|
|
|
DISC BRAKE ROUGHNESS LINK BRAKE TECHNOLOGY REPORT—FEV2 Written by Arnold E. Anderson COMMON KNOWLEDGE ABOUT DISC BRAKE ROUGHNESS Definition and description Disc brake roughness is a rigid body vibration, mostly at wheel rotation frequency, that is caused by brake torque variations. It is detected more readily by the driver, since brake roughness often feeds back through the brake pedal in the form of a tactile pulsation and may produce a similar feedback through the steering wheel. Both driver and passenger may feel brake roughness through vehicle vibrations. Occasionally, front end sheet metal may vibrate enough that roughness also may be seen as well as felt. Roughness Mimics Many vibrations can occur at wheel rotation frequency that are not brake roughness. For example, tire and wheel unbalance can cause tactile and visual vibrations that may be sensed at the steering wheel. However, such vibrations do not require application of the brakes and tend to occur only at specific narrow speed ranges (typically at 52-56 mph and 70-75 mph, sometimes at 35-37 mph). Poor suspension alignment, bent wheels, and some road surface irregularities can produce vibrations that are similar to brake roughness. Sometimes these may be more pronounced when the brakes are applied. Therefore, it is important to be careful in diagnosing and rating brake roughness on a vehicle. Proper vehicle instrumentation can be used to identify, quantify, and document brake roughness test data. Roughness Occurrence Disc brake roughness has been around for a long time, and has many root causes. Much now is known about its causes and cures, even how to test for it. Roughness may show up only with cold brakes, only with warmed brakes, sometimes for all brake applications. Most vehicles have suspension and steering systems that get excited into greater vibration amplitudes at certain vehicle speeds (e.g., 30 mph). Prior brake usage history affects brake temperature distributions, their resultant brake thermal distortions, and thus also the tendency toward roughness. Experienced test drivers often choose a smooth road, then use specific vehicle speeds and brake usage sequences to search for brake roughness. Different vehicle suspensions, different steering systems, different caliper designs, different brake rotors, and different brake linings all can change the occurrence and severity of brake roughness. New Vehicle Roughness New vehicle start-up time is often a time of major concern about NVH problems in general and brake roughness in particular. Prototype vehicles may have brake, suspension, and steering components that differ from the initial production parts. At times the new parts may appear to be better, closer to print nominal values, better finishes, etc. However, if they are different in any way, they possibly may show more roughness. Even if the production parts are unchanged, the larger number of vehicles from full production may provide some with disc brake roughness. Some new vehicles may exhibit a brake roughness, often including a pulsing brake pedal feel, especially during a light brake application. These symptoms may disappear after a few brake applications and not return. If so, they probably resulted from contamination of the rotor surface. Local rusting of the rotor and/or oil/grease/paint contamination of the rotor may be the causes. If the problem worsens with usage, a systematic diagnostic is justified. Rotors from problem vehicles should be measured for thickness variation (TV), lateral runout, and runout second harmonic. At a minimum, this should be measured at the rotor mid-plane, but preferably also near the OD and ID. Vehicles vary in their sensitivity to rotor dimensional characteristics. Such sensitivity studies should be performed using production brake linings for the vehicle. Some brake linings have different elastic and frictional properties, so they influence the rotor dimensional requirements for an acceptable brake rating. The brake linings used to evaluate brake roughness should be fully burnished. This is required to assure the rating corresponds to steady-state customer usage conditions. When rating tests are run, the brake mechanic needs to be extremely careful to assure that neither the test linings nor the test rotor rubbing surfaces are contaminated by finger contact, or oil, grease, paint, and similar materials. High Mileage Brake Roughness Some semimet and NAO brake linings cause brake roughness to worsen with time and vehicle mileage accumulation. This type of disc brake roughness results from a combination of abrasive brake lining surfaces and frequent highway/ expressway driving. At least 1500 miles (2500km) of highway driving conditions, with a minimum of brake usage, is needed to develop high mileage roughness. Since it is mileage and usage sensitive, high mileage roughness may not appear until after 20,000 miles. Many roughness symptoms only show up after ten thousand highway miles. It is not uncommon for drivers to first notice brake roughness after an extended driving vacation, since this type of driving hastens roughness occurrence. Why roughness gets worse with high mileage, but minimal brake usage is quite simple, but not obvious. Abrasive