Friction-induced Noise And Vibration Of Disc Brakes

Wear, 133 (1989)

39

39 - 45

FRICTION-INDUCED

NOISE AND VIBRATION

OF DISC BRAKES*

S. K. THEE, P. H. S. TSANG and Y. S. WANG ~~l~~~~-Sj~na~A~~#~o~~ue

Technical

Center,

900 West Maple Road, MI 48084

(U.S.A. j

Summary

Several noise excitation theories have been proposed in the literature. These theories are found to be unsatisfactory for explaining the noise excitation phenomenon. In this paper, we propose a simple mechanical impact (hammering) model for independent of friction variation speed.

brake noise generation. This model is during the period of decreasing sliding

1. Introduction -4 brake system must meet certain customer requirements for performance, durability and noise. In recent years, disc brake noise has become an issue of growing concern to the automotive industry, especially to the manufacturers of disc brake systems and friction materials. Brake noise is a very complex phenomenon owing to the design of the disc brake system and its operating conditions. Although numerous studies of this phenomenon have been carried out throughout the years, understanding of its excitation mechanisms remains rather limited. Several mechanisms for noise excitation have been proposed in the literature. North [l - 31, and Lang and Smales 141 gave excellent surveys of these mechanisms. In general, the various proposed excitation mechanisms can be roughly grouped into two major schools of thought. First, it is commonly believed that the brake squeal is caused by a rapid increase in the coefficient of friction with decreasing speed in braking, i.e. the p us. speed curve. Fosberry and Holubecki [ 5,6], among others> conducted extensive experimental investigations in this area. Secondly, in the case where there is no apparent change in friction, brake squeal is believed to be caused by a system instability related to the interaction of the structural components of the brake system. Earles and Soar [7], Milner [8] and others contributed significantly to the advancement of the instability theory. Most recently Murakami et al, [9] proposed “Paper presented at the International U.S.A., April 8 - 14, 1989.

Conference

on Wear of Materials, Denver, CO,

004%1648/89/$3.50

@ Elsevier Sequoia/Printed

in The Netherlands

40

that both schools of thought should be considered and reported that the excitation of brake squeal was influenced by both the P us. speed and the structural instability factors. The p us. speed, the instability, and the combined P us. speed/instability theories accurately describe the conditions under which brake noise might occur, but they do not clearly define the physical phenomenon which causes brake noise. In this paper, a simple “hammering model” for noise and vibration excitation is presented and discussed.

2. Discussion of noise phenomenon A disc brake consists of a caliper, two disc pads (friction material), a rotor and other components for attaching the caliper and the rotor to the vehicle. Usually, the brake is operated hydraulically. The caliper, which contains a cylinder with a piston, holds the two disc pads on either side of the rotor. The caliper may slide while the rotor is firmly attached to the wheel. Common friction materials are made of complex resin-based, short fiber-reinforced composites containing various friction modifiers and fillers. Typical compositions of friction materials are found in a paper by Jacko et al. [lo]. For stable friction and low wear, a good friction film (or glaze) is desirable on the friction surface of the pads and rotor, as shown by the investigations of Liu and coworkers [ll, 121. A disc pad assembly consists of a friction material attached to a steel backing plate of a certain thickness (about 4 mm) and configuration. Attachment can be achieved by either riveting or chemical bonding. A rotor is generally made of grey cast iron. It can be of either a simple solid rotor design or of a configuration with various vents for more effective cooling. Production brake systems vary greatly with respect to the configuration of caliper, disc pads and rotor. The use of noise insulators on the backing plate is also recently gaining popularity owing to the increasing concern over brake noise [ 13, 141. In braking, through the actuation mechanism of the caliper, the two disc pads are brought into contact with the rotor in motion. The resultant friction between the rubbing surfaces of the rotor and pads decreases the rotor and vehicle speed. In this braking process, the kinetic energy of the vehicle is transformed into frictional heat, part of which is transferred to the atmosphere by convection and radiation, and the rest of which is dissipated through the rotor and the friction material by conduction. The temperature of the rubbing surfaces of the rotor and pads will rise during braking. This temperature rise will significantly affect the chemistry and microstructure, as well as the physical properties of rubbing surfaces. In cases where this temperature rise is sufficiently high, the desirable friction films on the friction surfaces may be destroyed, and undesirable friction variation and high wear rates may result. Furthermore, a brake is usually operated at various conditions of speed, deceleration, temperature and load;

41

the tion and the

mode of applying the brake may also vary from a slow, steady decelerato hard, quick applications; or it may even be a combination of fast slow deceleration. Thus brake noise may be dependent not only upon system design but also upon the way that the brake is applied.

