Wear of copper alloy against steel in oxygen and argon

Wear of copper alloy against steel in oxygen and argon

Citation for published version (APA):

Zaat, J. H., & Gee, de, J. H. (1962). Wear of copper alloy against steel in oxygen and argon. Wear, 5, 257-274.

https://doi.org/10.1016/0043-1648(62)90129-1

DOI:

10.1016/0043-1648(62)90129-1

Document status and date:

Published: 01/01/1962

Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be

important differences between the submitted version and the official published version of record. People

interested in the research are advised to contact the author for the final version of the publication, or visit the

DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page

numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne Take down policy

If you believe that this document breaches copyright please contact us at: openaccess@tue.nl

providing details and we will investigate your claim.

WEAR

₂₅₃

WEAR OF COPPER ALLOYS AGAINST STEEL IN

OXYGEN AND ARGON

A.‘iV. J.IxzGEE hh'T)j, H. %%*T

Metal Rtwarch Institute T.N.O., Ddft (The Netherlands)

(Received for review November ‘~4~ 1g61; accepted January 21, x962)

VVear of 6 copper-zinc-lead alloys with zinc percentages of o-, I.T-, 3.5, 7.5-, 15.7-, and

rg.51/0 and a lead percentage of 2% against plain carbon steel has been studied as a function of sur-

face roughness in oxygen and argon. From the rate of wear, the volume of wear particles, the total amount of brass on the steel, and the appearance of the steel after termination of the experi- ment, it became obvious that two quite different mechanisms of adhesive wear can play a domi- nating part in the wear process, the mechanism by which the wear proceeds probably being con- trolled by the composition of the very thin oxide layers naturally present on the brass. The surface roughness of the steel appeared to have a very pronounced influence on wear when the formation of a semi-continuous layer of smeared brass on the steel was inhibited. This influence was studied extensively.

Finally, some information was gained about the influence of zinc content, composition of the atmosphere, and crystallite size on the strength of the brass-steel junctions.

Wear between two materials under rubbing conditions can proceed via many different mechanisms, for example, adhesive, abrasive or corrosive wear, or a combination of these. The mechanism of adhesive wear has been studied extensively, e.g. by BOWDEN AND TABOR~, KERRIDGE AND LANCASTER%, R~~~~owr~~3~ STEYNQ, HZIST AND

LANCASTERJ. Many investigators used the combination leaded brass 60140 (ix + ,3 alloy + 2% Pb) against steel, since this combination shows very pronounced ad- hesive wear.

The influence of varying chemical composition of the softer component, however, was investig,ated by ROACH, GOODZEIT AND HUNNICUTT~, who studied the influence of the “alloying tendency” on adhesive wear (see also DAVIES’), An adhesion- inhibiting action is ascribed to the oxide layers naturally present on the metal sur- faces8e9, and to other interlayers (e.g. smeared lead). It is generally held therefore that introduction of a neutral or reducing atmosphere greatly increases the contami- nation of the surface of the harder component by particles of the softer one. This is confirmed under certain experimental conditions*.

Throughout our experiments, we concentrated on discovering to what extent the composition of oxide films on the metal surfaces would affect the character of the adhesive wear process. Accordingly, we studied the influence of the composition of 6 or-C.u-En-Pb alloys during adhesive wear of these alloys against steel 60 in oxygen and in argon.

258

EQUIPMENT AND TESTING COXDITIONS

The experiments were performed on a pin-and-ring wear machine. In this machine, a cylindrical brass pin of 8 mm diameter is pressed against the cylinder wall of a rotating steel disk of 75 mm diameter (see Fig. I). During the experiments, the de-

crease in height of the pin was continuously measured and recorded electro-pneu- matically.

Load

Fig. 1. Principle of test apparatus.

It was possible to evaluate the rate of wear accurately from the curve giving the relationship between the height of the pin and time, since this curve in the range considered, viz. the steady state process period, is rectilinear. During testing, the specimens were kept in a plastic casing into which the desired gaseous medium was introduced. By maint~nil~g an excess pressure of about IOO mm Hz0 within this

cabinet, access of impurities was inhibited. The brass pins were turned to give a surface roughness of approx. 0.2 micron c.1.a. (cut-off value 0.03 in.).

