The development of a design methodology to reduce the probability of brake judder and drone due to thermo-elastic instabilities of the brake rotor

Until recently research efforts regarding brake judder has been overshadowed by other brake related issues such as brake noise for example squeal and groan. Although brake noise still remains a problem for vehicle manufacturers, past research effort has led to a greater understanding of its mechanism and solutions are now more readily available. This is not the case with brake "judder" as past research has been less focussed towards its occurrence or cause. As brake performance and reduced packaging demands increase then the thermal issues of braking become more of an issue. It is this aspect of thermal judder, rather than cold judder, which is most difficult to investigate as this can manifest itself as high frequency (often referred to as drone) or low frequency excitation, the latter causing excessive vibration in the suspension and steering systems.
The focus of this research is the analysis of thermo-elastic instability after both high energy braking and prolonged drag braking. Both types of braking result in high energy input into the brake rotor. The principal issue is that the rise in temperature, under some conditions, is localised with temperatures exceeding 730ºC. If this is followed by rapid cooling, then a phase change in the brake rotor material results. This phase change in the metallurgical structure of the rotor may be seen on the surface of the rotor as dark areas which are referred to as "hot spots" or `blue spots´. The resulting increase in grain growth associated with these localised areas causes a further increase in temperature and so the process continues. The disc surface becomes a series of raised areas giving localised disc thickness variation. It is during a subsequent light braking application that judder occurs due to the DTV with a high order of vibration - or "drone".
The paper is aimed at detailing the preliminary vehicle testing which has been performed which will allow for accurate validation of dynamometer testing and computational analysis. Promising results have been obtained and are presented and discussed. Conclusions have been drawn with recommendations for further work.