particles at the brake lining surface can be the first to contact the rotor. Under normal brake pressures, and when the brakes are heated, most abrasive particles are imbedded into the brake lining matrix. This limits their abrasive action. However, when driving at highway speeds with the brakes released and cool, a brake lining may gently and locally rub the rotor. The abrasive particles may ‘stand proud’ of the surface and dominate the contacts at such times. Eventually, this local contact of the rotor by the brake lining (especially by abrasive particles at the lining surface) locally wears the rotor. This local rotor wear provides a rotor thickness variation, called TV. TV produces uneven braking torques that may be especially noticeable on gentle brake applications. The resultant periodic brake torque variations, and their associated brake pedal pulses, provide initial brake roughness. Local Inboard Wear of Rotor It is important to remember that a very localized rotor wear produces most brake roughness. This local wear almost always is produced during vehicle usage when the brake is released. It is commonly worse when the brake linings are cool (below the binder resin glass transition temperature). Under these conditions, a small amount of local brake dragging wears the rotor at the local contact site. With most disc brakes, this wear is confined primarily to the inboard rotor face. The section Brake Design Factors provides a more complete explanation why the inboard brake linings cause most brake rotor TV problems. Thermal TV Once the rotor has developed significant TV, gentle brake applications provide uneven heating of the rotor. This thermally induced TV increases the initial rotor TV. Now the brake roughness is more severe. When the vehicle speed reaches one where this brake roughness input can excite some suspension or steering component, the observed brake roughness peaks. MEASURING DISC BRAKE ROUGHNESS Vehicle Roughness Measurements Drivers sense brake roughness through the brake pedal, steering wheel, seat assembly, floorboards, as well as through both visual and audible inputs. These are difficult to quantify repeatedly. Most customer complaints on brake roughness come from the drivers. Experienced brake test driver roughness ratings are fairly repeatable, and are needed for final vehicle ratings. However, they are not well suited for brake roughness diagnostic studies or for component evaluations. These require instrumentation ratings. Common methods of vehicle roughness instrumentation are strain-gaged drag struts and torque wheels. Both permit instrumented readings of brake torque averages and torque variations. All brakes have some torque variations, but not all torque variations are at wheel frequency and large enough to be detected as brake roughness. Instrumentation of the drag struts appears to offer both advantages and disadvantages, compared with the torque wheels. Torque wheels are self-contained, not requiring application of the instrumentation directly to each test vehicle. However, they may provide a different wheel offset, mass, and stiffness than the OE vehicle has. They also may affect brake cooling rates and temperature distribution. This may affect the brake roughness amplitude and occurrence conditions. When several test vehicles of the same make and model are to be evaluated, torque wheels can be quite acceptable. If a number of samples for a particular vehicle is to be evaluated, for example to obtain an initial quality rating, the use of torque wheels can be quite effective and efficient. It should be remembered that torque variations do not necessarily correspond with the brake force output variations, such as seen by the drag strut, so torque data alone may not correlate well with driver ratings. Drag Strut Measurements: Strut instrumentation is particularly useful to characterize individual vehicles for roughness sensitivity. For example, a known set of rotor/lining sets can be evaluated on a particular vehicle to establish that vehicle’s suspension sensitivity to roughness. It is known that soft suspensions and soft strut bushings make vehicles more sensitive to brake roughness inputs. An instrumented test was run with the same sets of brake rotors and linings on three vehicles. Each had a different suspension and/or strut bushing stiffness. The measured torque variations were over three times greater on the vehicle with soft strut and suspension bushings (Figure 1). It appears that strut bushing instrumentation is better for developing and tuning suspensions to minimize vehicles response to brake roughness inputs. Strut bushing test data (e.g., absolute amplitude or ratio of force variation to average force) has provided a good correlation to experienced test drivers' roughness ratings. Instrumented vehicle struts generally provide better vehicle roughness response data than instrumented torque wheels. Dynamometer Roughness Measurements: Most brake dynamometers have strain gage torque sensors that can provide the needed brake torque average and variation numbers. However, a brake dynamometer does not have the same brake mounting compliance as on a vehicle, and is connected to the drive motor and load inertia by means of a drive