner to fit the test rig was designed and manufactured. If any test rig restrictions necessitated a modification to the manufactured liner, an updated prediction of the attenuation was performed. The insertion loss of the liner installed on the test rig was measured by the standard acoustic instrumentation generally available on that test rig. Comparisons between the predicted and measured attenuation were computed and analyzed to evaluate the efficacy of the model and the liner. Figure 8 illustrates the design process. Impedance tube testing was conducted to achieve low TRL. Testing on ducted fan models evaluated the MDOF in the low-mid TRL range. Turbofan engine testing was conducted to investigate mid-TRL and finally, a flight test resulted in a high TRL demonstration. The timeline of the series of investigations is presented in Table 1.

Table 1.

Multi-degree of freedom liner development timeline.

TRL 2/3	TRL 3	TRL 4	TRL 5	TRL 7
2012	2014	2015	2017	2018
NIT/GFIT	ANCF	FMD	DART	Q-MAX
Impedance Tubes	Low-Speed Ducted Fan	High-Speed Ducted Fan	Small Turbofan	Flight Test

Investigation in normal incidence tube

The Normal Incidence Tube (NIT) is located in the NASA Langley LTF which is shown in the photograph in Figure 11. NIT testing is considered TRL 2. A detailed discussion of the NIT can be found in.⁵

Figure 11.

NASA Langley LTF photo.

NIT description

Generally, the NIT is the first step in an experimental evaluation of a liner concept. The NIT is simple to operate and only requires a small (inexpensive) sample and is therefore good for concept development.

Each liner concept used in this investigation was tested in the NIT, a 2 in × 2 in waveguide (see the sketch in Figure 12). It contains six 120 W compression drivers to generate a plane wave sound field that impinges on the surface of the liner and combines with reflections from the liner to create a standing wave pattern. The Two Microphone Method³⁵ is used to measure the complex acoustic pressures at two prescribed distances from the liner surface, and the no-flow acoustic impedance of the liner sample can be determined as a function of frequency.

Figure 12.

NIT Sketch.

NIT liner design and installation

As part of the MDOF liner development, two tests were conducted in the NIT. The first set tested only in the NIT is described in the next few sub-sections. The second set of liner samples was tested in both the NIT and GFIT. That design and the results are described in the Investigation in grazing flow impedance tube section.

The liner used in the first study³⁶ incorporated the Hexcel AcoustiCap® product. Each individual honeycomb cell in the acoustic liner can be customized. A cell can contain zero, one (Figure 13(a)) or two (Figure 13(b)) buried septa. The depth of each septum can vary from cell to cell – constructing either uniform-depth or variable-depth configurations. The DC flow resistance of each septum can vary from cell to cell. The variation can be customized into a grid pattern of different septum depths or resistances within the acoustic panel.

Figure 13.

Photos of NIT mesh-cap liners.

The test liner samples used in this study were fabricated with 0.375 in diameter honeycomb core material (flat-to-flat measurement, hexagonal shape). Acoustic mesh material with DC flow resistances from 300 to 1200 MKS Rayl were used for the buried septa (referenced as “mesh-caps” in the remainder of this paper). The mesh material is shaped for insertion into the hexagonal shape, and a punch is used to form the mesh and insert it into the honeycomb core. The mesh-cap is anchored to the honeycomb cell wall with an adhesive bond at the chosen depth. The liner is formable because the septum is a flexible mesh. The honeycomb structure maintains its strength since it does not have to be spliced at each septum. Each liner was mounted into a 2 in × 2 in sleeve with the same length as the multi-layer liner. This assembly was mounted onto a 1-inch thick rigid backplate.

Multiple two- and three-layer liners (see Figure 14) were tested in the NIT, with very consistent results. Note that a two-layer liner is not necessarily a 2DOF liner if within a layer the cell characteristics vary; likewise, a three-layer liner is not always a 3DOF liner. In those situations, usage of the term MDOF is more appropriate. Two sets of NIT tests were performed: (1) a core (with embedded septa) with a backplate and (2) facesheets over the core, and a backplate. For the sake of brevity, the results presented in this report are confined to one two-layer liner (MC1) and one three-layer liner (MC2) that are representative of all of the liners tested. The characteristics of these two liners are provided in Table 2.

Figure 14.

Two-layer (2DOF) and three-layer (3DOF) configurations sketch.

Table 2.

Multi-layer, mesh-cap liner samples dimensions.

	Liner #
Characteristic:	MC1	MC2
Layer 3 height, h3 (inches)	–	0.3
Layer 2 sheet DC flow resistance, (MKS Rayl)	–	300
Layer 2 height, h2 (inches)	0.5	0.7
Layer 1 sheet DC flow resistance (MKS Rayl)	700	300
Layer 1 height, h1 (inches)	1.0	0.5

The liners used in this portion of the study were fabricated without facesheets as the main focus of the study was to investigate the ability to predict the effects of the embedded mesh-caps.

NIT test assessment

Results are presented for the two mesh-cap liners. Figure 15 provides a comparison of absorption coefficient spectra measured for each of the mesh-cap liners. Both are observed to provide excellent sound absorption over the design frequency range (1.4 to 3.0 kHz). Recall that these liners did not contain facesheets. It is reasonable to assume that the addition of a facesheet should cause similar effects to those observed for the wide-chamber liners (i.e., shift frequency range of peak absorption downward and reduce absorption bandwidth).

Figure 15.

Measured absorption coefficient spectra comparison (in NIT).

The transmission line impedance prediction model described earlier provided good to excellent comparisons to the measured impedance spectra for each liner concept (impedance comparisons are detailed in reference36) For those liners with facesheets (tests were conducted with and without facesheets) a modeling approach that employs a lumped element facesheet model is acceptable for the thin facesheets currently in use. This study showed the model was useful in predicting the acoustic absorption spectra and variable-impedance liner configurations are amenable to aircraft applications.

