Periodic Reporting for period 3 - DARETOMODEL (DAta-dRivEn, low-rank, jeT-nOise MODELling)
Okres sprawozdawczy: 2021-10-01 do 2022-09-30
Aircraft noise constitutes a pressing environmental issue that continues to pose a problem for urban planning in densely populated regions near airports, and for the aircraft industry. One of the dominant sources of sound is associated with the propulsive jets. Computation of this sound requires computational power beyond that which is presently available. Simpler models are necessary, but these should be based on the correct underlying physics. Despite over 60 years of jet-noise research, some of the most basic questions regarding these physics remain unanswered, and it is for this reason that an efficient, reduced-complexity, robust modelling framework has not yet been obtained. Recent progress in simulation and experimental diagnostics has opened up new possibilities. Data-mining has been used, with success, to clarify the mechanisms that underpin both free and installed jet noise; this has provided clarification in terms of the modelling framework best adapted for the jet-noise problem. The goal of DARETOMODEL is to leverage these recent developments to build the robust, low-cost, modelling tool that industry requires. The close coupling of data-mining and theoretical modelling that has underpinned the most significant recent progress provides the foundation on which the project is built.
The main objectives of the project are to:
1. Use existing numerical and experimental databases to identify the minimum modelling requirements for robust jet-noise prediction in isolated and installed conditions, with and without flight effect.
2. Expand and complete existing databases with measurements dedicated to assessing the effect of flight.
3. Realise a fast, low-cost, semi-empirical modelling strategy for jet-noise prediction, tailored to require input data that is readily available from existing, industry-standard tools.
The main objectives of the project are to:
1. Use existing numerical and experimental databases to identify the minimum modelling requirements for robust jet-noise prediction in isolated and installed conditions, with and without flight effect.
2. Expand and complete existing databases with measurements dedicated to assessing the effect of flight.
3. Realise a fast, low-cost, semi-empirical modelling strategy for jet-noise prediction, tailored to require input data that is readily available from existing, industry-standard tools.
1. Development of a computational tool for modelling jet noise
2. Assessment of existing high-fidelity data: databases were exhaustively analysed to assess suitability for use as input to computational tool (1). Analysis identified hitherto unknown sources of error. Solutions were developed to enable error mitigation for existing data, and novel error-mitigation strategies for the generation of new data.
3. New theoretical guidelines: computational tool (1) was elaborated using state-of-the-art techniques. DARETOMODEL identified a hitherto unknown theoretical ambiguity implicit in these techniques. A correction was identified that both removes the ambiguity and shows how it may be beneficially exploited.
4. Combining (1), (2) & (3), best-practice guidelines were established for physics-based jet-noise modelling.
5. A novel data-analysis tool, RESPOD (Resolvent-based Extended Spectral Proper Orthogonal Decomposition), was developed for eduction of turbulence structure responsable for an observable of interest (e.g. drag, stability, noise,...).
6. Use of RESPOD to identify turbulence structure responsible for jet noise.
7. Construction, thanks to (6), of semi-empirical model for noise-producing turbulence structure.
8. Use of (7) to perform jet-noise prediction.
9. Evaluation of robustness of (7).
10. Development of second empirical modelling strategy that bypasses turbulence structure.
11. Transfer of tools (points 1, 7 & 10) and associated knowledge (points 2, 3, 4, 5, 6, 9) to Airbus.
12. Implementation of jet-noise modelling tool in industry environment, using industry tools.
13. Evaluation of robustness of the modelling tool in an industry environment.
14. Design of experiment to assess effect of flight on jets and their sound.
15. Velocity measurements of jets with flight stream.
16. Acoustic measurements of jets with flight stream.
17. Acoustic measurements of jet and wing, in static and flight conditions.
18. High-fidelity numerical simulation of the jets in flight.
19. Post-processing/analysis of new databases to assess effect of flight on jet.
20. Identification of effect of flight on flow structures.
21. Modelling the effect of flight on flow structures.
Results, exploitation and dissemination:
1. The main result of the project is a tool for predicting jet noise using input data from standard industry tools. The tool was developed and validated using data from experiments and high-fidelity numerical simulation. Signal-processing tools were elaborated for mining the data. Databases were used to guide/inform the jet-noise prediction tool, tailored and tested for robustness to changes in operating condition and errors in data. The prediction tool was transferred to an industry environment. Airbus engineers learned to use the tool; it was integrated into their suite of modelling tools; it was validated in the industry environment and shown to provide robust noise prediction with minimal input data from standard, fast-return, industry CFD.
2. Dissemination of results, in addition to the transfer of knowledge and tools to Airbus, took the form of publications in leading journals. 5 journal publications are in print; 4 are submitted or in preparation. Expected final output is 9 journal papers. Results were also presented at the 2022 AIAA/CEAS Aeroacoustics Conference; at the ICTAM conference in 2021; at the Euromech Colloquium, in Poitiers, 2021.
3. A Euromech Colloquium was organised in Poitiers 2021. DARETOMODEL results were showcased. A 33-page book of abstracts was produced, as was a 4-page scientific report.
4. DARETOMODEL databases will continue to guide future jet-noise research.
2. Assessment of existing high-fidelity data: databases were exhaustively analysed to assess suitability for use as input to computational tool (1). Analysis identified hitherto unknown sources of error. Solutions were developed to enable error mitigation for existing data, and novel error-mitigation strategies for the generation of new data.
