Bayer R.G. Mechanical Wear Fundamentals and Testing, Revised and Expanded

Подождите немного. Документ загружается.

The general goal of these bearing tests is to correlate design and application param-

eters with life. This is typically done by adjusting and controlling the bearing and test

conditions to those of interest, including such things as the materials, dimensions, and

lubrication system of the bearings. Also, the loading conditions of the applications are

simulated, as well as the nature of the motion, stop=start conditions, and the environment.

The general approach is to monitor the wear measure during the test. In the case of ball

and roller bearings, the duratio n of the test is usually until failure occurs. This is often the

case with journal bearing tests as well, but these tests may be carried out only to the point

needed to establish a stable wear rate. These stable wear rates are then used to project life.

Figure 9.84 Examples of testers used in journal bearing wear tests. (‘‘A’’ from Ref. 89; ‘‘B’’ from

Ref. 129; ‘‘C’’ and ‘‘D’’ from Ref. 130; ‘‘E’’ and ‘‘F’’ from Ref. 131; reprinted with permission from

Butterworth Heinemann Ltd.)

These tests provide a value for the life of the bearing for a set of operational con-

ditions and, if the simulation is complete enough, this can be a direct assessment of ﬁeld

life. However, there are more sophisticated approaches and uses of the data beyond this

direct and simple use of the test results. These methods of using the data vary with the

type of bearing. For ball and roller bearings, the tests are frequently used to establish the

parameters in the load-life relationships and adjustment factors used for these types

of bearings (86,90–93). For journal bearings, a common approach is to use a P-V

concept to interpret the data (94). In this case, the concept is either to identify acceptable

Figure 9.85 Examples of testers used in roller and ball bearing wear tests. (‘‘A’’, ‘‘B’’, and ‘‘C’’

from Ref. 131, reprinted with permission from FAG Bearing Corp.; ‘‘D’’, ‘‘E’’, and ‘‘F’’ from

Ref. 132, reprinted with permission from Butterworth Heinemann Ltd.)

Figure 9.85 (continued )

Figure 9.86 General conﬁguration of bearing wear tests: ‘‘A’’, journal bearings; ‘‘B’’, ball and

roller bearings.

combinations of pressure and velocity as a function of all the other variables or to deter-

mine the wear rate for conditions where the product of pressure and velocity are

constant as a function of the other parameters. These methods are covered in Secs.

2.7 and 2.8 (EDW2E).

Bearing tests of the type discussed are generally quite long, since they are similar to

applications in terms of operating parameters. Because of this and the complex nature of

these devices, these type of tests are generally combined with phenomenological tests.

Phenomenological tests provide initial screening to select candidate material and condi-

tions for the bearing tests. They also pro vide the opportunity to investigate in more

detail speciﬁc aspects that are involv ed in overall bearing performance (56,87,95).

A variety of sli ding and roll ing wear tests is used in this fashion. Pin-on-disk and crossed

cylinder tests (Secs. 9.2.8 and 9.2.7, respectively) have been used to simulate sliding

wear aspects in journal bearings under boundary lubrication conditions and between balls

and cages in ball bearings. The rolling wear test discussed previously in the pheno-

menological section (Sec. 9.2.10) and ball rolling tests have been us ed to simulate race

and roller wear in ball and roller bearings.

9.3.6. Brake Material Wear Tests

Dynamometers of various types are typically used to evaluate the wear and friction

behavior of brake material systems (96–98). Such tests are used to evaluate both the

friction (brake) material and the counterface (rotor) material. Two illustrations of typi-

caldynamometerconﬁgurationsareshowninFigs.9.88

and9.89. With this type of

apparatus, the wear behavior is evaluated under braking conditions which simulate

application conditions. This typically results in a complex testing procedure or sequence

to provide adequate simulation. A typical example of a test sequence used for automo-

tivebrakesisshowninTable9.6

. These sequences simulate the synergistic effects of

break-in, high speed stops, low speed stops, pulsed and continuous braking, etc. These

same test sequences are also used in conjunction with different ambient conditions to

Figure 9.87 ‘‘A’’ shows an example of the appearances of the wear scars on the races of bearings

during the initial stages of wear. ‘‘B’’, in the intermediate stages. ‘‘C’’, in the ﬁnal stages. Which stage

represents ‘‘end-of-life’’ depends on the application. For most applications ‘‘B’’ would be more

representative than ‘‘A’’, and ‘‘C’’ is generally unacceptable. (From Ref. 93. Original source SKF

Industries, reprinted with permission from Texaco’s magazine Lubrication.)

simulate use under different climatic conditions and at elevated temperatures. Typically,

mass loss or thickness measurements are used to determine wear of the brake material.

These techniques are usually not appropriate for the rotor material since their wear is

typically quite small in these tests. As a result, the degree or severity of rotor wear that

Figure 9.88 Basic conﬁguration of a research dynamometer used for wear and friction evaluations

of brake materials. (From Ref. 96.)

