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

Подождите немного. Документ загружается.

the accuracy required, provide proper alignment, and rotational control. Since galling is

an adh esive phenomena, proper cleaning of the specimens is a critical factor in this test

and this is emphasized in the test procedure. The information that should be reported with

resultsisshowninFig.9.32

. The method also provides guidance as to how to examine the

surfacesforevidenceofgalling.ExamplesfromthistestareshowninFig.9.33

This test is an example of one in which the appearance of the worn surface is used to

evaluate the wear resistance of materials. When such an approach is used, it is necessary to

deﬁne the type of technique to be used in evaluating the appearance. One aspect of this is

magniﬁcation. Different techniques could enhance or degrade the identiﬁcation of certain

features and variations in precision would result. In additi on, different techniques might

be sensitive to different aspect s of the phenomena and in effect would result in varying

criteria. To address these co ncerns, the test method speciﬁes unaided visual inspection for

determining whether or not galling has occurred. While qualitative approaches, like the

one used in this test, are more subjective than quantitative approaches and as a result tend

to be less precise, they can be effective and consistent. Good inter and intralaboratory

Figure 9.32 Data to be reported in the ASTM test for galling.

reproducibility has been found in interlaboratory testing done for the development of

the ASTM standard. Experience has shown that a 5–10 kpsi difference in the galling thre-

shold stress is needed before a difference in ﬁeld perfor mance is observed. Interlaboratory

testing programs, utilizing this standard method, generally have shown repeatability

within that range.

9.2.10. Rolling Wear Test

A conﬁguration that has been successfully used to address rolling wear is illustrated in

Fig.9.34

. Basically, it consists of a pair of driven rollers pressed against one another.

The typical procedure is to visually monitor the condition of the roller surfaces and deter-

mine the number of cycles for a selected level of surface damage to occur. This could be the

appearanceofcracks,surfacetexturechange,orspalls.Figure9.35

illustrates examples of

this type of damage and wear. These tests are usually quite long, extending for days or

weeks and inspections are done on a periodic or scheduled basis. This is another example

of the use of appearance criteria in a wear test. The longer the number of cycles, the more

resistance the pair is to rolling wear.

Critical elements of this test are control of the surface velocities of the two rollers,

alignment of the rollers, and geometrical tolerance of the rollers. With respect to the last,

Figure 9.33 Examples of wear scars produced in the test for galling. ‘‘A’’ shows examples of sur-

faces exhibiting galling in the test. ‘‘B’’ shows the results of a test sequence used to determine the

galling threshold. (From ASTM G98, reprinted with permission from ASTM.)

Figure 9.35 Examples of wear produced in a rolling test. ‘‘A’’ shows various stages in the forma-

tion of a single spall. ‘‘B’’ shows the appearance of the surface of a roller with extensive spalling.

(‘‘A’’ from Ref. 18, reprinted with permission from NLGI; ‘‘B’’ from Ref. 27, reprinted with

permission from Penton Publishing Co.)

Figure 9.34 Basic conﬁguration used in rolling wear tests.

particular attention has to be paid to the edge conditions of the rollers so that a signiﬁcant

stress concen tration condition does not occur. This means that the edges of rollers should

be well rounded. Use of rollers of the same length can help to minimize this exposure, as

well. Another approach that has been used is to use slightly curved or crowned rollers,

asisshowninFig.9.36

. This type of test has been used to address conditions of pure

rolling, in which case the surface velocities of the two rollers must be identical. In addi-

tion, the test has also been used to address conditions of mixed rolling and sliding. In this

case, the relative velocities must be control led so that the proper ratio of sliding to rolling

is achieved and maintained. There are two elements to controlling the velocity; one

element is the rotational speeds of the two rollers. The other is the radius of the rollers.

In addition, control of the preparation and cleaning of the rollers are important, as well

as the consistency and uniformity of the materials and lubrication, if used.