Investigation in grazing flow impedance tube

The Grazing Flow Impedance Tube (GFIT) is also located in the NASA Langley Liner Technology Facility as discussed in reference⁵ and shown in Figure 11. GFIT can be considered TRL 3.

GFIT description

The GFIT is often used in conjunction with the NIT to provide additional information on the basic acoustic characteristic of liner samples in the presence of grazing flow. It has a cross-sectional geometry of 2.0 in wide by 2.50 in high and allows evaluation of acoustic liners with lengths from 2.0 in to 24.0 in. The surface of the test liner forms a portion of the upper wall of the flow duct (highlighted in blue in Figure 16). Twelve acoustic drivers form an upstream (exhaust mode) source, and six acoustic drivers form a downstream (inlet mode) source. The acoustic drivers are used to generate tones one frequency at a time over a frequency range of 0.4 to 3.0 kHz, at source SPLs up to 155 dB, and 0.6 Mach number at the centerline. Ninety-five microphones are flush-mounted in the rigid walls of the duct, with fifty-three distributed along the lower wall over an extent of 40.0 in. Acoustic pressures measured with these lower-wall microphones are used for the majority of the analyses described in this report.

Figure 16.

GFIT sketch.

GFIT liner design and installation

Three liner configurations (2DOF, 3DOF, and MDOF) were evaluated in this second study.³⁷ Two sets for each type were fabricated: one for installation in the NIT and the other for installation in the GFIT. The dimensions of the GFIT liner samples were 2.0×2.5×22 inches. Figure 17 provides sketches of the three liner configurations. Figure 18 shows photographs of the GFIT liner sections.

Figure 17.

NIT/GFIT test liner sketches. (a) 2DOF configuration (b) 3DOF configuration (c) MDOF configuration.

Figure 18.

Test core sample photographs.

The 2DOF liner configuration contains one embedded mesh-cap in a chamber and every chamber is identical. Given the high resistance of the mesh-cap (2000 MKS Rayl; normalized flow resistance of 5), very little sound is expected to be transmitted into the lower air layer next to the rigid backplate. Therefore, this liner was expected to exhibit the acoustic characteristics of an SDOF liner with a core depth of 1.71 inches.

In the 3DOF configuration, each chamber contains two mesh-caps and is identical to every other chamber in the liner. The two embedded mesh-caps have identical flow resistance values (500 MKS Rayl) and are positioned 0.720 and 1.71 inches from the backplate.

The MDOF liner configuration is comprised of two unique 2DOF chambers, each of which contains one mesh-cap, and these two chambers are assumed to be replicated (in pairs) to form the entire liner. The resistance values for the mesh-caps in each chamber differ and have values of 280 and 590 MKS Rayl, respectively.

GFIT test assessment

This comparison highlights some of the key features that can be exploited in the design of parallel-element, embedded mesh-cap liners. The predicted acoustic impedance spectra from the as manufactured acoustic liners were computed using the described design approach. The measured attenuation spectra for the 2DOF, 3DOF, and MDOF liner samples are shown in Figure 19(a) (without wire mesh facesheet, no flow), Figure 19(b) (with wire mesh facesheet, no flow), and Figure 19(c) (with wire mesh facesheet and flow).

Figure 19.

Measured attenuation spectra in GFIT. (a) without facesheets (M#=0.0), (b) with facesheets (M# = 0.0), (c) with facesheets (M# = 0.3).

Without a facesheet (Mach # =0.0 test), the 2DOF liner provides good attenuation over a very narrow frequency range. The DC flow resistance for the mesh-cap installed near the bottom of the 2DOF liner has a very high resistance causing the liner to respond like a conventional SDOF liner. The 3DOF and MDOF liners provide much better attenuation across the majority of the frequency range used in this test. Adding a wire mesh facesheet significantly improved the attenuation achieved with the 2DOF liner. The addition of the facesheet moves the DC resistance toward optimum for the GFIT as designed. While the 2DOF liner then outperforms the 3DOF/MDOF liners, it is important to remember that the purpose of the investigation is to validate the methodology, not optimize the design. In Mach 0.3 flow, the attenuation spectra are generally reduced for all three liners while still providing significant broadband attenuation.

Summarizing, this investigation showed that the model described in the Concept background section favorably predicted the acoustic absorption spectra in the presence of grazing flow for all degree-of-freedom liner sample types. The 3DOF and MDOF liners provided much better attenuation without the facesheet and reduced attenuation with the addition of a facesheet. It is important to recognize that these results are not necessarily representative of those that would be achieved with other 2DOF, 3DOF, and MDOF liners but rather a demonstration of the design methodology. This comparison highlights some of the key features that can be exploited in the design of parallel-element, embedded mesh-cap liners.

Investigation on low-speed fan

A test on a low-speed ducted fan model was conducted to achieve TRL 3.

ANCF description

The ANCF,^6–8 shown in Figure 20, is a low-speed ducted fan testbed for measuring and understanding fan-generated aeroacoustics, duct propagation, and radiation to the far-field. It is a highly configurable 4-foot diameter ducted fan located in the Aero-Acoustic Propulsion Laboratory (AAPL)³⁸ at the NASA Glenn Research Center (for the time period covered in this paper). The maximum rotor tip speed of ∼500 ft/sec and duct Mach number of up to 0.16 is low but allows for limited studies on the effect of flow. The fan pressure ratio is a few inches of H2O – therefore it is not relevant to study turbofan performance utilizing the ANCF.

Figure 20.

The advanced noise control fan.

