Project Summary

Noise reduction is a crucial objective for the European aviation industry in the near future. Jet noise emitted by aeroengines stands as a primary target for reduction, achievable through active flow control — manipulating the flow structures responsible for noise emission. To effectively address this control problem, accurate methods are needed to discern how both the flow field and noise source are altered by control actions. Such insights are often elusive in experimental studies due to hardware constraints. The INFLUENTIA project aims to provide AI-based tools, integrating data from flow field measurements of jet flows and far-field sound measurements from microphone arrays. Establishing a clear relationship between these two quantities is not easy, but will facilitate turbulent-jet-noise estimation and control. To accomplish this objective bridging three distinct disciplines —fluid mechanics, acoustics, and statistics—, the project assembles a multidisciplinary research team proficient in cutting-edge techniques of experimental aerodynamics, acoustics, and data science. The team, with a solid background in PhD tutoring, will train a student in this intersectional research area. The project adopts a multi-faceted approach to varying complexities of the problem, initially determining and connecting flow field and sound recordings states, retrieved using quantization algorithms, then progressing towards incorporating the full dimensionality of the problem through the use of convolutional/recurrent neural networks, Fréchet random forests, and diffusion models adapted to complex data types. This approach will ensure meeting the final objective: establishing a connection between flow and sound. The outcomes of INFLUENTIA are poised to significantly contribute to aircraft noise reduction, offering valuable insights to the European aviation industry.

INFLUENTIA in a nutshell

Title: Learning flow and Noise Dynamics in turbulent jets via artificial Intelligence

Goal: establish the complex relation between flow fields and far field noise of turbulent jets via AI methods

Duration: 30 months (01/07/2024 – 31/12/2026)

Budget: 75.000 €

Funding: This work has been supported by the Madrid Government (Comunidad de Madrid-Spain) under the Multiannual Agreement with UC3M (INFLUENTIA-CM-UC3M).

Team

Results

Navarro-González, M., Gonzalez-Sanchez, B., García-Portugués, E., & Raiola, M. (2025). A Dual-Time Implementation of the Spectral Proper Orthogonal Decomposition. In AIAA AVIATION FORUM AND ASCEND 2025 (p. 3221). Las Vegas, USA ,21 – 25 July 2025