This test has been used for the evaluation of material pairs for rolling applications

such as gears, cams, roller bearings, and ball bearings. In these types of applications,

additional forms of wear might also be present. For example, different regions of gear

teeth are exposed to different modes of wear. Rolling predominates along the pitch line,

while sliding predominates at other locations along the tooth proﬁle. Nonetheless,

generally good correlation has be en found between this test and actual performance

for those regions where rolling is the major characteristic. One way in which this type

of test has been used is to develop data in conjunction with a model for rolling wear

(27,28). The basic concept of that model is that there is a power law relationship between

the number of cycles to failure and the stress level under which rolling takes place, as per

the following equation:

N ¼ N





ð9:5Þ

N is the number of cycles to failure at stress level S and N

is the number of cycles to

failure at stress level S

. The parameters of the model that need to be determined experi-

mentally are N

and n. This can done by performing a series of tests at different stress

Figure 9.36 Conﬁguration of a rolling wear test in which curved rollers are used.

levels and determining the number of cycles to failure for each of the stress level. By ﬁtting

these data to Eq. (9.5), the values can be determined or a curve deﬁned that represents that

type of equation. Examples of several curves characterizing pairs of material s are shown in

Fig.9.37

. This is a log–log plot and Eq. 9.5 is a straight line on such a plot. Consequently,

a mini mum of two tests at different stress levels is required to deﬁne the parameters.

It is generally desirable to do repeated tests and ones at additional stress levels. The

differences in stress levels and number of repetitions should be large enough that a good

estimate of n can be obtained. Changing geometry, size, or load can vary stress level.

However, it is usually more convenient and preferred to perform tests at different loads,

since specimens can then be made to a single size and shape. Changing size and shape

may introduce or change other factors, particularly the amount of slip and alignment,

which can result in increased variability or scatter.

The same basic test and procedure is used with different amounts of sliding intro-

duced.ExamplesofcurvesobtainedforthesemixedconditionsareshowninFig.9.38

Again, the data are ﬁtted to Eq. (9.5). In general, different values of N

and n are obtained

for different percentages of sliding. The percentage of sliding is deﬁned as

% Sliding ¼ 2 

 V

þ V

 100% ð9:6Þ

where V

and V

are the surface velocities of bodies 1 and 2, respectively. At some

percentage of sliding, the morphology of the wear changes to one more characteristic of

sliding wear (e.g., scratches, scuff ed appearance, adhesion, etc.). This would represent a

limiting value for the rolling wear model. This transition point can be established by

performing tests at different percentages of sliding and looking for this type of change

in morphology. Since the value of this transition point can vary signiﬁcantly with materials

pairs (e.g., 9–300%), it is often of engineering signiﬁcance.

Testing in this fashion (i.e., to determine values of parameters of a model) provides

the test with the element of being able to determine absolute performance, as well as accel-

eration. Eq. (9.5) can be used to predict or project actual performance if the stress level in

the application is known. Acceleration is provided by being able to deﬁne N

and n in

shorter periods of time by performing tests at stress levels above those in the application.

For the simulation of application, it is important to insure that the stress levels in the

Figure 9.37 Examples of the relationships between stress and number of revolutions obtained in

rolling wear tests with different materials. (From Ref. 27.)

test remain in the same range as that of the application. Generally, this means that the

loads used in the test should result in stresses in the elastic range. If the test loading is

such that plastic deformation occurs, the test method is not valid. This can be a practical

problem for some materials, like thermoplastics. For these materials, stress levels required

to avoid creep or deformation might result in very long test times and it may not be a

practical way to evaluate these materials.

9.2.11. Reciprocating Pin-on-Flat Test (Oscillating Ball-Plane Test)

Generically this test is very similar to the pin-on-disk test and used for the same purposes.

