The noise friction coefficient is a key to the development of a NVH optimised brake system. Nowadays friction measurement on noise test rigs is based on a torque transducer integrated into the drive shaft. This torque transducer faces several kinds of disturbances by dynamic effects inside the drive shaft e.g. torsional vibrations. The torque transducer based calculated noise friction level variation therefore is known to be significantly noisier as on a classical performance dyno with Torque Measurement via Tailstock. First choice for the analysis of the error margin would be a direct comparison of a noise dyno drive shaft, with and without torque transducer. This methodology is limited by torque measurement problems in a noise test rig brake assembly and control difficulties of the test rig. Therefore a model based approach was chosen. Based on the modal analysis of the drive shaft a simulation model was developed. This simulation model can be used to estimate the error margin by parameter variation. This leads to an error margin being about two up to three times higher as on a performance dyno. The simulation model is valid just for the test rig under evaluation, but generally spoken, it shows the limitations of torque transducer based drive shafts, which are used in every noise dyno design, independent of the manufacturer. To overcome this problem a model based new design of a drive shaft was developed and implemented in a noise test rig. The new design shows much better noise friction coefficient performance. The spread of Min- / Max- Friction is reduced by approx. 30 - 50%. The error margin regarding the friction coefficient of state of the art noise dynos was analysed using a model based approach. A model based suggestion for a new torque measurement system was developed and successfully tested.

Dohle, Achim; Zheng, Wenxin; - TMD Friction, Weiss, Dieter; - Horiba; Perzborn, Nils; Ostermeyer, Georg-Peter; - TU Braunschweig