2.1. Types of brake noise In general, there are several types of noise or vibration associated with a disc brake system. They are classified into two major categories: (a) low frequency rigid body vibration (about 100 - 1000 Hz), called brake roughness, judder, moan or groan; (b) medium and high frequency vibration (about 1000 - 18 000 Hz), called squeal or squeak. Each type of brake noise has unique characteristics and probably unique excitation mechanisms. The brake roughness or judder is a low frequency oscillation of the order of 100 Hz which can be detected by the driver’s foot or hands. The roughness or judder could be caused by (a) dimensional variations of brake components e.g. variation in thickness of the rotor, lack of parallelism in rotor or pad surfaces, distortion of lug bolts or wheel and (b) variation in thickness of the rotor as a result of either thermal effects, or massive friction material transfer to the rotor surface. The other low frequency phenomenon, moan, occurs at around 150 to 400 Hz and may be caused by rigid body oscillation of the caliper and its mount. Groan or creep groan is another kind of low frequency (100 - 400 Hz) audible vibration which generally occurs at around 12 mile h-’ (20 km h-l). Brake squeals or squeaks generally occur at or above 1000 Hz up to the limit of human hearing. They are usually associated with the continuous diametrical vibration of the rotor and may also be due to the bending and twisting modes of vibration of the pad/ shoe assemblies. Usually, brake squeals are observed towards the end of a stop. How are these vibrations at various noise frequencies excited? As mentioned previously, the most commonly accepted theory is that a rapid increase in friction with decreasing sliding speed causes brake noise. If this is so, a brake system being dragged at a slow constant speed would be less likely to generate noise than one being stopped, since the latter has the potential for a rapid increase in friction with decreasing speed. However, in practice, the reverse is found to be true. In other words, a brake being dragged generates more noise than one being stopped or snubbed (from a high speed to a lower speed without coming to a complete stop). Also, during a stop, noise can occur quite suddently within a brief period of time in the order of about 50 ms during which the sliding speed remains virtually constant or does not change appreciably. Thus it should be questioned whether or not this theory is correct. In fact, in comparing noisy stops with quiet stops in our inertial dynamometer study, no correlation is found between friction change and noise excitation, when the friction level at the time of noise occurrence is examined closely. Therefore a noisy stop does not always have a rapid increase in friction with decreasing sliding speed towards the end of the stop where noise tends to occur; nor does stable friction throughout a stop guarantee a quiet stop.

42

According to modal testing of disc brake rotors, some 70 to 80 natural frequencies and vibrational modes are observed in the range 1000 - 16 000 Hz. Similarly, 30 to 40 frequencies and vibrational modes are observed for a typical disc pad assembly (friction material and steel backing plate), and about the same number for a caliper. When a rotor, a disc pad assembly, or a caliper vibrates by bending and twisting at these high rates, it must be asked what might be occurring at the sliding interface and what is happening to its normal force and friction. It is very likely that the area of contact between the rotor and the friction pads is small while noise or vibration is occurring, and the contact location changes rapidly. In a case like this, what would be the level of instantaneous normal load and friction? It must be asked whether conventional experimental measurement techniques are adequate for detecting these rapid changes in normal load and friction. It appears that conventional p vs. speed curves do not accurately represent measurements of the real physical phenomena occurring at these high rates. Thus it is suggested that new experimental techniques be considered for observing and measuring these rapid changes in normal load and friction while noise is being excited and that a new model of noise excitation mechanism is necessary for explaining the phenomenon. It might also be suggested that the destruction of friction film (glaze) on the sliding surfaces yields rougher rubbing surfaces, and causes a subsequent change in friction and wear behavior, as well as brake noise. If this is true, brake noise would occur more frequently under more severe braking conditions at high temperatures, where friction film destruction takes place more readily. Contrary to this notion, brake noise is more frequently observed under mild braking conditions of slow speed, slow deceleration, light pressures and relatively low temperatures in the neighborhood of 150 . 250 “c. 2.2. The hammering model for noise activation As stated above, modal analysis of brake components with either a hammer or a shaker shows that the rotor has 70 - 80 natural frequencies and vibrational modes, the friction pad or shoe assembly has 30 - 40, and the caliper also has 30 - 40 of its own. It is important to note that the noise frequencies identified in a vehicle test of the same brake also match those from modal analysis of the brake components. Based on this observation, the authors hypothesize that the brake noise and vibration might be activated by a “hammering” type of mechanism, not dissimilar to the modal analysis in which the vibration of the brake components is actually induced by a hammer or a shaker. The “hammering” during braking might occur between the friction pad/shoe assemblies and the rotor, or between the pad/ shoe assemblies and the caliper. It must then be asked how the hammering is initiated. In other words, which component gets “hammered” into natural modes of vibration and subsequently begins to hammer the other component( s)?

43

Abendroth’s experimental study on brake noise with a derotator system on a brake dynamometer indicates that after a certain number of braking stops or snubs, the rotor surface can develop 9 or 13 equally spaced high_ spots, which appear as waviness or hills and valleys on the surface [15]. The recent study by Inoue [16] also shows that thermal deformation of the rotor can occur during braking as a result of the frictional heat which causes the formation of localized hot spots on the rotor surface. It is easy to imagine that if the friction pad has a length approximately equal to two neighboring hills on the rotor surface, the pad may be tilted or cocked when slid across these hills and valleys, in a fashion rather similar to a small boat rocking up and down against a series of waves of certain wavelength. Therefore we can also imagine that “hammering” during braking may be initiated by the rocking action of the disc pads when they slide across the rotor surface against the “hills and valleys” which are formed by thermal distortion (or by any other mechanical distortion such as uneven rotor wear or massive friction material transfer to the rotor). The impacted pads might in turn hammer the caliper or the rotor, and thus start a chain reaction of hammering among the brake components. During this process one or more components might be excited into natural modes of vibration or resonance which result in noise and vibration, The order as to which component would go into resonance first probably depends on the specific system design and its operational conditions. Thus, this simple mechanical impact (hammering) model can explain the excitation mechanism of brake noise without having to deal directly with the frictional force, or the destruction of the friction film. The model is also useful in explaining the observation that brake squeal is more likely to occur at lower speeds. At relatively low speeds, the disc pads have to slide over the hills and valleys of the rotor surface in a rocking or cocking motion, which leads to eventual hammering and noise as described above. At higher speeds, however, we believe that it becomes possible for the pads to skim over the hills and valleys without resulting in severe rocking motions, just as a speed boat skims over the water surface. In other words, the hammering action could be avoided at higher braking speeds, and the brake would have a lower tendency to induce noise under such conditions. The model is not inconsistent with the instability theory. The action of hammering can be viewed as an effect of an “unstable state” induced by unfavorable interaction among structural components. While the instability theory depicts the conditions of component interaction which can lead to the excitation of noise or vibration, the hammering model describes the physical phenomenon by which the brake noise is excited. It is generally believed that the phenomenon of increasing friction, with decreasing sliding speed is due to the difference between a high static friction and a lower dynamic friction, or stick-slip. According to Schallamach’s work on the stick-slip phenomenon [17], the more elastic member of the two rubbing surfaces buckles at the interface during the stick-slip motion, and as it

44

recovers from this buckling, a wave of detachment sweeps across the contact region, with a movement similar to that of a caterpillar. We believe that this sweeping wave of detachment produces the same effect as a series of tiny hammers, which can provide additional energy for exciting the components into noise or vibration. Therefore the stick-slip motion can be viewed simply as a secondary hammering mechanism. It was mentioned earlier that the use of noise insulators was gaining popul~ity for reducing disc brake squeal. A typical noise ~sulator consists of an elastomer of high damping capacity, sandwiched between two thin stiff steel plates with certain adhesives. This sandwich assembly is then attached to the back of the steel backing plate of the disc pad. It has been reported that the shearing action of the high damping elastomer between the plates during braking is responsible for reducing the brake squeal [ 13, 141. We believe that the efficiency of the noise insulator can also be equated to its effectiveness in minimizing the hammering action between the disc pad and the caliper structure. 3. Summary

and conclusions

Brake noise is a complex phenomenon. Several theories on brake noise excitation have been proposed in the literature. However, these theories are found to be inadequate in explaining the noise excitation phenomenon. We propose a simple mechanical impact model (hammering) for brake noise excitation. This physical model does not rely on friction variation to explain the excitation of noise and ~bration. The model is not inconsis~nt with the “instability” theory, although it differs from the latter in that it depicts the physical process by which the noise and vibration are activated. Moreover, stick-slip can also be treated as a secondary mechanism of hammering by this simple mechanical impact model. We also believe that more advanced and accurate measurement techniques other than the conventional experiment techniques for measuring the normal load and frictional force are needed to aid study of and increase understanding of the complex but important noise phenomenon.

References 1 M. R. North, Disc brake squeal-a theoretical model, MIRA Research Rep., 197215. 1972. 2 M. R. North, Disc brake squeal, Proc. Conf. on Brake of Road Vehicles. 1976, Institution of Mechanical Engineers, 1976, pp. 169 - 176. 3 M. R. North, A survey of published work on vibration in braking systems, MIRA Bull., 4, 1969. 4 A. M. Lang and H. Smales, An Approach to the Solution of Disc Brake Vibration Problems, Paper C37/1983, Institution of Mechanical Engineers, 1983. pp. 223 231.

45 5 R. A. C. Fosberry and Z. Holubecki, An investigation of the cause and nature of brake squeal, MIRA Research Rep., 1955/2, 1955. 6 R. A. C. Fosberry and Z. Holubecki, Disc brake squeal; its mechanism and suppression, MIRA Research Rep., 1961/2, 1961. Ti S. W. E. Earles and G. B. Soar, Squeal noise in disc brakes. Symp. on Vibration and Noise in Motor Vehicles 1971, Institution of Mechanical Engineers, 1971, pp. 61 - 69. 8 N. Milner, An analysis of disc brake squeal, SAE Paper 780332, 1978. 9 H. Murakami, N. Tsunada and T. Kitamura, A study concerned with a mechanism of disc-brake squeal, SAE Paper 841233, 1984. 10 M. G. Jacko, P. H. S. Tsang and S. K. Rhee, Automotive friction materials evolution during the past decade, Wear, JO0 (1984) 503 515. 11 T. Liu and S. K. Rhee, High temperature wear of “semi-metallic” disk brake pads. In W. A. Glaeser, K. C. Ludema and S. K. Rhee (eds.), Proc. Int. Conf. on Wear of Materials, 1977, American Society of Mechanical Engineers, New York, 1977, pp. 552 - 554. 12 T. Liu, S. K. Rhee and K. L. Lawson, A Study of Wear Rates and Transfer Films of Friction Materials. In W. A. Glaeser, K. C. Ludema and S. K. Rhee (eds.), Proc. Jnt. Conf. on Wear of Materials, 1979, American Society of Mechanical Engineers, New York, 1979, pp. 595 - 600. 13 T. M. Lewis and P. Shah, Analysis and Control of Brake Noise, SAE Paper 872240, 1987. 14 C. T. Hoffman, Damper Design and Development for Use on Disc Brake Shoe and Lining Assemblies, SAE Paper 880254, 1988. 15 H. Abendroth, A New Approach to Brake Testing, SAE Paper 860080, 1985. 16 H. Inoue, Analysis of brake judder caused by thermal deformation of brake disc rotors, 21st FISITA Congr., Belgrade, June 2 - 6, 1986, 1986, pp. 2.213 - 220. 17 A. Schallamach, How does rubber slide?, Wear, 17 (1971) 301 - 312.