shaft and couplings -- not a wheel and tire. In its basic form, a brake dynamometer can provide useful data on brake roughness. The ratio of peak-peak torque amplitude to average torque provides a measure of the brake roughness input. A brake dynamometer can measure differences in this torque ratio for different test temperatures, different brake apply pressures, and at different times during a simulated brake application. Brake roughness output, the observed vehicle response, varies substantially with this input. Both brake roughness input and output measurements are needed to determine the best approach to reduce brake roughness in the vehicle. Brake dynamometers normally do not provide information on how brake torque variations may interact with such things as suspension geometry and component compliance. Few brake dynamometers have the capability to include an entire vehicle corner -- complete brake assembly, suspension, and structural components. Very few brake dynamometers absorb torque through tire/wheel assemblies. However, almost any brake dynamometer can roughly simulate brake roughness deflections by the addition of a spring element (even an actual strut bushing) to the brake tailstock reaction arm. This spring should be installed in series with the brake torque load cell. The spring allows a test brake on a dynamometer to have nearly the same vibrational frequency as the wheel/tire/brake assembly on a vehicle. It is not known if dyno windup springs improve the correlation of roughness data from brake dynamometers to vehicle drivers. The important consideration is that the brake dynamometer readily provides brake roughness input data, and can be modified to provide some simulated output data. Vehicle roughness response characteristics, for the same brake input, may be quite different from one vehicle to another. It may be preferable to measure vehicle suspension response versus frequency behavior using shakers at both front wheels to simulate brake torque variations. This needs be done for each vehicle platform. Such data then can provide brake roughness torque variation bounds to achieve different roughness ratings. Figure 1 shows the drag strut response signal on three different vehicle platforms with the same brake, under the same testing conditions. Most, but not all, of these differences were attributed to the strut bushing stiffness differences. Note the clear differences of signal amplitude and frequency that occurred before the brake was applied. CAUSES OF DISC BRAKE ROUGHNESS Brake roughness is excited by excessive brake torque variations. These may result from one or more of several brake-related sources, most of which are first order. By first order, this means that a significant event occurs only once per wheel revolution. Examples are:
Brake torque variations have their roots in brake design, materials, manufacturing, and usage history. However, there is more to brake roughness than simply the excitation by the brake. Vehicle Design Factors The same brake hardware, installed in different vehicles, can provide large differences in reported brake roughness. Even when tested by the same drivers, the roughness ratings are clearly different for different suspensions and steering systems. As with most vibrations, the brake roughness response is a function both of the brake excitation and of the vehicle system response to that excitation. Since the vehicle response to brake roughness inputs also is intimately tied to vehicle ride and steering behavior, brake engineers usually have to be content to address brake roughness problems primarily through brake system modifications. Such constraints make roughness fixes difficult to achieve on luxury vehicles with soft suspensions. This report does not address vehicle suspension and steering design changes to reduce observable brake roughness. However, the non-brake contributions to reported brake roughness problems should be recognized. Brake Design Factors -- Sliding Calipers Most disc brakes today have sliding calipers, either pin or rail slider types, with pistons that use their seals for retraction. With such a design, if the outboard brake lining starts to drag against the rotor, its caliper readily moves over to reduce the contact travel to a minimal value. This happens because the stiffness of the caliper assembly is high and its sliding force is low. On the other hand, if the inboard brake lining drags against the rotor, the caliper piston (suspended by the rubber seal) has a low stiffness, so it can move readily. The predisposes sliding caliper brakes toward dragging of their inboard linings. This is further biased by the normal displacement of the rotor, during cooling, toward the inboard lining. The bottom line is that sliding caliper disc brake designs have an inherent tendency toward dragging of the inboard lining. The caliper piston travel, using its seal for a spring, may be 0.0015 to 0.002" for a dragging brake with rotor runout. The rotor contact, as might be expected, is along the runout ramp before, and after the point of maximum runout. Measurements tend to show brake dragging contact from about 50 to 70 degrees before maximum runout to about 10 to 50 degrees after maximum runout on