CE40 - Mathématiques

Efficient simulation of noise of rotating machines – NORMA

Efficient simulation of noise of rotating machines

Prediction of the sound field of rotating machines with hybrid turbulence models and accurate and parallel implicit algorithms. <br />An increasing degradation of our environment by the noise generated by rotating machines (helicopters, drones, wind turbines, etc.) is observed. There is a need for small and large companies to be able to accurately perform aerodynamic and aeroacoustic simulations of single-rotor and multi-rotor rotating machines.

Development and implementation of accurate and efficient numerical methods and turbulence models for the simulation of the noise of rotating machines.

The objectives of this project are the development and implementation of innovative numerical methods and turbulence models for the simulation of the noise of single-rotor and multi-rotor machines. An important work is to develop low dissipative and low dispersive discretization methods allowing accurate propagation of acoustic waves on unstructured meshes. Another significant task is to propose scale-resolving simulation techniques for turbulent flow predictions which are adapted to aeroacoustics. A third important point is to implement a simple and efficient method for taking mobile geometries into account. The challenge is to combine these different approaches and adapt them to the aeroacoustic simulation of rotating machines.

One of the important tasks of this project is to develop low dissipative hybrid turbulence models that are suitable for aeroacoustics. Several hybridization strategies based on the RANS, DDES and dynamic variational multiscale of LES approaches are developed and evaluated. Another significant work is to develop low dispersive and low dissipative high-order vertex centered finite volume schemes, with an appropriate reconstruction of the solution at the interfaces of neighboring cells for each edge, allowing precise propagation of acoustic waves on unstructured meshes. Another task concerns the extension of the immersed boundary method, based on the Brinkman penalization method, to the simulation of blades using hybrid turbulence models and dynamic mesh adaptation strategies. Concerning time-advancing algorithms, a finite volume multirate approach is extended with the objective of accurately predicting aeroacoustic problems. A parallel LU-SGS method for implicit time integration is also investigated.

The ecology of areas inhabited by man is constantly deteriorating due in particular to noise generated by rotating machines (RM): mainly helicopters and wind turbines. This will amplify significantly with the irruption of a growing army of new single-rotor machines and multi-rotor systems (small multi-purpose helicopters, aerial taxis, drones, quadcopters, etc.) in the smart cities of near future. Noise mitigation can be obtained by the optimization of the shapes of RM's blades and fuselages and of other design characteristics, optimization which can be based on serial numerical predictions of aerodynamic properties and acoustic radiation generated by these new RMs. Providing the needed accuracy at acceptable computational cost for such complicated geometrical configurations is a difficult numerical challenge which the partners propose to address.
The Consortium involves specialists having proposed new methods in turbulence modelling, novel high-accuracy numerical schemes for unstructured meshes and efficient parallel algorithms applicable for predicting complex flow near RM's blades and fuselages via scale-resolving turbulent flow simulation, and the associated far field acoustic fields.
The strength of the proposal relies on the further development of a few innovative methods which the partners have introduced, a notable part of these methods being derived in common due to a long cooperation history.
Complex and moving RM's geometries needing dynamic mesh adaptation will be addressed thanks to higher-order edge-based reconstruction schemes for unstructured meshes (jointly developed by partners from 90'), particularly adapted to aeroacoustics. The novel ENO and WENO edge-based reconstructions combined with the recently proposed localized adaptive artificial viscosity will improve the treatment of discontinuities on unstructured meshes. Multiple moving geometry ability will be further amplified thanks to a specific level-set based Brinkman penalization method to mimic solids, new moving mesh method for dynamic mesh adaptation to solid surfaces and mesh adaption with a continuous metric. The very high-order time advancing will be efficiently applied thanks to a novel multirate formulation. The all-Mach methods will be implemented for the vertex-centred edge-based formulation used to improve the simulation of flow around RM's blades. The new hybrid RANS/LES models including transition models and LES filters of extremely low dissipation will permit the accurate generation and propagation of noise. The recent enhanced DES with new subgrid scale accelerating the "numerical" transition in turbulent shear layers will also be used for cross validation.
Two families of geometries for single and multiple rotor systems (helicopter main rotor with fuselage, quadrotor system with fuselage respectively) will be considered for evaluating the achieved improvement in efficiency and accuracy of the developed massively parallel simulation tools.

Project coordinator

Monsieur Bruno Koobus (Institut Montpelliérain Alexander Grothendieck)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.


IMAG Institut Montpelliérain Alexander Grothendieck
INRIA Centre de Recherche Inria Sophia Antipolis - Méditerranée
KIAM Keldysh Institute of Applied Mathematics of RAS / Computational Aeroacoustics

Help of the ANR 280,354 euros
Beginning and duration of the scientific project: February 2020 - 48 Months

Useful links

Explorez notre base de projets financés



ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter