Servizio Comunitario di Informazione in materia di Ricerca e Sviluppo - CORDIS

FP7

ALLEGRA Sintesi della relazione

Project ID: 308225
Finanziato nell'ambito di: FP7-JTI
Paese: Ireland

Final Report Summary - ALLEGRA (Advanced Low Noise Landing (Main and Nose) Gear For Regional Aircraft)

Executive Summary:
The ALLEGRA project has been developed in response to the requirements of the European Clean Sky Joint Technology Initiative to assess low noise technologies applied to both nose and main landing gear architectures. ALLEGRA consisted of a consortium of well recognized universities (Trinity College Dublin, KTH Sweden), a well-known aeroacoustic wind tunnel company (Pininfarina SPA) and SME partners (Eurotech, Teknosud) from around Europe supported by a leading landing gear manufacturer (Magnaghi Aeronautica). This group has well demonstrated competencies in landing gear design and manufacturing, computational aeroacoustics, noise measurements and data analysis. The negative impact of aircraft noise includes effects on population’s health, land use planning and economic issues such as building restrictions and operating restrictions for airports. Thus, the reduction of noise generated by aircraft at take-off and approach is an essential consideration in the development of new commercial aircraft.

The key issues associated with the landing gear noise problem are:
• Landing gear contributes to approximately 30% of the overall noise emission of the aircraft during approach phase
• The noise signature is broadband in nature covering frequencies from approximately 90Hz to 4KHz. The annoyance level associated with noise within this frequency range is high for exposed communities
• Landing gear consists of numerous structures, surfaces and components which are generally not optimised from an aeroacoustic point of view. Turbulence from these non-optimised components of the landing gear is a direct noise source
• The wake of the landing gear structures can interact with other airframe components and generate an indirect noise source.

In the past full-scale models of landing gear have rarely been tested due to the large test facilities required. Most experimental airframe noise research has been performed using small-scale models. This leads to great difficulty when using model-scale results for full-scale noise predictions due to the lack of details in the geometrical modelling. One of the significant contributions of ALLEGRA is that a full representation of the landing gear detail and associated structures (e.g. bay cavity, bay doors, belly fuselage etc.) have been included and addressed at a realistic scale. The nose landing gear is designed at full scale and the main landing gear at half scale.

All technical and scientific objectives of the project have been met. The results of the two wind tunnel test campaigns have been analysed and the best performing low noise technologies ranked and identified.

Project Context and Objectives:
ALLEGRA was developed in response to the requirements described in the Clean Sky- ITD-GRA call for proposals referenced under SP1-JTI-CLEAN SKY-2011-3 and named “Advanced low noise Main and Nose Landing Gears for Regional Aircraft -Trade off concept studies, large-scale mock-ups design manufacturing & WT testing”. ALLEGRA consists of three work packages, one for each of the landing gear to be tested and a management work package.
The objectives of the activity under this CFP are the conceptual design and experimental validation of low-noise devices for both main and nose landing gear architectures. Complete landing gear architectures have been considered, including the gear strut, wheel pack, bay cavity, bay doors and belly fairing. On the basis of baseline architectures (CAD models) provided by the GRA member, the Partners provided input to a GRA down selection of potential low noise technologies to be investigated in the ALLEGRA test campaigns. Following this down selection process the Partners designed and manufactured two large scale wind tunnel models of both the main and nose landing gear. The test articles were manufactured to be realistic representations of the complete gears architectures as above described (gear strut, wheels, bay, doors, fairings).
The Partners then performed aero-acoustic wind-tunnel tests on both the above landing gear test models installed in an experimental aero-acoustic facility to assess the selected low-noise solutions against respective baseline/benchmark configurations. Tests were performed in the Pininfarina aeroacoustic facility open jet semi-cylindrical wind tunnel in Turin, Italy, that holds a test section of 8m x 9.60m x 4.20m. The facility contains a low noise high speed fan-drive system of 13 fans in order to increase the wind speed up to 260 km/h and reduce the background noise level to 68 dBA at 100 Km/h. The complete test data and low noise technology assessments were released to the GRA in the final test reports.

Project Results:
Reference surfaces of the landing gear and fuselage sections were provided to the consortium by the topic manager. In order to design a model suitable for wind tunnel testing considerable work to optimize and implement missing parts was required. The input received included scaled 1:1 surfaces of an half part of a section of fuselage in the region of the MLG. The MLG model therefore required scaling to 1:2 and a mirroring of the missing surfaces to complete the model. The final wind tunnel models reproduces simplified external lines of the Advanced Turboprop lower fuselage and main landing gear fairing, and main landing gears in extended position; a 4° angle of attack is implemented to take account of the test requirements.