the inboard rotor face. When the rotor has a high runout, the worn zone usually stays within 0.0015" of the maximum. This makes the worn zone narrower, with a resultant sharper brake torque pulse from the TV. Brake lining drag wear is typically only on the inboard side of the rotor for most pin and rail slider caliper designs. When outboard wear is found, it normally is only a fraction of that found on the inboard side, and 180 degrees offset in location. Road crowns tend to provide a greater contamination to the right side brake assembly in the US and other countries with right hand traffic. Consequently, the typical situation is for more contamination-based rotor wear to show up on the right side rotors. Fixed Calipers Fixed calipers can and do get TV wear on both sides of the rotor. If the brake linings are abrasive, the outboard wear can be about the same as that of the inboard. Fixed calipers are less common than sliding calipers. They tend to be used with rotors that have less runout, less tendency toward distortion, and are likely to have suspension systems that are insensitive to brake roughness. There is not a great deal of data available on fixed caliper disc brake roughness. Old data, from early Lincoln and Thunderbird fixed caliper disc brakes, indicated their roughness was more noted when very hard brake lining were used, and when the vehicles were driven in regions where abrasive road dust was prevalent. MANUFACTURING-RELATED CAUSES Thickness Variation Since disc brake roughness is directly related to brake torque variations, it is logical that variations in the thickness of a rotor, called thickness variation or TV, would be important. Most caliper disc brakes have a very limited tolerance for TV before the brake roughness becomes unacceptable. For this reason, disc brake rotors are generally machined on both rubbing faces at the same time. This may involve a straddle cutting on a lathe, or a grinding operation that machines both faces on the same setup. Since cutting tools and grinders have some compliance, it is important that the roughing operations provide a minimal level of TV as well. Runout All disc brake rotors have some runout. It is not possible to eliminate all runout, since this involves bearing machining, bearing seats, bearings, machine setups, and so forth. A small amount of runout generally will not induce a detectable brake roughness, at least initially. Large amounts of disc runout require the caliper and brake lining assemblies to move laterally with the runout, or the brake clamping forces will vary with angular position. If it does, the brake may develop roughness immediately due to the brake force variations. It also may develop brake roughness during a prolonged low-pedal-force brake application, for example during a slowing for a freeway exit or a downgrade. During such braking, the rotor will become heated unevenly as a result of the uneven clamping forces. This uneven heating of the rotor can increase the rotor runout and provide a significant increase of rotor thickness variation as well. Mass Imbalance Rotors may have castings that provide uneven mass distributions with angular position. These will respond to an even heating from brake application with an uneven change of thickness. This thermally induced TV tends to be self perpetuating, once initiated. Ideally, the rotor contact faces should not vary in thickness with angular position. Residual Stress Some TV change after machining is possible if the gray iron casting is not stable. Initial heating of the rotor has been reported to produce permanent changes of most dimensions, with runout changes being larger than those for TV. Surface Texture Uneven or irregular surface texture is not often a source for disc brake roughness. However, the initial roughness rating for new vehicles has been found to be sensitive to grinder alignment and bearing effects, when they produce an uneven surface texture on the rotor. Lathe turning is not known to produce uneven surface texture, but poor casting, with porosity, hardness, inclusion, free ferrite variations can result in finished rotor rubbing surfaces that vary with angular position. Surface Coatings Rotors at times are given a surface treatment, for example to provide rust protection. It is important to be aware that any coating that affects the friction level, or is depleted through wear, be thoroughly tested for its effect on brake roughness. While such coatings may only temporarily affect brake roughness, any detectable adverse effect can elicit a strong negative first impression by the customer. USAGE-RELATED CAUSES Brake Parasitic Drag Wear When a disc brake is released, the piston seal roll-back retracts the piston several thousandths of a inch. This small retraction is needed to minimize brake pedal travel for initial lining contact. However, the small seal roll-back may result in some local brake lining contact when the brake is released. This is called parasitic drag. Normally this drag is small, about the same as wheel bearing or seal drag. However, it can have serious brake roughness consequences under certain circumstances. Abrasive Brake Linings Some brake linings contain abrasives as a part of their composition. For