It differs only in the type of motion. The generic features of a reciprocating pin-on-ﬂat or

ball–planetestareshowninFig.9.39

. One difference between the ﬂat and disk tests is the

shape of the ﬂat member of the contact. In the pin-on-disk, it is a disk to accommodate

rotation, while with the ﬂat and plan test conﬁgurations, it is normally a rectangular block

or ﬂat specimen. However, the fundamental difference is with the type of motion that each

provides. The motion is generally unidirectional at a constant speed in the pin-on-disk

test. In the pin-on-ﬂat test and ball-on-plane test, there is a reversing of the direction of

sliding and the speed may vary throughout the cycle. One of the consequences of the

change in directions is that each cycle contains acceleration and deceleration portions,

an element that is not present in the pin-on-disk test. The velocity proﬁle tends to vary

with different apparatuses and depends on the nature of the drive mechanism used. For

example, the proﬁle is sinusoidal if a rotating eccentric is used. If a linear stepper motor

is used, it would have a square wave proﬁle. These differences in the motion can inﬂuence

wear behavior for a variety of reasons including the inﬂuence of debris, build-up of trans-

fer and third-body ﬁlms, lubrication, and fatigue wear mechanisms (which ca n be inﬂu-

enced by stress reversals). Consequently, the motion in one type of test could provide

Figure 9.38 Examples of the relationships between stress and number of revolutions obtained in

rolling tests involving slip. (From Ref. 27.)

better simulation for an application than the motion in the other type. While this potential

exists and must be recognized, it generally does not appear to be a major factor. Both tests

have been used effectively to address wear concerns in both unidirectional and oscillatory

applications.

It is also desirable to measure friction in conjunction with a pin-on-ﬂat or ball-on-

plane wear test, as it is with the block-on-ring and pin-on-disk tests. Therefor e it is a good

practice to incorporate this capability into these types of apparatuses. Many of the ones

described in the literature (29) have this capability. One advantage or use of these

oscillating tests is that its oscillatory nature makes it suitable for the simulation of fretting

or fretting corrosion situations. This is done by reducing the amplitude of the motion

to the range of associated with fretting (30,31).

Like with the pin-on-disk test, these reciprocating test s can be used to rank material

pairs in terms of wear resistance. ASTM G133 describes a standard test method for such a

purpose, which utilizes a sphere or spherical-ended pin. Like the ASTM test method for

pin-on-disk, ASTM G99, wear volume after a ﬁxed amount of exposure (number of cycle)

is used. The coefﬁcients of variation from interlaboratory test programs are between 20%

and 30% within a laboratory and about 50% between laboratories. This test method and

ASTM G99 (pin- on-disk test) are useful guides for conducting these types of tests. The

ball–plane test has also been used in a different manner to address engineering wear situa-

tions (30–35). With these uses the same methods for providing control apply but with

some modiﬁcations that are associated with the taking and analysis of the data. However,

the measurements made and the analysis techniques tend to be different for these two uses.

If the standard method is used, the volume of wear generated after a speciﬁed amount of

sliding is used to rank the material pairs (with the caution that the wear curves in these

tests are frequently nonlinear). In the engineering use, a linear wear dimension (e.g., such

as scar width or depth) is often used and a wear curve developed, rather than a single

measurement.

The following is a description of a method that can be used for engineering evalua-

tions with a ball–plane test and extended to the pin-on-disk test as well. In addition,

some elements of this approach can also be applied to the block-on-ring test. In many

of the engineering applications of this test, the primary wear measure is the depth of the

wear scar. This is usually determined by means of a proﬁlometer measurement. Fig. 9.40

shows typical traces for the ball and the plane. With the plane, the trace automatically

Figure 9.39 Conﬁguration of reciprocating pin-on-ﬂat test, using a spherical specimen as the

‘‘pin’’. The shapes used for the pin are the same as with the pin-on-disk test. (See Fig. 9.25.)

provides a reference level to determine wear depth. For the ball, some technique com-

paring the unworn proﬁle with the worn proﬁle is often used. A graphical over-lay tech-

niqueisshowninFig.9.41

for a ﬂat wear spot on the ball and for a more general

condition in Figs. 9.28 and 9.30. Analytical techniques, which are based on the measure-

ment of the width of the spot, can also be used (36). To improve accuracy with an over-

lay technique, it is desirable to use unworn proﬁles of the actual specimen rather than a

theoretical or nominal shape.