The Configurable Fan Artificial Noise Source (CFANS)³⁹ – a unique ANCF feature – was utilized to generate and control circumferential modes and to generate radial modes to provide an even wider distribution of acoustic sources for the assessment of the efficacy of liner concepts.

ANCF liner design and installation

For the more complex 3 D acoustic environment in a ducted fan, the broadband liner design process begins with a series of propagation predictions (at selected ﬂow conditions and frequencies) over a speciﬁed impedance design space. These calculations are used to obtain a predicted optimum impedance spectrum over the full range of ﬂow conditions of interest. Since CDL can accept arbitrary source speciﬁcation, it is convenient to specify the acoustic source distribution in terms of duct modes. For situations in which the source pressure is available, this greatly simpliﬁes the conversion to the required acoustic potential. However, when source information is not available, an assumption on the source description must be made. In this study, the source modal powers (and hence, amplitudes) and modal phases were allowed to vary randomly and independently.

Two liner cores were designed and manufactured for this test program. The first liner incorporated a septum with a constant depth to provide a liner with a constant impedance liner design (CIL) which can be considered a 2DOF liner. The second design incorporated a unique variable depth septum creating a variable impedance liner (VIL) which can be considered a MDOF liner. The design process, acoustic specifications, and predicted performance of the liner cores are detailed in references.^40,41 The liner cores were manufactured separately and individually integrated into an existing 48-inch spool piece. The effective axial length of each liner core was 16 inches. Descriptions of the liner core properties are in Table 3.

Table 3.

ANCF liners.

Liner type:	CIL	VIL
Characteristic:		Cell group 1	Cell group 2
Core (inches)	2.0	2.0	2.0
Septum flow resistance (MKS Rayl)	800	1100 1100	1200 1100
Septum depth (inches)	1.25	1.25	1.30
Face sheet resistance (MKS Rayl)	20	20	20

Photographs of a liner assembly sample are shown in Figure 21. Both cores had a total thickness of 2.0 in. The CIL had cells of uniform depth and septum resistance; the VIL core was comprised of repeated groups of four cells. Two of these cells had a septum penetration height of 1.25 inches, with DC flow resistances of 1100 and 1200 MKS Rayls, respectively. The other two cells had a penetration height of 1.30 inches and a septum flow resistance of 1100 MKS Rayls. The septum is seen in Figure 21(a). Figure 21(b) shows the liner assembly installed in the ANCF spool (the perforated screen is partially removed for viewing – it was not tested this way). Figure 22 shows the liner spools installed on the ANCF with the CFANS integrated, in two orientations.

Figure 21.

ANCF liner spool build-up photographs. (a) Cross section of liner core. (b) Installation in ANCF spool.

Figure 22.

Liner installation on ANCF. (a) Stack-up in vertical, off-stanchion orientation (no flow). (b) Build-up in horizontal, standard on-stanchion orientation (flow & no-flow).

The CIL and VIL cores were tested in separate assemblies. The CIL and VIL were both tested in the vertical orientation (no flow) with a 24 in center-body. The VIL was also tested with a 36 in center-body in the vertical configuration. A set of frequencies were generated from the CFANS while Rotating Rake data were acquired at the entrance and exit of the installed liner for each liner planform (hard-wall, and fully exposed). The VIL liner was then tested on the ANCF on-stanchion, with flow – the normal operating condition of the ANCF. Rotating Rake data were acquired upstream and downstream of the installed liner with the fan as a source with 0 (rotor alone) or 14 vanes over the standard rpm range. Far-field directivity data were acquired for the hard-wall and fully exposed liner planforms.

ANCF assessment

Selected insertion loss predictions from reference42 are compared with the experimental measurements in Figure 23. The predictions are a mean of multiple runs of random amplitudes and phases, while the experimental measurements are from a single set of broadband modes (which were created by sending a unity amplitude signal with a random, but fixed, phase to each driver). The 95% confidence distribution bar for the predictions is placed on the bar graphs. (From experience the measurement uncertainty is +/−1 dB).

Figure 23.

ANCF insertion loss comparison (experimental vs predicted). (a) Constant impedance liner. (b) Variable impedance liner.

Figure 23(a) shows the insertion loss vs. frequency for the CIL installed on the ANCF in the vertical orientation (no-flow) and the CFANS generating the acoustic signature. Most of the points could be considered within the uncertainty range or just outside. Figure 23(b) shows the comparison for the VIL. For the 24 in center-body the comparisons are similar, within the uncertainty range or just outside. The difference between the predictions and measurements for the 36 in VIL center body is more generally outside the uncertainty, but the trend of the curves match. Because of the underlying assumptions (parabolic approximation, reflections are ignored) of the CDL, it typically provides conservative estimates of the predicted attenuation. The relative performance of the two liners with the 24 in center-body (compare across sub-plots) compares well for the predictions and measurements. When the VIL is coupled with the 36 in center-body the insertion loss is greater, which also was predicted by the design methodology. Added to this plot are data from the VIL installed on the ANCF with the rotor alone as a source. More points are plotted here because the first three harmonics of the blade passing frequency (BPF) are plotted for each fan rpm. The breaks in the lines indicate the jump to a higher harmonic. Note that the ANCF center-body varies from the rotor plane to the constant area section at the exit, the constant area section has a 24 in center-body, and about 75% of the liner is in this section; the leading 25% is in the converging section. This would clearly affect the liner performance.

The generally higher measured insertion loss compared to the predicted may be a consequence of duct reflections affecting the experimental results, perhaps due to the center-body boundary conditions. Another possibility may be the physical difference between running a single, phase-randomized, but fixed, distribution of modes compared to a fully random set of many modal distributions. Overall, the predicted and measurement trends agreed very well – both liners perform reasonably similar across the frequency range, as designed.