Figure 9.89 Conﬁguration of a high-vacuum dynamometer used for wear and friction evaluations

of brake materials. (From Ref. 133, reprinted with permission from ASME.)

occurs is usually characterized in terms of roughness changes. These wear measurements

are taken at various points in the test sequence.

Since the typical use of these dynamo meter tests is to provide a basis for selection

of materials and design parameters for more extensive and costly ﬁeld evaluations, these

tests are designed to provide a relative ranking rather than an absolute determination of

wear behavior or wear coefﬁcients. One approach is to simulate the contact conditions

with a standard conﬁguration and a small sample of the brake material. In this case, the

dynamometers are smaller and less complex, and the tests are easier to implement. An

example of this approach is the use of the Chase dynamometer for these evaluations

(99). While this approach is attractive from an implementation standpoint, the reduction

in scale reduces simulation with the result that there is often poor correlation with ﬁeld

performance (99). Another approach that is used is to evaluate full-sized brake systems.

This generally results in the need for large r, more complex, and expensive dynam-

ometers and more complex tests. However, this approach has generally been found to

correlate well with ﬁeld performance (99). Because of this, full-sized testing is the recom-

mended method for establishing material rankings. The smaller scale tests are used for

more general purposes, such as investigating the effect of vacuum on performance and

determining general trends (100), but care must be taken in extrapolating the results to

speciﬁc applications.

The rankings in dynamometer tests are determined directly by the amount of wear

generated in the test sequence. The best performer is the one that has the least amount

of wear at the end of the test. While there is generally good correlation with test performed

with full-sized brakes, the tests do not provide universal rankings since the rankings are

for speciﬁc applications. Different rankings can be obtained with other tests. This is

because the test sequence is selected to simulate a speciﬁc application and, when a full-size

brakeisused,thewearperformanceisrelativetothatdesign.Table9.7

shows the results

of tests on four different materials using two different test apparatuses and test sequences.

In these tests, the rankings are based on thickness chan ge of the brake material. It can

be seen that differences in rankings are obtained with these two tests.

The complex and interactive nature of these full-scale dynamometer tests provides an

effective means of assessing wear performance in terms of application parameters. For

example, these tests provide a means of determining the effect of relative humidity, pul-

sing, or rotor roughness on brake performance. At the same time, this same nature inhibits

the determination of basic wear parameters or coefﬁcients of fundamental wear relation-

ships. As an example, they do not provide a means of determining the coefﬁcients of a

Table 9.6 Test Sequence Used in a Chase Sample Dynamometer Wear and Friction vs

Temperature Procedure

Initial burnishing Twenty min drag at 312 rpm with 100 lb load

with a maximum temperature of 200



Speed of drum 325 rpm

Load 350 in.-lbs

Test sequences,

Drum temperatures (



F) 250, 350, 450, 550, 650, 780, 250, 350, 450

Applications Forty of 20 sec duration at each temperature

Wear measurements Sample weighted and thickness measured after

each test at a different temperature

Source: Ref. 99.

fundamental equation proposed for wear of brake materials, Eq. (9.24), or verif ying the

applicability of the equation (101).

W ¼ KP

ð9:24Þ

In Eq. (9.24), W is the wear volume; P, the normal load; v, velocity; T, tempera-

ture. The situation is too complex and interactive in these full-scale dynamometer tests

to isolate the effect of these parameters and to allow the determination of the coefﬁ-

cients, a, b, and c associated with them. This is a co mmon feat ure of operational wear

tests. The effect of these parameters can be isolated and the coefﬁci ents, including K,

determined with phenomenological tests, which are generally simpler but less simulative.

9.3.7. Engine Wear Tests

Many of the wear situations encountered in engines are complex and are difﬁcult to

simulate. One aspect is the simulation of the local environment, for example, the com-

bination of combustion products and temperature that exists in various locations within

the engine. Another aspect is associated with the complex lubrication phenomena that

take place in engines. Lubricants contain active agents which, along with co mbustion

products, can react with the surfaces and form a variety of surface layers which can

inﬂuence wear beh avior in the boundary lubrication regime. The high speeds, amount

of lubricant involved, and the conforming nature of the components can result in

hydrodynamic lubrication as well. At many points within an engine, the occurrence

of wear is associated with the presence or absence of this type of lubrication. These

different aspects are interrelated as well. For example, temperature can inﬂuence both

boundary and hydrodynamic lubrication, which in turn inﬂuence friction, which in turn

inﬂuences temperature. Wear also inﬂuences the conformity between surfaces, which

inﬂuences hydrodynamic and boundary lubrication, which inﬂuence wear. With this

type of complexity to simulate laboratory engine tests provide a ‘‘natural’’ simulation

and are often used as tests to address engine wear concerns. Additional advantages with

this type of test are that naturally occu rring wear points can be identiﬁed, and many

wear points can be evaluated simultaneously.

