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

Подождите немного. Документ загружается.

accuracy to provide material ranking in a reasonable length of time. In ad dition, they were

also chosen to avoid complicating elements, such as excessive heating of the rubber

or degradation of the abrasive. In the standard test, different practices are deﬁ ned for

different classes of materials to achieve these goals. Milder conditions are speciﬁed for less

abrasive resistant materials and coatings.

The ASTM test procedure also advises the visual examination of the wear scar to

insure that there is good alignment and no problem wi th the ﬂow of abrasives. In addition,

the test method also uses tests with reference mate rials to insure consistency. The ASTM

procedure also calls for the reporting and measuring of a number of parameters associated

withthetest.TherecommendeddatasheetisshowninFig.9.3

. This test conﬁguration can

be used for other purposes other than ranking materials, such as evaluating the abrasivity

of different materials. When used in this fashion, the standard test procedure can be used

as a guide in establishing proper control, etc.

One point should be made regarding the weight loss method for wear determina-

tion used in the procedure. To provide valid comparisons, the volume of wear should be

used; this requires knowledge of the specimen’s density. If this is not done, ranking

might be in error because of density differences. This is of particular concern when

Figure 9.2 Examples of wear scar morphology from the dry sand–rubber wheel test. (From

Ref. 113.)

coatings are evaluated, since the densities of the coatings may be unknown. Therefore

othertechniques,suchascomputationofthewearvolumefromwearscar

dimensions,

might yield better resul ts.

An important element to recognize in this test is that while it simulates a wear

mode and correlates to ﬁeld performance, it does not directly provide a wear constant

or parameter that can be applied to different situations. To determine absolute perfor-

mance in an application, a reference material for which there is ﬁeld and test data must

be used and a scaling factor established for that application. This is because the test does

not address the question of relative severity in the test and in application, either on an

empirical or a theoretical basis. This is also the case in several of the other tests that will

be discussed. This implies that there is an unknown amount of acceleration provided by

the test and that the acceleration could be different for different applications. The

amount of acceleration that exists can be signiﬁcantly dependent on differences in

the amounts and type of abrasives in the test and in application.

9.2.2. Wet Sand-Rubber Wheel Abrasion Test

This test also simulates low stress scratching abrasion and is very similar to the dry

sand-rubberwheeltest.AdiagramofthewetsandtestisshowninFig.9.4

. It can be

seen that the basic test concept is very similar to the dry sand test, namely using a

rubber wheel to drag abrasives across the face of a wear specimen. In the wet sand test,

however, the abrasive is in the form of a water slurry. The wear in both tests is deter-

mined by weight loss and converted to volume for material rankings. Both tests have

been used extensively to investigate scratching abrasi on and both have been found to

correlate well with many practical applications (1–3). While there is the potential for

differences, similar material behavior is generally found with both tests. The wet sand

Figure 9.3 Data sheet for the dry sand–rubber wheel test. (From Ref. 114, reprinted with

permission from ASTM.)

test or test conﬁguration offers an advantage in that the normal procedure can be

modiﬁed to utilize a slurry more representative of an application. In this way, chemical

effects associated with an application can be addressed.

As with the dry sand test, a standard version for material comparisons has been

developed by the ASTM (ASTM G105). Similar repeatability is found with both the

wet and dry sand tests when the procedures are followed. A coefﬁcient of variation of

5% is characteristic of the wet sand test. To achieve this, the test procedure controls many

of the same factors that are controlled in the dry sand test. These include design and

construction of apparatus, measuring techniques, specimen tolerances and preparation,

and rubber wheel and abrasive speciﬁcations. A procedure and interval for dressing the

rubber wheel are also given. However, some of the approaches used are different.

As with the dry sand test, the feeding and control of the abrasive is a signiﬁcant

factor in the test. Composition of the slurry is deﬁned in terms of abrasive size, type

and source, and the type and amount of water. A uniform supply to the interface is

achieved by having the contact below the surface of the slurry. To maintain uniformity

of the slurry during the test, paddles on the side of the wheel provide agitation

(Fig. 9.4). The test procedure also requ ires that the wear specimens be demagnetized prior

to the test. This is to avoid problems with magnetic wear debris adhering to the wear

surface or clustering in the slurry.