In some cases, the insertion losses were under-predicted relative to the measurements, likely due to the statistical source distribution used to compute the value for the prediction being more conservative than an actual measurement with a fixed acoustic source.

Investigation on high-speed fan

A test on a high-speed fan ducted model was conducted to achieve TRL 4.

9 × 15 Ultra-high bypass fan description

The 9x15-Foot LSWT at NASA Glenn Research Center is used for acoustic and performance testing of research fans at realistic fan tip speeds and pressure ratios. The tunnel walls are acoustically treated, and a large muffler mitigates noise from the drive motors and compressor. Several reports overview the facility including aerodynamic test capabilities and acoustic quality.⁴³ The research fan was powered by the Ultra-High Bypass (UHB) drive rig.⁴⁴ This drive rig has a four-stage air turbine driven by compressed air generated by a 450-psi central air system. A photo of the high-speed fan model in the 9x15 test section is shown in Figure 24.

Figure 24.

Photo of high-speed fan model in 9x15 wind tunnel test section.

NASA and Honeywell Aerospace conducted a model (22-inch diameter fan) – the Honeywell Fan Module Demonstrator (FMD) – acoustic wind tunnel test in the fall of 2014 in the LSWT. Details of the model and configurations tested are in reference.⁴⁵ The model has an active booster, on the same shaft as the rotor that pressurizes the core flow path.

9 × 15 Fan module demonstrator design and installation

Two liners were tested on the FMD (Figure 25). The “traditional” liner was an SDOF liner using a perforated facesheet over an aluminum honeycomb, designed by Honeywell using flight hardware techniques. The outer wall liner depth was larger than the depth of the inner wall liner. The intent was to assess the ability to achieve more broadband attenuation than with the same depth on each wall. The combined pair of liners are intended to expand the frequency range of attenuation since each depth targets a different range of frequencies. So, while each individual liner is SDOF, the combination could be considered MDOF.

Figure 25.

MDOF liner installation on honeywell fan model demonstrator.

An advanced MDOF liner was designed at NASA Langley using a fabrication process developed and patented by Hexcel Corp as discussed earlier. This liner had a perforated metal facesheet over a polymer honeycomb, with porous mesh-caps bonded into the honeycomb cells to split the depth. Each mesh-cap can also have a unique DC flow resistance, thereby providing even more degrees of freedom in the design process. For this test, a combination of different cell types was used to create MDOF liners. An MDOF liner typically provides more attenuation than a SDOF liner over a wide frequency range, except at the specific frequencies for which the SDOF liner is designed. Design and implementation details for the SDOF and MDOF liners are provided in a separate report on liner design.⁴⁶ Three liner depths were used for the MDOF, one for the outer wall and two for the inner wall. The overall depths of each zone were chosen based on the available spacing in the inner and outer walls of the fan rig. In the outer wall, the maximum depth was approximately 1 in over the entire length of the treated section. The inner wall had two distinct regions. The maximum depth was 0.76 in over the forward portion, and 0.36 in over the aft portion. Those portions for which the available depth was larger allowed for a wider range of septa depths and, hence, more variability in the liner design.

9 × 15 Fan module demonstrator assessment

Attenuation of the interaction modes were measured by the Rotating Rake mode measurement system and then summed to determine the overall attenuation. The liner attenuation levels relative to the hardwall for three fan speeds and the first three harmonics are summarized in Figure 26. As most of the power is in these modes, this is a good approximation to the overall tonal power level (PWL) reduction. The advanced MDOF liner outperformed the traditional SDOF liner at all fan speeds and at the first harmonics by ∼2–4 dB PWL. The highest attenuation in mode power level occurred at blade passing frequency, 93%, rpm-c (8.8 dB for the SDOF liner and 13.1 dB for the MDOF liner). The predicted attenuation levels generally following the measured levels within the uncertainty range. More analysis can be found in reference.⁴⁷

Figure 26.

MDOF liner attenuation of the sum-of-interaction circumferential mode orders.

The metric important to commercial aircraft noise certification is the EPNL. For the results summarized here, straight and level flyover at an altitude of 1500 ft (457.2 m) at Mach 0.1 with one engine at a scale factor of 1 for the wind tunnel data were used in the computation of EPNL. Doppler shift and tone correction penalties were included. Further description is provided in reference.⁴⁹

Evaluating liner performance, the EPNL calculation shows a larger range of results than the sound power. Figure 27 has a comparison of the liner attenuation as a function of percent speed for both liners. Figure 27(a) shows the liner performance as a reduction in total sound pressure level from the hard-wall configuration. When measured by sound power the attenuation values were as much as 2.5 dB at 50% speed, dropping to 1.9 dB at 93% speed. Figure 27(b) has a comparison of the liner attenuation as a function of percent speed for both the Honeywell and NASA MDOF liners. The MDOF liner provided a decrease in EPNL of 3.9 dB at 50% speed, and 1.1 dB at 93% speed; with a nominal improvement of 0.5 to 1.0 dB over the SDOF liner for fan speeds up to 80%.

Figure 27.

MDOF/SDOF liner attenuation in the far-field (a) liner insertion loss – total sound power level (b) EPNL reduction due to liner attenuation.

This investigation showed that the MDOF liner concept can outperform a traditional, though expanded, SDOF liner in an annular duct with a realistic Mach number. The prediction methodology generally showed that to be the case, although the absolute levels showed a wider variation due to the statistical analysis used by the methodology.

Investigation on a static turbofan engine

A test on a small turbofan was conducted to achieve TRL 5.