The method involves the use and analysis of a wear curve, generally plotted on a

log–log scale. Conducting tests of different duration generates the wear curve. The inter-

vals and overall duration of the test vary with the situation or application being

addressed. The concept is to generate a well-deﬁned wear curve which is representative

of the wear in the application. As a rule, this means adjusting the total measurement

interval to extend from the smallest amount of sliding required to produce a measurable

Figure 9.40 Proﬁlometer traces across wear scars occurring in ball–plane wear tests. ‘‘A’’ and ‘‘B’’

are for the ball specimen; ‘‘C’’ and ‘‘D’’, for the plane surface. The radii of the spheres used and the

magniﬁcations of the traces are different.

wear depth to that required to produce a signiﬁcant increase in the wear depth (e.g., for

example an order of magnitude increase or more). It is generally appropriate to select

logarithmic or half-logarithmic intervals since the wear depth typical increases in a less

than linear fashion. To characterize the wear behavior the data are then ﬁtted to a

power relationship

h ¼ C  S

ð9:7Þ

where h is the depth of wear and S is the amount of sliding, number of cycles, or time. This

is usually best done by plotting data on log–log paper and ﬁtting the log-form of Eq. (9.7)

to the data. Best-ﬁt values for C and n are used to characterize the wear behavior of the

system.TypicalwearcurvesobtainedinsuchafashionareshowninFig.9.42

. In effect,

the standard test method only uses C for this characterization, which presumes that n

is a constant.

There can be poor correlation between the wear data and Eq. (9.7) in some instances,

as in one of the cases shown in Fig. 9.40, curve 2. In such cases, the wear data are ﬁtted in a

piece-wise fashion. Usually, it is sufﬁcient to separate the data into tw o regions, such as

initial and long-term behavior. The data in each region are ﬁtted by an expression of

theformofEq.(9.7)(Fig.9.43

). In such cases, there is a signiﬁcant difference in n for

the two regions. Generally, this piece-wise ﬁt is necessary as a result of transitions

in wear behavior, which can be associated with observed changes in friction and wear

scar appearances . In some cases, additional regions may need to be considered in the

same manner (34).

Once the wear curve is deﬁned in terms of one or more pairs of C and n values, the

results can be interpreted in terms of various models or theories. These models can then be

Figure 9.41 Illustration of the overlay technique used to determine wear depth on a curved surface.

‘‘A’’ shows the proper technique; ‘‘B’’, an improper technique. In using the technique, the wear scar

should be at the apex of the trace.

extended to the a pplication. The following example for the case of a ﬂat spot being

worn on a ball illustrates this type of consideration. Because of the geometries involved,

a linear relationship between S and wear volume results in a value of n close to 0.5. If

the value obtained for n is close to this, it would imply stable wear behavior, as well as

the wear being consistent with models for abrasive and adhesive wear. However, if a value

of n signiﬁcantly greater than 0.5 were obtained, this would impl y that there is some

transition in wear occurring, as is illustrated in Figs. 9.43 and 4.46. Likewise if a value

of n lower than 0.5 is obtained, this suggests a wear mode, referred to as a variable energy

wear mode, which is associated with fatigue wear and stress-dependent, is dominant (24).

Figure 9.43 Example of the piece-wise ﬁt of wear data to relationships of the form, Wear ¼ K 

Usage

Figure 9.42 Example of wear curves obtained in ball–plane tests. Curves 1, 2, and 3 are for the

depth of wear on a steel sphere sliding against ceramic ﬂats of different roughnesses and with

different lubrication conditions. Curves 4 and 5 are for the depth of wear on two different steel ﬂats

being worn by ceramic spheres of different radii and with different loads. (From Ref. 136.)