In the wet sand test, three rubber wheels of different duromet er (A50, A60, and

A70 durometer) are used to minimize the effect of wheel variation. This is done by using

the wear obtained with these different durometer wheels to deﬁne a best-ﬁt linear

relationship between durometer and log of the wear. The value of wear reported for

the test is that obtained for A60 durometer from that curve. Graphically, this is

illustratedinFig.9.5

Another difference between the two tests is that the wet sand procedure requires a

break-in cycle with the A50 durometer wheel. This is to minimize the effects of surface

Figure 9.4 Sketch of the wet sand–rubber wheel test.

defects and variations. After each of these tests, including the break-in, it is speciﬁed that

the slurry chamber be cleaned and new slurry used. The technique also requires reposition-

ing of the specimen, which is controlled by the ﬁxturing and associated tolerance. Visual

inspections and statistical criteria are associated with the procedure as well. For example,

tolerance on the correlation coefﬁcient for the linear ﬁt is speciﬁed. Also, if the coefﬁ-

cient of variation is beyond 7%, the test is considered out-of-control and the test procedures

and apparatus should be examined to determine the reason for this condition.

Like the dry sand test, the wet sand provides a ranking of materials but does not

provide absolute values. Different test conditions are speciﬁed for different classes of

materials to provide suitable resolution and separation in wear resistance.

9.2.3. Slurry Abrasivity

This test is also associated with low stress scratching abrasion. However, it was primarily

developed to address slurry abrasion problems in pipelines, which handle slurries. It has

been used extensively for these applications with good correlation (4,5). The initial focus

for this test was to rank the abrasivity of the slurries encountered, but the test method was

expanded to provide a ranking capability for wear resistance against a particular slurry.

For the former, the test results in what is termed a Miller Number; for the latter, a

SAR Number. SAR stands for slurry abrasion response. Both are based on a weight loss

technique for measurement of wear, although a wear rate is used for ranking. This test

is also an ASTM test procedure (ASTM G75).

The basic conﬁguration of the test is a ﬂat rectangular wear specimen sliding back

and forth across a rubber lap, ﬂooded by a slurry; both the wear specimen and the lap

aresubmergedintheslurry(Fig.9.6

). The ﬂat wear specimen is mounted on an arm, which

moves it back and forth at a controlled speed. The arm lifts the specimen at the end of each

Figure 9.5 Illustration of the method of analyzing wear data in the wet sand–rubber wheel test.

(From Ref. 3.)

stroke to allow the slurry to come between the specimen and the lap. The tray that holds

the slurry has tapered sides so that with the motion of the wear specimen, mixing and cir-

culation of the slurry occurs. The wear specimen is removed periodically and weight loss

determined. By varying the slurry and keeping the wear specimen constant, the abra-

sivity of slurries can be ranked. This is essentially the Miller Number. By keeping the

slurry constant and varying the wear specimen, wear resistance of different mate rials

to that slurry can be ranked. This is used to generate the SAR Number.

The reproducibility of the test is good, as the coefﬁcient of variation is approximately

6% for the Miller Number and 12% for the SAR Number. The larger value for the SAR

Number reﬂects the decrease in control that this version of the test has. A carefully

controlled reference wear specimen is used in the Miller Number test, whereas in the

SAR version the wear specimen is simply a material of interest. One element contri-

buting to this increased variation would be lack of precision in the value for speciﬁc gravity

of this specimen; other factors are variation and nonuniformity in other material

properties, such as composition and grain structure.

The test procedure associated with each of these tests involves a well-designed and

speciﬁed test apparatus, cleaning and specimen preparation steps, speciﬁcations and

controls on the materials involved, and carefully outlined steps for conducting the test.

The test method calls for replacement of the rubber lap with each test to avoid problems

associated with degradation and co ntamination of the lap. In particular, extensive details

are provided to insure proper alignment of the wear specimen to the lap, including a

method of checking alignment prior to the start of the test. Procedures for preparing

the slurry are also given.

The analysis method involves the developing of a wear curve and using a wear rate to

rank either the abrasivity of the slurry or the abrasion resistance of a material to a given

slurry. Measurements are made after 4, 8, 12, and 16 hr. Using a least square method, the

wear data obtained are ﬁtted to the following equation:

mass loss (mg) ¼ A  t

ð9:1Þ

Figure 9.6 Sketch of the slurry abrasivity test apparatus. 1, Molded plastic tray; 2, neoprene lap; 3,

tray clamp; 4, splash guard; 5, block lifting cam; 6, standard wear block—27% Cr-iron; 7, dead

weight; 8, adjustable plastic wear block holder; 9, pivoted reciprocating arm; 10, sand slurry; 11,

molded plastic ﬁller ‘‘V’’ channel; 12, tray plate.

where t is time. Using the value of A and B obtained in this manner, the wear rate (mg =hr)

atthe2hrpointisdetermined.ThismethodisillustratedinFig.9.7

. The wear rate is then

multiplied by a scaling factor of 18.18 hr=mg to provide the dimensionless Miller Number.