3. New theoretical guidelines: computational tool (1) was elaborated using state-of-the-art techniques. DARETOMODEL identified a hitherto unknown theoretical ambiguity implicit in these techniques. A correction was identified that both removes the ambiguity and shows how it may be beneficially exploited.
4. Combining (1), (2) & (3), best-practice guidelines were established for physics-based jet-noise modelling.
5. A novel data-analysis tool, RESPOD (Resolvent-based Extended Spectral Proper Orthogonal Decomposition), was developed for eduction of turbulence structure responsable for an observable of interest (e.g. drag, stability, noise,...).
6. Use of RESPOD to identify turbulence structure responsible for jet noise.
7. Construction, thanks to (6), of semi-empirical model for noise-producing turbulence structure.
8. Use of (7) to perform jet-noise prediction.
9. Evaluation of robustness of (7).
10. Development of second empirical modelling strategy that bypasses turbulence structure.
11. Transfer of tools (points 1, 7 & 10) and associated knowledge (points 2, 3, 4, 5, 6, 9) to Airbus.
12. Implementation of jet-noise modelling tool in industry environment, using industry tools.
13. Evaluation of robustness of the modelling tool in an industry environment.
14. Design of experiment to assess effect of flight on jets and their sound.
15. Velocity measurements of jets with flight stream.
16. Acoustic measurements of jets with flight stream.
17. Acoustic measurements of jet and wing, in static and flight conditions.
18. High-fidelity numerical simulation of the jets in flight.
19. Post-processing/analysis of new databases to assess effect of flight on jet.
20. Identification of effect of flight on flow structures.
21. Modelling the effect of flight on flow structures.
Results, exploitation and dissemination:
1. The main result of the project is a tool for predicting jet noise using input data from standard industry tools. The tool was developed and validated using data from experiments and high-fidelity numerical simulation. Signal-processing tools were elaborated for mining the data. Databases were used to guide/inform the jet-noise prediction tool, tailored and tested for robustness to changes in operating condition and errors in data. The prediction tool was transferred to an industry environment. Airbus engineers learned to use the tool; it was integrated into their suite of modelling tools; it was validated in the industry environment and shown to provide robust noise prediction with minimal input data from standard, fast-return, industry CFD.
2. Dissemination of results, in addition to the transfer of knowledge and tools to Airbus, took the form of publications in leading journals. 5 journal publications are in print; 4 are submitted or in preparation. Expected final output is 9 journal papers. Results were also presented at the 2022 AIAA/CEAS Aeroacoustics Conference; at the ICTAM conference in 2021; at the Euromech Colloquium, in Poitiers, 2021.
3. A Euromech Colloquium was organised in Poitiers 2021. DARETOMODEL results were showcased. A 33-page book of abstracts was produced, as was a 4-page scientific report.
4. DARETOMODEL databases will continue to guide future jet-noise research.
Progress beyond the state of the art
- Hitherto unknown limitations of high-fidelity simulation data for jet-noise modelling were identified and novel mitigation strategies elaborated.
- Identification of a theoretical ambiguity in state-of-the-art use of the mean-flow-based linear modelling of turbulent shear flows,
- Proposition of approaches to remove ambiguity,
- Proposition of approaches to exploit ambiguity to improve modelling.
- Development of a tool (RESPOD) for identifying turbulence interactions that drive jet noise (or other observable).
- Development of second jet-noise tool that bypasses turbulence structure.
- Measurement and modelling of non-modal coherent structures in jets with flight.
- Demonstration of the effect of flight on non-modal linear mechanisms.
- A modelling tool optimised to provide jet-noise prediction with simplified industry tools requiring minimal input data.
- Integration of tool in AIRBUS acoustics department.
- Project objective achieved: a jet-noise modelling tools adapted for use in industry environment. Tool requires significantly less empirical input than current state-of-the-art models and shows excellent robustness to changes in jet operating condition and to errors associated with industry implementation.
Potential impacts
- The availability of the fast-return jet-noise modelling tool will improve the capacity of AIRBUS to design lower-noise aircraft.
- The competitiveness of AIRBUS will be improved.
- Lower-noise aircraft means a societal benefit in terms of lower noise pollution.
- Hitherto unknown limitations of high-fidelity simulation data for jet-noise modelling were identified and novel mitigation strategies elaborated.
- Identification of a theoretical ambiguity in state-of-the-art use of the mean-flow-based linear modelling of turbulent shear flows,
- Proposition of approaches to remove ambiguity,
- Proposition of approaches to exploit ambiguity to improve modelling.
- Development of a tool (RESPOD) for identifying turbulence interactions that drive jet noise (or other observable).
- Development of second jet-noise tool that bypasses turbulence structure.
- Measurement and modelling of non-modal coherent structures in jets with flight.
- Demonstration of the effect of flight on non-modal linear mechanisms.
- A modelling tool optimised to provide jet-noise prediction with simplified industry tools requiring minimal input data.
- Integration of tool in AIRBUS acoustics department.
- Project objective achieved: a jet-noise modelling tools adapted for use in industry environment. Tool requires significantly less empirical input than current state-of-the-art models and shows excellent robustness to changes in jet operating condition and to errors associated with industry implementation.
Potential impacts
- The availability of the fast-return jet-noise modelling tool will improve the capacity of AIRBUS to design lower-noise aircraft.
- The competitiveness of AIRBUS will be improved.
- Lower-noise aircraft means a societal benefit in terms of lower noise pollution.