The project addressed the problem of squealing disc brakes of railway rolling stock, in particular the TGV and AGC (Autorail Grande Capacité) emitting high levels of noise when stopping at stations.
Train operators and train manufacturers wished to reduce significantly or eliminate this noise on their rolling stock. To this end, they sought to specify the brake linings (consumables) with respect to the noise whereas suppliers aimed to design quieter pads.
The project aimed at developing tools to support noise trim specification and design. The ambition is to improve the essentially empirical approach usually adopted and based on the development of physical models capable to reproduce the main phenomena causing the screeching of braking systems.
From a scientific perspective, this implies the development of numerical methods for calculating vibrational instabilities related to friction, adapted to industrial numerical models, and the application and experimental validation of these methods to TGV and AGC systems.
The project has been organized around five main tasks:
- R&D and efficient numerical methods
- Characterization and mechanical modelling of the brake components
- Characterization and modelling of brake squeal
- Development and validation of industrial tools finalized
- Design and testing of working models by suppliers
- Analysis of a large and diverse panel linings must allow better assess the validity and robustness of the tools developed.
One of the key results achieved was the development of an efficient numerical method for analysing vibratory instabilities of structures in contact rubbing, especially the non-linear dynamic transient and steady self-sustained vibrations causing the screeching (Task 1).
The second advance, made possible by the use of this method lies in modelling the TGV and AGC’s squeal braking systems and the confrontation with the experimental results obtained on the SNCF brake test bench (Task 3).
These results allowed to develop a simplified software tool to be used by manufacturers.