The higher this number, the more abrasive the slurry is. The scaling factor was determined

so that the Miller Number of a reference slurry of sulfur is 1 and of a reference slurry of

corundum is 1000.

The method of analys is is the same for the SAR Number. However, the wear rate at

the 2 hr point is multiplied by the scaling factor and the ratio of the speciﬁc gravity of the

reference material to that of the test specimen

SAR Number ¼ (wear rate at 2 hr)  18:18 ðh=mgÞ

7:58

ð9:2Þ

where p is the speciﬁc gravity of the test specimen. To reduce scatter, the method of

analysis uses the average value of two runs for the ﬁt. The test apparatus is usually built

with two separate trays so that two tests can be performed at the same time. In addition,

the test procedure calls for the samples to be rotated 180



between each measurement to

reduce effects of misalignment and orientation. Since separat ion of the slurry can occur

while these measurements are made, the procedure requires stirring of the slurry prior

to continuing the test.

TheformatforreportingtheresultsofthetestisshowninFig.9.8

, which shows that

more data and observations regarding the test are recorded than simply the mass loss of

the wear specimen(s). Included in this are the temperature and pH of the slurry, observa-

tions regarding the degree of wear seen on the lap, and a detailed identiﬁcation of the

slurry used in the test, including the preparation technique of the slurry. The recom-

mended slurry for the test is a 50% mixture of the solid extracts from the slurry in question

and water. However, the procedure allows the use of other types of slurries, such as the

slurry in question or the mixing of the dry extra cts with other media.

The test parameters and procedures were selected and developed with the use of

additional studies, aimed at insuring simulation with the intended application, minimum

scatter, and a convenient test (4–6). One focus of the studies was the inﬂuence of

Figure 9.7 Example of the method of analysis used in the slurry abrasivity test. The curve repre-

sents the least squares ﬁt of the data, using Eq. (9.1). Best ﬁt values of A and B are, respectively,

13.4262 and 0.749182, which result in a wear rate of 8.47 mg=hr at 2 hr and a Miller Number of

154. (From Ref. 6.)

concentrationandslurrymedia;anonlineareffectwasfound(Fig.9.9). The Miller

Number becomes less sensitive to concentration as the concentration increases. This

means that performing the test at higher concentration can minimize variation. The use

of higher concentration also enhances simulation in that many of the applications involve

high concentration (e.g., in the 50% range).

These same studies also showed effects associated with the liquid phase of the

slurry. For example, oil-based slurries tend to result in lower wear than water-based slur-

ries. Some of this behavior can be associated with the different ﬂuids providing different

lubrication and dispersion. In addition, this type of behavior can also be the result of

differences in chemical or corrosion effects that occur with different ﬂuids. Several things

were done to control this element in the test. One was to use a design, which employed

the use of nonconducting materials to eliminate electrochemical effects; another was to

use a chromium iron, which is used in some applications and is somewhat corrosion

Figure 9.8 Format for reporting wear data in the slurry abrasivity test. (From Ref. 6, reprinted

with permission from ASTM.)

resistant, as the standard material for the Miller Number . These effects also inﬂuenced

the selection of a water=dry solids mixture as the preferred slurry for the test. However,

because of the varied and uncontrolled ionic nature of the dried solids from various

applications, these do not necessarily eliminate corrosi on effects from tests. To help

separate and assess the signiﬁcance of corrosion on the test results, it is suggested that

a second test be done with a slurry speciﬁcally buffered to eliminate corrosion. This

concern is reﬂected in the pH data requested in the data sheet. It also underscores the

desirability to test with the actual slurry from an application to enhance simulation

and the use of the SAR Number test. Since the material of interest is tested in the

SAR Number test, the speciﬁc corrosion effects on that material are included in a test

with the actual slurry. It can also be seen that both the SAR and Miller Numbers must

be referenced to a parti cular slurry.

The use of several different test conditions suggested with the slurry abrasion test

procedure illustrates the need in some cases to modify standard wear tests to improve

correlation with an application.

The primary intention of both the dry and wet sand test and the slurry test is to rank

materials in terms of their abrasion resistance. Neither the wet nor the dry sand test is as

sensitive to the corrosion element as the slurry test. In the dry sand test, it is eliminated

since a ﬂuid is not used. In the wet sand test, corrosion effects are limited by the use of

a nonionic abrasive in water. From a test for pure abrasion resistance standpoint, the

dry and wet sand tests have some advantage in that the wear situation is less complex.

