In order to solve brake related noise problems, a clear insight in the fundamentals of self-exiting oscillations is of utmost importance. When a brake path is contacted to a rotating disc or drum, a non-steady friction force is generated. Under certain, usually poorly controlled, boundary conditions this reaction force will interact with the dynamics of the brake system. Typically this friction force will excite a resonance of the brake disc or caliper, amplifying the friction force, and so on.
The amplitude of this force self-sustained vibration is determined by energy considerations: during each period of vibration the amount of dissipated energy equals the amount of externally supplied energy. The dissipation of energy is determined by the damping of the system. The supply of energy is function of the disc/brake pad friction force and the system’s reponse to this force. Several theoretical models exist that describe in very detail this energy balance. None of these model is perfect, but it is straightforward in all models that an increase of damping is required in order to control or reduce the amplitude of vibration.
However, these models also prove that by a simple addition of damping, self-exciting oscillations cannot be avoided and a more fundamental change in the dynamics of the brake systems is required. Therefore a detailed description of1 the behaviour of the brake system under excitation (operational deflection shape and/or operational modal analysis) is necessary for solving brake related noise. The frequency of the brake noise determines the scope of this investigation. For low frequency problems not only the brake system should be considered, but also the entire wheel suspension. While for higher frequencies it can be sufficient to describe only the behaviour of the disc, caliper and brake pads.
In this case a sports car is equipped with ventilated disc brakes at the rear. The rear wheel suspension consists of a twist beam construction that links the left and right wheel. A very disturbing brake squeal is present around 380 Hz.
Unfortunately for the customer, this noise is easily generated and appears under a wide range of boundary conditions. This omnipresence of the noise facilitates acquiring the operational deflection shapes. The vibration of the disc has been measured by a laser vibrometer. The static parts’ deflection has been acquired by miniature high temperature accelerometers. The deformation consisted of a rigid body movement of the entire brake system without any relative motion between the brake pads and the disc. The disc itself was also wobbling about as a rigid body. The wheel spindle was relatively motionless while the highest vibration levels were encountered at the twist beam. This beam has a V-shape cross section and both arms of the V where moving in counterphase.
Modal analysis indicated that there was a coincidence in resonance frequency of the wobbling mode of the brake system and the local deformation of the twist beam. Although the source of the noise is situated at the contact between disc and brake pads, the noise radiator proved to be the twist beam of the wheel suspension. As a consequence, there is a crucial energy flow from the brake system towards the wheel suspension. Countermeasures can then be situated at different levels : the excitation at the source level (friction force), the energy flow from the source to the noise emitter and the noise emittor itself.
This brake squeal has been tackled by several successful countermeasures at the prototype level (tuned absorbers, asymmetric disc, stiffened twist beam, surface treatment disc,…). However, the industrial solution consisted of a reinforced brake support. Stiffening ribs were added at the cast iron interface between the brake and the wheel spindle. The additional costs for this modification were almost neglectable while a 100% successful solution was obtained.