Tests were performed in the Pininfarina aeroacoustic facility which is an open jet semi-cylindrical wind tunnel in Turin, Italy, that holds a test section of 8m x 9.60m x 4.20m. The facility contains a low noise high speed fan-drive system of 13 fans in order to increase the wind speed up to 260 km/h and reduce the background noise level to 68 dBA at 100 Km/h. The velocity produced by the wind tunnel is very uniform, since it varies by only 0.5% over the area. The turbulence level is 0.3%. The instrumentation deployed in the test campaigns is briefly described here.

Four planar microphones arrays were installed inside the wind tunnel: a 3 meter diameter half-wheel array of 66 microphones positioned on one side, parallel to the model axis and at a distance of 4.22m from it, a far field linear array of 10 microphones on the same side and at the same distance from the axis model, a 78 microphone 3 meter diameter wheel array on the ceiling at a distance of 1.82 meters from the model and a spiral front array of 15 microphones placed upstream the landing gear plane and at an angle of 10 degrees to the same plane. Data were acquired for 10 seconds at a sampling rate of 32768 Hz.

Datasets from all sensors sampled at 32768 Hz for time duration of 10 seconds are processed using NI Sound and Vibration toolkit incorporated into custom software in Labview. For each sensor a 1/3 octave band analysis between 20 Hz and 10 kHz is performed using both linear and A-weighting filters in accordance with IEC 1260:1995 and ANSI S1.11-2004 standards.

"A numerical simulation of the test-set up was performed to compliment the experimental investigation and the results were compared and contratsed." A brief summary of the project results is provided here.

An identical test set up in terms of sensors and wind tunnel operating conditions were used for the NLG and MLG tests.

WP1 - MLG

The noise sources produced by the extended main landing gear are broadband in nature but considerable energy is found at low frequencies of between 100-200 Hz. While the sound produced by the extended landing gear can be as much as 12dB above the sealed fuselage in certain octave bands the increase in OASPL is in the region of 7dBA for most emission angles and flow speeds.

The low noise technologies ML4 through to ML8 consist of distinct approaches to noise source reduction. They affect different areas and components of the landing gear and as such may be applied either in isolation or in combination. ALLEGRA looked at the individual performance of the low noise technologies compared to the baseline case. No combination technologies were planned as part of the MLG test campaign.

ML4 - Perforated leg fairing
ML5 - Bay cavity absorber
ML7 - Mesh
ML8 - Hub caps and wheel axle fairing

For the purposes of the ranking the individual technologies ML4, ML5, ML7 and ML8 are considered and the modified bay opening FOB is included for comparison purposes. For each model configuration the information from the Linear Far Field array was entered into a ranking database. The performance of each configuration in terms of minimum noise emission was ranked both for OASPL and as function of octave band using linear weighting. These rankings were then summed for all flow speeds and yaw angles.

Following this ranking procedure in each octave band a global ranking which considers all octave bands was generated. From this it was clear that the technology ML7 was the best performing technology considering both frequency and emission angle along the Linear Far Field array. The second best technology was ML8 which was a better performer at the lower frequency octave bands between 400Hz and 1250Hz. It is interesting to note that these technologies could be applied in combination since they do not overlap in location on the landing gear.

In order to further evaluate the directivity of the noise generated by the landing gear, spectrograms as a function of frequency and emission angle were generated using the data of the Linear Far Field Array. Data in the range of 50 Hz - 10000 Hz from the 1/3 octave band spectra were used. These results showed that the MLG sources are greater than 8dB over a wide range of frequencies and angles with a peak of up to 14dB in the 100Hz octave band at all emission angles. This is due to a very strong bay cavity tone at this frequency.

The numerical analysis can shed further light on the source mechanisms of the MLG. The numerical sources can be compared to the outputs of the experimental beamforming results which also localize sources on the landing gear components. Landing gear is a complex system with many components which are capable of generating noise.