DGEN Aeropropulsion research turbofan test description

The NASA Glenn Research Center’s DGEN Aeropropulsion Research Turbofan Test (DART) is based on the Akira Mecaturbines’ (formerly Price Induction) DGEN380. The DGEN380 is a geared two-spool, unmixed-flow turbofan, with a maximum static thrust of 560 lb_f (250 daN) and a high bypass ratio of 7.6. It is a small turbofan engine, with a length of 44.3 in (1.126 m), a fan inlet diameter of 13.78 in (350 mm), and a maximum diameter of 18.5 in (0.469 m) (at the exit plane of the bypass nozzle). The maximum overall pressure ratio is 5.3 (1.2 from the fan, 4.6 from the centrifugal compressor). Figure 28 shows the approximate dimensions of the DGEN380. The DGEN380 is integrated into the AAPL facility by mounting the engine and its associated pylon on a moveable platform. Figure 29 is a photo of the DART located in the center location of AAPL. Initial acoustic baselining of the DART was presented in references.^13,48–49

Figure 28.

DGEN380 turbofan engine dimensions.

Figure 29.

Photo of DART in AAPL.

Dart liner design and installation

The liner design process is described in a previous paper.⁵⁰ As indicated in that reference, the CDL propagation code was used to determine target optimum impedance values, at selected frequencies, for a given flow speed. With this being the first operational test of DART, limited information was available on DART source content and flow conditions. Therefore, the target values were obtained at one-third octave center frequencies ranging from 1000 to 8000 Hz at a flow condition expected to be nominally representative of 100% of design rpm-c (Mach # = 0.5 at the fan face).

To achieve the desired broadband performance, candidate liner designs consisted of multi-layer configurations incorporating septa (or “mesh-caps”) embedded into a honeycomb core.³⁶ The honeycomb cells were allowed to contain up to two mesh-caps for this investigation. The mesh-cap depths and DC flow resistances were allowed to vary from cell to cell. The liner modeling tools were used to obtain predicted design impedance values.

A number of MDOF liner concepts were considered. Because of the small diameter of the DGEN380 fan (∼14 inches), it becomes a significant challenge to bend the core to achieve a flush fit. This limited some of the choices for manufacturing of the liners. To gain insight that could be used to better optimize liners for future tests with this engine, two liner concepts were selected for this study to allow acquisition of sufficient data to demonstrate the validity of the liner design process. Figure 30 is a close-up photo of the liner cores. The first, shown in Figure 30(a), is a liner that contains a septum with a resistance sufficiently high to effectively eliminate sound transmission into the lower chamber. It should be considered as an SDOF liner from an acoustics point of view. The second, shown in Figure 30(b) is a 3DOF liner that contains two septa in each honeycomb chamber. Each septum has a unique resistance and location within a chamber, and the same distribution is used in each chamber. Figure 30(c) shows the fully assembled liner cores.

Figure 30.

Photos of DART liner cores. (a) SDOF liner core. (b) 3DOF liner core. (c) Assembled liner cores.

An inlet spool was designed and manufactured by the DGEN380 vendor, Akira Mecaturbines, to accommodate the liner cores. The liner cores were placed in a liner-holder spool. Figure 31 is a drawing illustrating the liner spool holder in the DGEN380 inlet. Figure 32 is a photo of the extension spool installed in the inlet of the DGEN380 engine.

Figure 31.

Schematic of the inlet spool holder installed in the inlet of the DGEN380.

Figure 32.

Photo of DART in the center of AAPL: hardwall inlet extension shown.

The hardwall inlet insert was tested first to re-baseline the acoustic levels resulting from the additional inlet length. The paradigm was to evaluate the insertion losses of the liners and compare them to the predictions as a validation of the concept and design tools, not necessarily as a direct proof of greater efficacy of the 3DOF to SDOF liners. It is the view of the authors that the 3DOF liner design concept is more suited to the exhaust acoustic/flow environment, but the aforementioned restrictions resulted in the selection of the inlet for this initial study. The two liners were then tested sequentially.

The effects of source distribution on predicted directivity were also illustrated through the flexibility of the statistical source model. As part of the comparison process, the overall design and evaluation capability was extended such that external observer locations may be included in the optimization process. This enhancement provides a much wider design space in designing advanced broadband liners.

Dart assessment

While the liner design point Mach number fell outside the test range, a great deal of information remains to be gained through a comparison of predicted and measured results.^51,52 The directivity of PWL loss is shown in Figure 33: for the SDOF for the first (a) and second (b) fan harmonics; and for the 3DOF liner for the first (c) and second (d) harmonics. The integrated PWL loss is shown in a bar graph in the upper left of the figure. The 3DOF liner provided slightly improved performance for the conditions considered. The predicted attenuation trend matched the experimentally measured results at 1xBPF but with a significant underprediction. Over the rpm range tested, the integrated PWL loss prediction vs experimental measurement comparisons at 2xBPF is much better, within 1 dB. The design process did create a liner that was predicted to have greater attenuation at the frequencies around BPF, ∼1–2 kHz (but also had some attenuation up to 8 kHz) as shown in Figure 34, and this was verified.

Figure 33.

DART SDOF/3DOF measured vs predicted attenuation on the 10-ft. arc array. (a) SDOF Liner @ 1xBPF (b) SDOF Liner @ 2xBPF (c) 3DOF Liner @ 1xBPF (d) 3DOF Liner @ 2xBPF.

Figure 34.

Prediction attenuation loss vs frequency.

The insertion tonal PWL loss of each liner for the range of rpm tested is shown in Figure 35. This is obtained by integrating over the forward microphone array, weighted by the subtended area. The total PWL insertion loss for the 3DOF liner is generally higher at the lower fan speeds (less than 70% RPMc), mostly due to the variation in tonal attenuation which is 2.8< ΔPWL< 4.6 for the SDOF liner and 3.8< ΔPWL< 5.4 for the 3DOF liner. The broadband PWL insertion loss is consistent between the two liners across the rpm range (SDOF:2.4< ΔPWL< 2.9; 3DOF:2.8< ΔPWL< 3.4). These values, as well as a detailed analysis of the core tone components and fan-core interactions are presented in^58,59

Figure 35.