However, from an application standpoint, they may be poorer since the actual situation

may involve corrosion. Consequently, in performing evaluations for an application, which

involves ﬂuids, the slurry test provides the better simulation in that the confounding effects

of corrosion can be assessed. This may or may not be signiﬁcant, depending on the relative

degree of pure abrasion and corrosion, in any given situation.

Figure 9.9 The effect of slurry concentration on the Miller Number. (From Ref. 6.)

These three tests, dry and wet sand, and slurry abrasivity, have a common feature in

that they use a rubber lap as a counterface. In simulating low stress scratching abrasion,

the use of rubber as a counterface has an advantage in the way that it helps to reduce the

stress on the abrasive pa rticle and thereby to reduce fracture. Also, since they tend to

provide large contact area with the particles, they tend to preferentially hold the particles

anddragthemacrossthewearspecimen.Figure9.10

illustrates this for both a rubber and

rigid (e.g., steel), counterface. This action plus the use of a much larger surface area for the

rubber counterface results in preferential wear of the wear specimen. These two effects

account for the apparent superior abrasion resistance of the rubber laps in comparison

with the much harder materials evaluated in these tests. However, in a situation wher e

the abrasive would be dragged across the rubber surface, the rubber materials would show

a large reduction in wear resistance and inferior behavior to many of the mate rials

evaluated in the test.

The slurry abrasion test allows for different procedures for different material

categories. It does not provide absolute values of wear performance but rather a material

ranking.

9.2.4. Erosion by Solid Particle Impingement Using Gas Jets

Gas jets have been used to investigate solid particle erosion and to rank materials in

termsofresistancetothismodeofwear(Fig.9.11

). It has been found to correlate well

with erosion situations characterized by near normal particle impacts, such as erosion of

valves in coal gasiﬁcation and similar equipment (7,8), but not with situations which

involve grazing impacts. Examples of these latter situations would be erosion by

wind-blown dust and erosion of airfoil surfaces. Differences in the impingement condi-

tionsforthesecasesareillustratedinFig.9.12

. The effect of incident angle on wear scar

morphologyisshowninFig.9.13

Using the conditions and procedures for this type of test outlined in ASTM G76,

coefﬁcients of variation in the range 5–20% are achievable. The standar d conditions for

thetestareshowninTable9.1

,which lists the signiﬁcant parameters to be controlled

in the test and tolerances are speciﬁed for each. The test procedure also requires the

routine use of a reference material and the monitoring of the nozzle for signs of wear

(erosion). If the diameter of the nozzle increases by 10% or more, the nozzle is to be

replaced. Procedures for specimen preparation, cleaning, and repeatability are also

addressed in the standard.

Weight loss is the method used for determining the amount of wear that occurs.

However, the resi stance to erosion is measured in terms of the wear volume per gram

Figure 9.10 Behavior of abrasive particles trapped between two surfaces, which are sliding relative

to one another. ‘‘A’’, when one surface is an elastomer; ‘‘B’’, when both surfaces are metals.

of abrasive. This is obtained through the use of a wear curve that is generated by

measuring the mass loss at different time intervals . The slope of this curve is then used

to determine an average wear rate. Examples of wear curves obtained in this test are

showninFig.9.14

. The mass loss rate is converted to a volume loss rate by dividing

by the density of the specimen. This volume wear rate is then normalized to the abrasive

ﬂow rate to provide the erosion value, which is deﬁned as wear volume per gram of

abrasive. The smaller the erosion value, the more wear resistant the material. Guidelines

for the test duration are provided with the intention that the measurements be made in a

period of stable wear behavior. Since 2 min or less is typically required for stabilization,

it is speciﬁed that the ﬁrst measurement be taken after 2 min. The test should be carried

out for at least a total of 10 min but should not go beyond the point where the scar

depth exceeds 1 mm. The reason for this limit is that beyond that depth the shape

ofthescarbecomessigniﬁcantindeterminingimpactangle.Figure9.15

shows a typical

scar and the effect of wear geometry on angle of impac t.

The data and information that are to be recorded and reported in conjunction with

theerosionvalueareindicatedinFig.9.16

. When nonstandard test conditions are used,

Figure 9.11 Schematic of the solid particle erosion test.

Figure 9.12 Comparison of near-normal and grazing-normal particle impact.