MLG sources were investigated from a numerical perspective. In the framework of the Ffowcs Williams-Hawkings (FWH) integral method, a source surface (or volume) and a receiver are defined, and the contribution of the source terms over the surface/volume are integrated to compute the pressure at the far-field receiver. In the receiver timeframe, the pressure signal at the receiver is constructed by summing the contributions of each computational cell covering the source surface (different cells of the surface contribute to the receiver at different times). Therefore, for a given time t in the receiver timeframe, the contribution of each cell of the source surface can be computed, such that the integral of these contributions over the surface yield the fluctuating pressure at the receiver at time t.

There are some strong sources present in the CAA results on the shock absorber and folding stay likely due to the wake from the main leg. Both of these sources have been beneficially treated by the additions of the leg fairing.

A time domain delay-and-sum beam forming algorithm with microphone self-noise removal and shear layer correction is used to process the data. The top array has been used to scan the noise sources in each 1/3 octave band utilizing band pass filtering of the data prior to application of the beam forming algorithm. The algorithm was capable of detected and identifying the noise sources generated by various components and allowed comparisons with the numerical results.

WP2 - NLG

The noise sources produced by the extended nose landing gear are broadband in nature but considerable energy is found at low frequencies of between 200-300 Hz. While the sound produced by the extended landing gear can be as much as 12dB above the sealed fuselage in certain octave bands the increase in OASPL is in the region of 5dBA for most emission angles and flow speeds.

The low noise technologies NL1 through to NL4 consist of distinct approaches to noise source reduction. They affect different areas and components of the landing gear and as such may be applied either in isolation or in combination. This section looks at the individual performance of the low noise technologies compared to the baseline case.

NL1 - Ramp Door Spoiler
NL2 - Wheel axle wind shield
NL3 - Wheel hub caps
NL4 - Perforated fairings

NL1 is the ramp door fairing and shows clear reductions of 2-3dBA at upstream angles to the landing gear, greater than any other individual technology. At downstream angles the differences between the low noise technologies are less clear with all achieving some reduction with compared to the baseline.

NL5 combined the technologies NL2, NL3 and NL4. NL2 consisted of a wheel axle wind shield where as NL4 consisted of multiple perforated fairings including one in the region of the wheel axle. In order to allow for the combination of these technologies to be tested the wheel axle fairing of NL4 was excluded due to the identical placement of the component with the NL2 technology.

Considering the OASPL of the three component technologies as well as the combination technology it can be seen that the performance of NL5 does not achieve an additive benefit of the component technologies. As a general observation it achieved better A-weighted noise reduction than NL3 and NL4 but is equivalent to or worse than the isolated NL2 technology at many emission angles for the three flow speeds.

NL6 combined the technologies NL1, NL2 and NL3. The technology NL1 achieved the greatest A-weighted noise reduction of the individual technologies. NL2 achieved the second greatest noise reduction of the individual technologies. Due to this there is an expectation that NL6 will perform well. NL6 achieves a clear additive benefit of the technologies. The A-weighted noise reduction is 1-2dBA greater than any individual technology and is achieved over the full range of emission angles. Of the two combination technologies tested it seems that only NL6 has achieved an increase in noise reduction over the component technologies.

For the purposes of the ranking the individual technologies NL1, NL2, NL3 and NL4 as well as the combination technologies NL5 and NL6 are considered. For each model configuration the information from the Linear Far Field array was entered into a ranking database. The performance of each configuration in terms of minimum noise emission was ranked both for OASPL and as function of octave band using linear weighting. These rankings were then summed for all flow speeds and yaw angles.

Following this ranking procedure in each octave band a global ranking which considers all octave bands was generated. From this it was clear that the combination technology NL6 was the best performing technology considering both frequency and emission angle along the Linear Far Field array. The majority of the noise reduction was achieved by NL1 which ranked second overall.

The comparison of the experimental and numerical analysis can shed further light on the source mechanisms of the NLG. The numerical sources can be compared to the outputs of the experimental beamforming results which also localize sources on the landing gear components. The nose landing gear is a complex system with many components which are capable of generating noise. A time domain delay-and-sum beam forming algorithm with microphone self-noise removal and shear layer correction is used to process the data. The side array has been used to scan the noise sources in each 1/3 octave band utilizing band pass filtering of the data prior to application of the beam forming algorithm. The calculation plane is parallel to the related microphone array, at a distance of 4.22 m from it, and is centered on the landing gear vertical axis at the midpoint of the wheel axle.