DART SDOF/3DOF insertion losses vs. RPM.

These results provide further confidence in the design tools, as well as the enhancements made to the overall liner design process including the enhancement of projecting results to the far-field. The capability of the manufacturing process and the efficacy of the 3DOF liner in a representative environment was demonstrated.

Investigation on aircraft flight

A flight test was successfully conducted to achieve a TRL 7 demonstration of the MDOF liner concept. While the MDOF design tools were utilized to generate the liner characteristics, the result was more properly a 3DOF, i.e. all chambers within a layer were identical. The term MDOF is being used here to maintain consistency with earlier papers. The flight test was conducted at Moses Lake, WA, between 27 July and 6 August 2018. This effort was reported in public sources.^53–55

Flight test article description

A CFM International Leading Edge Aviation Propulsion (LEAP) 1B turbofan engine⁵⁶ was installed on the #2 engine (right location, from the pilot’s view) of a Boeing 737–7 narrow body airliner. The CFM LEAP 1B is a high-bypass ratio (9:1) turbofan generating approximately 30,000 lbs (130 kN) of thrust from a ∼70 in (175 cm) diameter fan.

The 737 MAX inlet had been in production for about three years at the time of this test. The airplane identified for this test was provided with two production inlets. The inlets went through the typical fabrication and inspection processes. No modifications were made to the acoustic treatment. The acoustic liner consists of circular perforations over a two-degree-of-freedom honeycomb core. Figure 36(a) shows the LEAP engine as a baseline and Figure 36(b) shows the NASA MDOF inlet liner build.

Figure 36.

Photos of 3DOF inlet configurations (a) baseline (b) NASA MDOF low-drag inlet.

Flight test liner design and installation

NASA led the conceptual design of the MDOF acoustic liner and supplied the acoustic requirements for the honeycomb core with dual septa and lower drag slotted facesheet. The design process to optimize the acoustic liner impedance is described in reference.⁵⁷ The acoustic liner optimization process aimed to improve broadband fan noise reduction relative to standard production liners. The design requirements were fine-tuned such that the hardware would meet manufacturability requirements per the Boeing Company and the aggressive test schedule.

The MDOF core was fabricated by the Hexcel Corporation. The core design contains two layers of mesh septa within each honeycomb cell. Figure 37 shows a sketch of the acoustic core design parameters and an example cross-section of the fabricated MDOF core. The combined system of septa has a total flow resistance of 190 CGS Rayl at 105 cm/s.

Figure 37.

3DOF liner concept used on LEAP. (a) Schematic of 3DOF septa configuration. (b) Photo of 3DOF septa configuration.

There were three inlet configurations of interest tested for this project: 737 MAX Production inlet, NASA MDOF low drag inlet, and a hardwall (taped) inlet. All inlets were without discontinuities, except at the small fan inlet temperature probe installation location.

Flight test assessment

The flight test methodology is detailed in reference.⁵⁸ The fan tone PWL in the forward arc (30°-80°) is shown in Figure 38 as a function of engine fan rpm-c. Relative to the hardwall inlet, the MDOF liner achieves 6 dB reduction at low speed and up to 12 dB at the higher fan speeds. The repeat points show good repeatability for both configurations. At approach power fan rpm-c settings, the fan tone PWL reduction realized was approximately 4 dB.

Figure 38.

Total forward arc fan tone power level comparison between NASA inlet and hardwall inlet (from reference⁵⁹). (a) Takeoff power (b) Approach power.

The broadband analysis required a combination of far-field microphone and phased array data. The phased array details are provided in reference.⁵⁹ A component separation method was utilized on the data acquired for each configuration to obtain the broadband attenuation attributed to the liner. Details are provided in the two prior references.

The isolated inlet component and airplane EPNL reductions are shown in Figure 39 for the approach, cut-back, and sideline flight conditions. The highest reduction achieved is at the sideline flight condition, with 1.9 dB and 0.4 dB EPNL reductions attributed to the inlet, and aircraft, respectively. The cumulative EPNL reductions achieved by the MDOF liner are 3.2 dB (inlet component) and 0.7 dB (total aircraft) as shown in Table 4.

Figure 39.

EPNL reduction on sample 737MAX to production on an inlet component basis (from reference⁵⁹).

Table 4.

3DOF liner cumulative benefit.

Condition	CUM (inlet component EPNL	CUM (airplane EPNL)
Benefit (re-production liner)	3.2 EPNdB	0.7 EPNdB

The NASA MDOF low-drag inlet liner flight test resulted in a measurable reduction in airplane noise on the 737 MAX. The design and prediction methodology, was validated on a flight test with a production aircraft. There were a number of constraints that limited the retrofitting that could be implemented on the 737 MAX and the flight test development schedule. There is potential for better acoustic performance in a future airplane liner design if optimization of core depth, facesheet, or inlet area could be conducted, as is the case with conceptual aircraft. Potentially, applications of this concept will likely have even greater benefit to future aircraft designs that are expected to have relatively lower jet, airframe, and aft fan noise sources.

Summary

A series of increasing Technology Readiness Level investigations of Multi-Degree-of-Freedom liners were conducted. These investigations centered on a liner design for a specific test rig and were coupled to the development and validation of a liner design process that combined several tools. The liner design process was improved and the efficacy of the MDOF liner concept was demonstrated in increasingly realistic environments, as well as enhancing the manufacturing technology required for MDOF construction. The investigations culminated in a flight test where an MDOF liner was installed on a production turbofan mounted on a commercial airliner, during which inflight acoustic reduction was measured. The potential for introduction into the commercial airline fleet is high, both as a retrofit and integration into a new aircraft design.

