Search Results

Although the radial migration of hot bands has been frequently observed, a systematic investigation of this phenomenon has not yet been performed. The ring-shaped temperature maximum, which occurs on the brake disk, is undesirable because the focused temperatures destroy the local materials in contact. Moreover, a hot band carries a dominant portion of the frictional load. If a hot band moves radially, the braking torque is directly influenced. It is supposed that material wear influences the radial hot band migration. New models demonstrate that wear is indeed the mechanism that triggers hot band migration. First, a minimal model including thermal expansion and a load-dependent loss of material is introduced. The simplicity of the model allows for an understanding of the impact of wear, as well as the mechanisms that lead to a periodic load distribution. This model can be analyzed in terms of complex eigenvalues, showing a periodic load distribution in the sliding plane.

For brake performance measurements, many standardized test procedures and machines were developed. A characteristic problem of these highly complex machines is the occurrence of measurement uncertainties, which usually are not predictable and rather difficult to explain. One part of explanations is driven by the huge but not yet complete list of influencing factors for the high load friction processes. Another part of explanations is given by f.i. manufacturing tolerances, load histories of machines and specimens and so on. A systematic investigation of these influences is very time-consuming and could be difficult to realize on professional test-machines [1]. On the other hand, simple laboratory tribometers are rather good for sensitivity investigations in principle but the friction systems are too far from the real system.

The Automated Universal Tribotester (AUT) is developed by the Institute of Dynamics and Vibrations (TU Braunschweig) and represents a reduced scale brake dynamometer. The setup is based on the pin-on-disc principle and the down-scaled test specimen are brought to contact to the disc and loaded via the specifically designed load unit. The AUT’s load unit is designed as a combination of parallel and serial leaf springs, resulting in a friction free motion. The stiffnesses in radial and tangential directions are much higher than in normal orientation. For the investigation of wear debris over time, changes in loads (e.g. forces, speeds, temperatures) are applied. Those varying loads result in tilting of the contact surface of the test specimen due to small elastic deformations. A change of the contact area is inevitable, and long time periods are needed to adopt the contact area to the new conditions. This prevents from investigating fast changes in the above mentioned loads.

Inertia dynamometers are commonly used to determine the friction coefficient of brake assemblies. Dynamometers are a well-established platform, allow testing under controlled conditions, exhibit a good correlation to many situations encountered in real driving, and are comparatively economical and less time-consuming than full vehicle test. On the other side of the spectrum is the use of scaled tribometer. These test systems make possible a test without the entire brake corner. This separation allows the investigation of the frictional-contact only (frictional boundary layer) speedily and independently of a given brake system or vehicle configuration. As the two test systems (inertia dynamometers and tribometers) may have different users with possibly different tasks, the question remains regarding how comparable the two systems are. These issues provide incentives to better define the fields of investigations, correlation, and applicability for the two systems.

Friction materials for automotive brakes are known to exhibit a time-dependent tribological behavior. When examining these dynamic effects special demands are made on the measurement device: The influences of the brake system should be minimized and parameters like velocity, contact pressure and temperature should be controlled closely and independently. Furthermore, special test procedures need to be designed. This can ideally be achieved using a scaled tribometer like the High-Load-Tribometer at the Institute of Dynamics and Vibrations in Braunschweig. Former investigations [1] have shown that a kind of memory effect can occur for a low-met brake pad rubbing on a cast iron disk. A variation of the initial disk temperatures has revealed that a temporary increase of the coefficient of friction can occur at slightly elevated temperatures. This effect is memorized by the material as a certain procedure needs to be performed in order to achieve a regeneration.

In today’s research and development of brake systems the model-based prediction of complex vibrations and NVH phenomena plays an important role. Despite the efforts, the high dimensional computational simulation models only provide a limited part of the results gained through experimental measurements. Several reasons are discussed by the industry and academic research. One potential source of these inadequacies is the very simple formulation of the friction forces in the simulation models. Due to a significant shorter computation time (by orders of magnitude), the complex eigenvalue analysis has been established, in comparison to the transient analysis, as the standard method in the case of industrial research, where systems with more than one million degrees of freedom are simulated.

The dynamics and, in particular, the NVH phenomena in brakes are still in the focus of research. Recent investigations of for example Rhee et al. show two principal vibrational forms of the linings on the rotor [1]. The first form is characterized by vibrations where both linings are in-phase (minimal differential torque between the inner pad and the outer pad). This produces in-plane vibrations of the rotor and results in high-frequency squealing events in the brake. The second form is an antiphase vibration of the brake linings with respect to each other (increased differential torque between the inner pad and the outer pad). This produce directly out-of-plane vibrational modes of the disc, which results in lower-frequency caliper and rotor oscillations. One hypothesis is that different wear densities of the linings essentially characterize the two vibrational modes. The wear behavior is not taken into consideration of this paper as it will be discussed in further publications.

There are few principal excitation mechanisms that brake system NVH simulations are based on, especially the high frequency squeal simulations. These mechanisms can be described by some simple mechanical models that exhibit excitation or self-excitation effects induced by friction [1, 2]. These models use very simple friction laws of Coulomb type, described by a friction coefficient that is either a constant or simple functions of some state variables, taking into account a Stribeck characteristic. Measurements from the AK-Master or SAE J2521, however, show that the friction coefficient is not a simple function of some state variables, describing a steady state behavior of friction. In the past several years, material dependent descriptions of the frictional brake interface have started attracting attention [3]. These aspects are greatly influenced by the tribological effects at the frictional interface, which can be characterized by typical wear patterns.

For more than a decade, investigations of the surface topography and chemical behavior of the boundary layer have been performed. This is of great importance for the understanding of friction, as it provides valuable insights into the dynamic processes that influence the brake pad's performance. The adaption of the boundary layer to a new load condition is of particular interest because it allows for the identification and description of mechanisms that influence macroscopically-observable effects, and quantities such as the coefficient of friction. It is possible to determine time constants that describe this process of adaption. The investigations discussed in this paper combine tribometer tests and optical investigations at the same time. Therefore, the measured coefficient of friction can be related to the observed surface topography of the brake pad.

Dynamic aspects in the understanding of friction are increasingly coming into focus. Therefore, test arrangements are required, which allow dynamic measurements as well, especially with defined and reproducible sliding speeds. In addition, investigations of vehicle brakes require high power. A simple, reliable measurement device, based on the pin-on-disc principle, that meets these requirements was developed and will be presented here. This high load tribometer is based on a specially altered lathe. One key benefit of a lathe as a basic unit is the fact that the main mechanical part is standardized industrial equipment. Additional components and measuring devices can be integrated with little effort and great flexibility. The built-in 15kW AC motor and the embedded gearbox permit speeds and torques to the service areas of vehicle brakes, also of heavy duty vehicles and sports cars.

In recent years, characteristic structures in the boundary layer of high-load contacts such as brakes have been reported, which have an important impact on the dynamics of the tribological contact. Usually, local assumptions concerning the friction of these patches are used to reach global conclusions about the brake system. Several numerical methods (e.g. Cellular Automata) have been developed which make use of such assumptions. The validation of these methods through measured data tends to be laborious and costly. Sprag-Slip elements are friction elements which are typically considered to exclusively undergo static friction. Such elements have been sporadically utilized towards describing friction in brake applications. In this paper, many locally distributed Sprag-Slip elements are used to model the global dynamics of braking friction. The results show good agreement with the measured characteristics of brakes.