Considering the numerical results achieved by the Ffowcs Williams-Hawkings (FWH) integral method, a source surface (or volume) and a receiver are defined, and the contribution of the source terms over the surface/volume are integrated to compute the pressure at the far-field receiver. In the receiver timeframe, the pressure signal at the receiver is constructed by summing the contributions of each computational cell covering the source surface (different cells of the surface contribute to the receiver at different times). Therefore, for a given time t in the receiver timeframe, the contribution of each cell of the source surface can be computed, such that the integral of these contributions over the surface yield the fluctuating pressure at the receiver at time t. This gives a measure of the contribution (per unit area) to the pressure fluctuations at the receiver for each cell of the source surface.

The results on the full model (landing gear + fuselage) highlight the location of the main sources for the receiver. The main sources can be observed on the wheels, the wheel axle, the pinion steering, the doors and a small region on the fuselage downstream of the LG assembly. The structure of the source distribution is very similar for the 50m/s and 65m/s cases, with higher amplitude of the sources at the higher velocity case.

The CAA results on the LG assembly show that the sources on the tyres and torque links are significantly modified by axle shield fairing (NL2) and by the hub caps (NL3). Two remarkable features:
- Both NL2 and NL3 dramatically reduce the source contributions on the lower part of the torque links (but NL2 imposes a penalty on the higher part).
- The source distribution on the tyres becomes asymmetric with NL3, probably due to the asymmetry of the torque links which create a blockage effect between the inner hub caps.
The sources are concentrated on the downstream part of the LG assembly, where the flow is separated. The results focusing on the torque links show that the fairing NL2 shields the lower part of the torque links (but evidences a penalty on the upper torque link). NL3 also reduces the sources on the lower torque link, but less than NL2. The axle shield is also successful at reducing the noise sources on the wheel axle, and parts of the inner tyre. NL3 suppresses the influence of the rim cavity, although new surface sources appear on the hub caps.

The same conclusions regarding the effects of NL2 and NL3 on the surface source distribution hold for the case at 65m/s and zero yaw. These conclusions from the numerical analysis are supported by experimental results for NL2 and NL3. In the experimental results we can see that the sources associated with the torque link in the 1250Hz to 2500Hz bands have reduced by at least 3dB and the other sources associated with the doors and other components have become the dominant sources. NL3 also reduced the 800Hz band level by 1dB which is primarily associated with the wheel region.

Potential Impact:
The main results of the project will be published in the scientific journals and presented at conferences. The consortium has an excellent track record for publication in the premium journals relating to aero-acoustics. The work within this project has and will continue to generate research publication within the fields of experimental aero-acoustics, noise source identification and novel noise reduction technologies. This will produce a number of international conference level publications and the AIAA/CEAS conferences on aero-acoustics have been identified as promising events for dissemination during the project. The initial results of the numerical analysis were presented at this conference in Atlanta, June, 2014 with follow on results presented in Dallas, June 2015.

The work will also produce journal level publications suitable for the highest level research publications. The track record of the consortium for presentation at the AIAA and the publication history will insure the success of these activities. TCD will utilise its role as a national contact point within the X-Noise thematic network to promote and disseminate the outputs of this research project to the European aero-acoustics community. TCD will also present the outputs of the work at the INTERNoise2015, San Francisco, Conference and the 2015 X-Noise/CEAS workshop namely the 19th Workshop of the Aeroacoustics Specialists’ Committee of CEAS, simultaneously 5th Scientific Workshop of the European X-Noise EV Network, dedicated to “Broadband Noise of Rotors and Airframes”.

The results of the ALLEGRA project will integrate with research activities in the wider Clean Sky and Green Regional Aircraft programs and pave the way for flight testing of the demonstrated low noise technology. In addition, the project will serve to sustain the momentum of a number of national programmes in Ireland, Sweden and Italy from which the consortium partners have been drawn. This has led to the further development of specific expertise in advanced experimental techniques, numerical modelling and sound propagation. Interaction and exchange of knowledge between the project partners supports the development of a synergistic approach to the design of novel airframe concepts and the development of new analysis tools for the European aerospace community.

List of Websites:
Project Coordinator:
Trinity College Dublin
Dr. Gareth J. Bennett
gareth.bennett@tcd.ie

Informazioni correlate

Documenti e pubblicazioni

Contatto

Deirdre Savage, (Research Accounting Manager)
Tel.: +35318961942
Fax: +35317071633
E-mail
Numero di registrazione: 182116 / Ultimo aggiornamento: 2016-05-17
Fonte d'informazione: SESAM