It was also noted that as the TRL increases the amount of measured acoustic reduction decreases due to added complexity and the presence of other sources. By testing at lower TRL, the direct effect of the concept under development can be measured and is more useful to validate codes and provide physical insight. The TRL progression in this paper showed a reduction on the order of 50 dB SPL in low-TRL test articles, to 15 – 5 dB PWL over the mid-TRL range, and ultimately a reduction on the order of 2 EPNdB component (∼0.5 EPNdB total) in a high-TRL test flight. This is generally expected and should be taken into account during the research and development process.

Coupling the prediction and test evaluation process ensured stronger confidence in the methodology and provides support for wider applicability. That is, one-off designs are avoided, and technical risk is mitigated. This process and philosophy documented herein should be considered a template for successful technology development. The support of several concurrent and consecutive NASA ARMD programs, inter-center, and academia-industry-government co-operation is also a model to be emulated.

Footnotes

Acknowledgements

This work was performed with support from several Aeronautics Research Mission Directorate programs (and underlying projects) over the 10-year period: The Fundamental Aeronautics Program (Fixed Wing and the Subsonic Fixed Wing Projects), the Advanced Air Vehicles Program (Advanced Air Transport Technology Project), and the Integrated Systems Research Program (Environmentally Responsible Aviation Project). The authors would like to express their appreciation for the efforts of the NIT/GFIT, AAPL, and 9x15 LSWT staffs for their expertise and dedication. Vantage Partners, LLC provided design integration for several of the liners. Hexcel, Inc was a key to the liner core development and manufacturing. A contract with Price Induction provided the liner holder design and analysis. Studies referenced in this report were conducted as part of Space Act Agreements between NASA and Hexcel Corporation, and between NASA and Honeywell.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Daniel L Sutliff

Douglas M Nark

Michael G Jones

References

Huff

DL.

Noise reduction technologies for turbofan engines. NASA/TM-2007-214495.

Hubbard

(ed.). Aeroacoustics of flight vehicles: theory and practice: noise sources, NASA Reference Publication 1258, Vol. 1, WRDC Technical Report 90-3052, Acoustical Society of America, Woodbury, NY, USA.

Hubbard

(ed.). Aeroacoustics of flight vehicles: theory and practice: noise control, NASA Reference Publication 1258, Vol. 2, WRDC Technical Report 90-3052, Acoustical Society of America, Woodbury, NY, USA.

Woodward

Hughes

Jeracki

, et al. Benefits of swept and leaned stators for fan noise reduction. NASA/TM-1998-208661.

Jones

Watson

Nark

, et al. A review of acoustic liner experimental characterization at NASA Langley. NASA/TP-2020-220583.

McAllister

Loew

Lauer

, et al. The advanced noise control fan baseline measurements. NASA/TM-2009-215595, also AIAA-2009-624.

Loew

Lauer

McAllister

, et al. The advanced noise control fan. NASA/TM-2006-214368, also AIAA-2006-3150.

Sutliff

DL.

The advanced noise control fan: a 20 year retrospective of contributions to aeroacoustics research. NASA/SP-2019-643.

Arrington

Gonsalez

Calibration of the NASA Lewis Research Center 9- by 15-foot low speed wind tunnel. NASA CR-195438, 1994.

10.

Arrington

Gonsalez

Flow quality improvements in the NASA Lewis Research Center 9- by 15-foot low speed wind tunnel. NASA CR-195439, 1995.

11.

Dahl

Woodward

RP.

Acoustical evaluation of the NASA Lewis 9- by 15-foot low speed wind tunnel. NASA TP-3274, 1992.

12.

Balan

Hoff

GE.

Propulsion simulator for high bypass turbofan performance evaluation. SAE Technical Paper Series 931410, 1993.

13.

Sutliff

DL.

Acoustic directivity of the DGEN aero-propulsion research turbofan at multiple far-field array locations. AIAA-2018-3279.

14.

Brown

Sutliff

DGEN aeropropulsion research turbofan (DART): lossless projection of measured engine noise spectra to a 1-Foot-Radius arc. AIAA-2018-3280.

15.

www.boeing.com/commercial/737max/ (accessed 27 July 2020).

16.

Nark

Jones

MG.

Broadband liner optimization for the source diagnostic test fan. AIAA-2012-2195.

17.

Howerton

Jones

Buckley

JL.

Development and validation of an interactive liner design and impedance modeling tool. AIAA-2012-2197.

18.

Koch

Jones

MG.

Measurement of the acoustic absorption of natural reeds and comparison to aircraft engine acoustic liner prototypes. NASA TM-2016-218851.

19.

Sutliff

Jones

MG.

Low-speed fan noise attenuation from a foam-metal liner. AIAA J Aircr 2009; 46: 1381–1394.

20.

Gazella

Takakura

Sutliff

, et al. Evaluating the acoustic benefits of over-the-rotor acoustic treatments installed on the advanced noise control fan. AIAA-2017-3872.

21.

Bozak

Dougherty

RP.

Measurement of noise reduction from acoustic casing treatments installed over a subscale high bypass ratio turbofan rotor. AIAA-2018-4099.

22.

Sutliff

Jones

Hartley

TC.

High-speed turbofan noise reduction using foam-metal liner over-the-rotor. AIAA J Aircr 2013; 50: 1491–1503.

23.

Koch

LD.

NASA’s bio-inspired acoustic absorber concept. In: Summit 2017: nature-inspired exploration for aerospace/biomimicry summit, 4–6 October 2017, Cleveland, OH, USA.

24.

Koch

LD.

