• Research and/or engineering questions/objective To reduce the non-exhaust particle emission of railway applications the main influencing factors have to be identified. Besides the brake-stop load cases in different applications the physical, chemical and performance properties of a wide range of friction materials and counterparts (wheel and brake disc materials) has to be considered. Therefore, a benchmark test scenario was created to compare different combinations of friction couples. The result of the survey was used to identify main correlations that can be used for simulation of particle emissions. • Methodology On a railway subscale dyno a test scenario with nine different load cases was used to measure friction performance, temperatures, wear and particle emissions of a wide range of friction materials and counterparts. The PM value obtained by an ELPI+ device was analyzed to find a suitable particle mass distribution equation. Based on a wear model the simulation of PM values of user defined full scale dynamometer tests, e.g. route profiles, was executed and compared with test results. • Results Based on subscale dynamometer test results some general relations in regard to particle emissions were found. By application of a suitable particle emission distribution function together with a tool to predict the friction material wear a simulation of PM values is feasible. • Limitations of this study For brake pad materials acting on brake discs the influence of the geometry of the brake pad could be considered. For the different brake block configurations available in service more investigations at full scale dynamometer are necessary to optimize the scaling procedure. • What does the paper offer that is new in the field in comparison to other works of the author In this investigation a benchmark study on many different railway friction materials is shown. Based on this some general relations between load cases, wear and particle emissions were found. • Conclusion The simulation of particle emissions in service is possible based on wear prediction and a distribution function for PM values. Some influences of load case conditions on the distribution of PM values were clarified and are implemented in the used model to simulate the PM values.

Dr. Dirk Meyer, Director Friction Technology, Wabtec Faiveley Becorit; Dr. Martin Rustemeier, R&D chemist, Wabtec Faiveley Becorit; Dipl.-Ing. Daniel Rabiega, Teamleader Testcenter, Wabtec Faiveley Becorit