example, many semi-met friction materials contain fused magnesium oxide of a particle size that can be abrasive to the rotor. Abrasive materials also may occur as unwanted, 'tramp' constituents in brake linings. Silicon carbide is a known tramp abrasive material that may be found in synthetic graphite. Accumulated surface materials, such as road dust or rotor rust particulate may collect on the brake lining rubbing surfaces. Some abrasive material is possible in and on a brake lining surface. The harder the brake lining matrix, the smaller the abrasive particles need be to wear the rotor. For this reason 'soft' brake linings and warm brake linings tend to wear the rotors much less aggressively. When the brake is nominally released, but with some parasitic drag, the brake lining surface periodically contacts a portion of the rotor surface. If the brake lining surface that contacts the rotor is abrasive, even this light contact may result in a local rotor surface wear. Such wear results in usage TV. This wear is generally on the inboard face of the rotor. The reason for this is that most disc brake calipers have their pistons on the inboard side, with a sliding mechanism of the caliper for the outboard shoe loading. If abrasive contamination comes from road dust, the outboard rotor wear can be much less, especially for vehicles with closed disc wheels or with closed wheel covers, These minimize abrasive particulate entry to the brake. If spoked wheels with large opening are used, both rotor faces may wear about the same from road-borne abrasive contamination. Road crowns tend to provide a greater contamination to the right side brake assembly (in the US and other countries with right hand traffic). Consequently, the typical situation is for more contamination-based rotor wear to show up on the right side rotors. Runout Induced TV A TV change of 0.0015" was measured on a large passenger car rotor during a slow brake application from 80 to 45 mph. The initial thickness variation was under 0.0001" on this rotor, so it was not considered as a possible cause for brake roughness. It is important to remember that a caliper disc brake is fundamentally unstable in terms of thickness variation. The first and second order components of the rotor runout result in some variation of brake lining contact pressure. This starts some thermally-induced thickness variation. Any TV will tend to increase with time during a prolonged light brake application. Some brake linings are more likely to generate regional hot spots, and associated brake roughness. Soft (in compression) brake linings are better than rigid materials, as they tolerate the runout with less frictional force variation. Lighter weight calipers and free moving calipers similarly reduce the vehicle sensitivity. On one brake, an increase of hydraulic brake line size also reduced the runout induced TV. It was presumed that the caliper piston was 'softer' with the larger hydraulic line, again reducing the brake lining drag force variation with runout. Disc Brake Roughness Case Histories may help in the understanding of brake roughness. These vehicle-specific cases have been edited to short summaries, and all vehicle model and vendor ID eliminated. Case 1. Instrumented steering wheels on car with severe case of roughness -- recorded angle was as high as 64 degrees (peak-to-peak) steering wheel oscillation. Light steady braking from highway speeds made it worse. Both rotors were involved, with TV ‘phasing’ important in the magnitude of steering wheel response. Fixing one side stopped all steering wheel oscillation, but brake pedal pulsation remained. Driver complaints were found to occur primarily when both rotors had excessive TV and when the rotor phasing caused steering input. This phasing effect caused the roughness to vary substantially, even for the same testing conditions. Case 2. Composite rotor – A stamped rotor hat section provided greater runout than typical of full-cast or composite rotors. Roughness was found to be increased after extended highway-type usage with cold brakes. Semimet linings had abrasive content that caused inboard face wear under these usage conditions. Inboard rotor surface wear was found to occur mostly between 70 degrees before peak runout to10 degrees after. Highway usage-induced roughness did not occur when drivers used brakes enough to keep the linings above their glass transition temperature of 82C. Case 3. Different disc brake behavior was noted on 3 platforms—using the same brake parts, but with different suspensions and drag strut bushings. The drag strut bushing spring rate was the greatest single variable that affected observed roughness. Initial rotor TV was over 7 times greater in effect than initial rotor runout. However, runout was the greatest root source for high mileage TV increase and high mileage driver complaints of roughness. Case 4. New Utility Vehicle launch–proprietary coating on rotor rubbing surfaces was not always even, causing brake roughness until coating was fully worn off. LINK ENGINEERING COMPANY © 1995, All rights reserved reproduced on ifriction.com with permission of Link Engineering For more information visit: Link Engineering Or Contact: Link Engineering Email |