Investigation of a bio-inspired liner concept. Acoustics Technical Working Group, 11–12 April 2017, Hampton, VA, USA.

25.

Ingard

On the theory and design of acoustic resonators. J Acoust Soc Am 1953; 25: 1037–1061.

26.

Blackstock

DT.

Fundamentals of physical acoustics. Hoboken: Wiley, 2000.

27.

Parrott

Jones

MG.

Parallel-element liner impedances for improved absorption of broadband sound in ducts. Noise Control Eng J 1995; 43: 183–195.

28.

Motsinger

Kraft

RE.

Design and performance of duct acoustic treatment: aeroacoustics of flight vehicles; Chapter 14, Vol. 2: noise control. NASA RP 1258, August 1991.

29.

Jones

Parrott

Watson

WR.

Uncertainty and sensitivity analyses of a two-parameter impedance prediction model. AIAA-2008-2928.

30.

Parrott

Jones

MG.

Assessment of NASA’s aircraft noise prediction capability, Chapter 6: uncertainty in acoustic liner impedance measurement and prediction. NASA TP 2102-215653, July 2012.

31.

Nark

Farassat

Pope

, et al. The development of the ducted fan noise propagation and radiation code CDUCT-LaRC. AIAA-2003-3242.

32.

Dougherty

RP.

A wave-splitting technique for nacelle acoustic propagation. AIAA-97-1652.

33.

Dougherty

RP.

A parabolic approximation for flow effects on sound propagation in nonuniform, softwall, ducts. AIAA Paper 99-1822.

34.

Lan

JH.

Turbofan duct propagation model. NASA CR 2001-211245.

35.

Watson

Jones

Parrott

TL.

Validation of an impedance eduction method in flow. AIAA J 1999; 37: 818–824.

36.

Jones

Howerton

Ayle

Evaluation of parallel-element, variable-impedance, broadband acoustic liner concepts. AIAA-2012-2194.

37.

Jones

Nark

Baca

, et al. Applications of parallel-element, embedded mesh-cap acoustic liner concepts. AIAA-2018-3445.

38.

Cooper

BA.

A large hemi-anechoic chamber enclosure for community-compatible aeroacoustic testing of aircraft propulsion systems. Noise Control Eng J 1994; 42, 1–7(7).

39.

Sutliff

Walker

BE.

Artificial noise systems for parametric studies of turbo-machinery aero-acoustics. Int J Aeroacoust 2016; 15: 103–130.

40.

Nark

Jones

Sutliff

DL.

Modeling of broadband liners applied to the advanced noise control fan. AIAA 2015-2693.

41.

Nark

Jones

Sutliff

DL.

Improved broadband liner optimization applied to the advanced noise control fan. AIAA-2014-3103.

42.

Sutliff

Jones

Nark

DM.

In-duct and far-field experimental measurements from the ANCF for the purpose of improved broadband liner optimization. AIAA-2014-3231.

43.

Arrington

Gonsalez

Flow quality improvements in the NASA Lewis Research Center 9- by 15-foot low speed wind tunnel. AIAA-95-2390, 1995.

44.

Balan

Hoff

GE.

Propulsion simulator for high bypass turbofan performance evaluation. SAE Technical Paper Series, 931410, 1993.

45.

Miller

Stephens

Sutliff

DL.

Acoustic testing of a high tip speed fan with bypass-duct liners. NASA/TM-2019-220162.

46.

Nark

Jones

Sutliff

DL.

Further development and assessment of a broadband liner optimization process. AIAA-01-2784.

47.

Sutliff

Nark

Jones

MG.

Efficacy of a multiple degree of freedom acoustic liner installed in the bypass of a scale model high speed fan. AIAA-2016-3005.

48.

Sutliff

Brown

Bayon

, et al. Far-field acoustic characteristics of the DGEN380 turbofan engine as measured in the NASA Glenn aero-acoustic propulsion laboratory. AIAA-2016-3006.

49.

Brown

Sutliff

DL.

DGEN aeropropulsion research turbofan (DART): projecting acoustic data to 1-foot-lossless. AIAA-2018-3280.

50.

Nark

Jones

Schiller

, et al. Broadband inlet liner design for the DGEN aeropropulsion research turbofan. AIAA-2018-3608.

51.

Sutliff

DL.

Acoustic directivity and insertion loss measurements of advanced liners installed the inlet of the DGEN aeropropulsion research turbofan. NASA/TM-2019-220161.

52.

Sutliff

Nark

Jones

, et al. Design and acoustic efficacy of a broadband liner for the inlet of the DGEN aero-propulsion research turbofan. AIAA-2019-2582.

53.

Francis S, Acousti-Cap technology from Hexcel reduces noise in flight test, CompositesWorld, www.compositesworld.com/news/acousti-cap-technology-from-hexcel-reduces-noise-in-flight-test (accessed 27 July 2020).

54.

Quiet Quest. Aviation week & space technology, 13–16 September 2018.

55.

Norris G , Boeing-NASA Low-Drag, Low-Noise Engine Feature Beats Expectations. https://aviationweek.com/air-transport/boeing-nasa-low-drag-low-noise-engine-feature-beats-expectations (accessed 27 July 2020).

56.

www.cfmaeroengines.com/engines/leap/ (accessed 27 July 2020).

57.

Nark

Jones

MG.

Design of an advanced inlet liner for the quiet technology demonstrator 3. AIAA-2019-2764.

58.

Wong

Nesbitt

Jones

, et al. Flight test methodology for NASA advanced inlet liner on 737MAX-7 test bed (quiet technology demonstrator 3). AIAA-2019-2763.

59.

Brusniak

Wong

Jones

, et al. Acoustic phased array analysis of quiet technology demonstrator 3 advanced inlet liner noise components. AIAA-2019-2765.