An example of this type of test is in a study of wear between cylinder liners and pis-

ton rings (102). In this study, three identical engines were laboratory tested under different

conditions of service. One engine was operated at high speed, another was operated at high

load, and the third was operated under mixed conditions. At the end of the test sequence,

Table 9.7 Wear Data Comparison as a Function of Fade and

Recovery Performance Sequences

Total wear (in)

Friction material

Chase

(Schedule J661A)

Inertial dynamometer

(Schedule 111)

A 0.009 0.006

B 0.011 0.019

C 0.022 0.027

D 0.017 0.031

Source: Ref. 99.

the cylinder liners and piston were removed and examined for wear. Several wear regions

could be observed on these parts, each with diffe rent wear characteristics. Figure 9.90

shows a cut-away section of a worn liner from these tests. The general nature of this wear

(e.g., small with irregular outlines), along with the size and shape of the parts, makes it

difﬁcult to measure wear in terms of such measurements as volume, mass loss or dimen-

sional change. This is a common situation with these types of tests and therefore as a result

a variety of exami nation techniques and measur ement are used to assess the severity of the

wear and to provide a basis for comparison. In the liner study SEM, optical microscopy,

X-ray ﬂuorescence spectroscopy (XFS) , energy dispersive X-ray spectroscopy (EDS), and

metallographic techniques were used for this purpose. Examples of some of the results of

theseanalysistechniquesareshowninFig.9.91

. Wear rates were estimated using the

amount of original surface roughness remaining. In other engine-wear studies rad ioactive

doping techniques have been used to quantify the wear (103) as have such measures as the

average width of a wear scar or amount of edge rounding. Mass loss is usually not possible

because of the large mass of the parts and the tendency for combustion products to coat

the parts, as well as multiple wear locations.

The states of wear found for the three different operating conditions were very simi-

lar and were not considered as signiﬁcantly different. A fourth test, which was a ﬁeld test,

was also done as a part of this study. The liners and pistons from the ﬁeld test were

examined in the same manner as those from the laboratory tests. Little difference was

observed between the ﬁeld and laboratory tests. This illustrates the good simulation that

laboratory tests of this type (e.g., full scale) can provide. It should be recognized that at

the same time it is difﬁcult to quantify the results of these tests in terms of the more

fundamental measures of wear.

9.3.8. Tests for Glazing Coatings on Plastics

The purpose of these coatings is to protect plastic surfaces from optical degradation, either

in the form of reduced luster or transparency. These coatings are used extensively in the

mass transit industry to coat surfaces of such components as windows, windshields, lights,

and transparent panels. These surfaces can be worn or damaged by a variety of mechanisms

Figure 9.90 View of a cut cylinder liner from an engine test, indicating regions of different wear

behavior. (From Ref. 102, reprinted with permission from ASME.)

and as a result a variety of tests have been developed to address these different situations

(73). One of the modes encountered is the wear that is associated with the use of large,

rotating brushes in the washing of trains and buses; another is the wear associated with

the action of a windshield wiper. In both of these situations, sand or other abrasive par-

ticles can be drawn across the surface of the plastic by the action of the brush and the

wiper, causing wear. Operational tests have been developed to simulate both of these situa-

tions (104). The similarity of the tests to the wear situations is apparent in Figs.

9.92

and9.93. Controlled slurries, selected to be representative of those encountered in

practice, are used to provide the abrasive acti on in both tests. Loads and speeds

representative of these applications are used in these tests.

Wear in these tests is quantiﬁed by measuring the haze of the surface, which is

deﬁned in ASTM D1003. These measur ements are used in different manners in the two

tests. In the Brush Abrasion Test, the change in the haze value from the initial value is

used as a measure of wear for comparing and ranking coatings. In the Wiper Test, the

amount of haze produced in the test is the measure since the specimens are initially trans-

parent. This type of functional wear measure is commonly used for the evaluation of

wear on optical surfaces.

Figure 9.91 WearmorphologyatthelocationsindicatedinFig.9.90.‘‘A’’, unworn surface, A1;

‘‘B’’, A2; ‘‘C’’, A3; ‘‘D’’, A4; ‘‘E’’, A5. (From Ref. 102, reprinted with permission from ASME.)

Both of these tests correlate well with ﬁeld experience. This is to be expected as a

result of their high degree of simulation provided by these tests. However less simulative,

more phenomenological tests can provide good correlation as well, provided they

simulate the basic wear mechanisms involved in the application. For example, tests with

Figure 9.92 Diagram of test conﬁguration used to simulate the wear caused by brushes during

the cleaning of vehicles.

Figure 9.93 Diagrams of the test conﬁguration used to simulate the wear between wiper blades and

windshields.