The steel disks were ground in a circumferential direction to degrees of roughness from 0.1 to 0.5 micron c.l.a., as required. The experiments were performed in pure

oxygen or in argon, contaminated with 0.5% residual oxygen, at a load of I kg and

at a circumferential speed of 46 m/min.

MATERIAJS Chemical composition and structure

In view of the objectives of this research, six Cu-Zn-Pb alloys containing less than 20% zinc were chosen as wearing components against steel 60.

The compositions of these alloys are given in Table I. The materials were applied in as-cast condition, Material 2 was also used after heat treatment for 2 h in vacuum at

TABLE I CHEMICALCOMPOSITION OFBRASSALLOYS 3 95.0 3.2 1.8 4 90.8 7.5 I.7 2 82.4 _‘5.7 I.9 78.7 X9.5 1.8 Wear, 5 (196~~) r.j7-.z74

7ot I%. The respective structures of materials I, 2, 3 and 4 in as-cast condition

Chi

uacterized by the occurrence of a chiefly globular structure with relatively SI

CrJ

&allites. The lead, in undissolved condition, mainly occurs at the edges of

rstallites (see Fig.

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON

₂₅₉

are

nall

the

Fig. 2. Structure of alloy z in as-cast condition. (x IOO)

Fig. 3. Structure of alloy 5. (x IOO)

260 A. 1%'. J. DE GEE, J. H. ZAAT

Materials 5 and 6 display a pronounced dendritic structure with large crystallitcs.

In this case, lead is deposited predominantly along the original dendrite arms and is

therefore mainly within the crystallites (see Fig. 3).

Figure 4 shows the structure of material z after heat treatment. Compared with the

as-cast condition, the crystallites have grown considerably, so that the lead has hc-

come distributed in a mainly intra-crystalline manner.

Fig. 4. Structure of alloy L in heat-treated condition. (x IOO)

The structure of the annealed material corresponds almost entirely with that of materials 5 and 6 in the as-cast condition.

Steel 60 is a plain carbon-steel whose structure is represented in Fig. 14. Composition of oxide layers

An impression of the nature of the oxides occurring on the surface could be gained through oxidation of small cylinders of the various alloys in oxygen at temperatures

ranging from 200 to 700°C.

On oxidation, the alloys of low zinc content become covered with a black layer, which was identified by X-ray analysis as a mixture of cupric and cuprous oxide. The alloys of high zinc content are covered with a whitish yellow layer, which was identified as zinc oxide. It was established that the tendency to zinc oxide formation at rising temperature shifts to higher percentages of zinc in the alloys. The results of

these oxidation tests at 300°C are given in Plate I (see p. 266). The seventh sample

shown contains 30% Zn and 2% Pb.

Smearing of lead

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON 261

Since lead undoubtedly plays an important part in wear, it is of interest to investigate the extent to which smearing of lead between wearing surfaces of leaded brass can be expected, and also to what extent the composition of the brass influences this effect.

For the purpose, previously polished specimens of the several alloys were scratched, using a diamond needle with a radius of curvature of 300 p. In Plate 2 (p. 266) part

of such a scratch is represented. It is seen that smearing occurs locally.

By taking the total surface of smeared lead as a percentage of the scratch surface, a measure for the smearing effect is obtained; Table II gives a survey of results. It was found that no significant differences can be established. We therefore concluded that any differences in wear behaviour as a function of zinc content of the alloys used cannot be due to smearing of lead.

TABLE II

SMEARING OF LEAD ON SCRATCH SURFACES

I 2.3 3.5 2 2.5 3.7 2 (heat-treated) 3.9 3.9 ; 2.5 2.3 3.6 4.5 RESULTS

Ges-wal picture qf a wea experiment

Results obtained with regard to metal transfer, growth of transferred fragments and removal of these fragments as wear particles agreed well with data from the literature (e.g. ref. 2, 4 and 9).

After an inactive period, during which both metals are in contact with each other without causing any appreciable surface damage, a short incubation period occurs, during which the wear process is initiated by the first transfer of brass from the pin to the steel disc.

The beginning of the wear period is marked by the first loosening of transferred fragments from the surface of the disc. However, during the first part of this period, the rate of transfer of brass from the pin to the disc still exceeds the rate of removal of fragments, so that the total amount of brass on the steel increases. This is the non- steady state $eriod.

After some time, however, equilibrium is reached; then the rate of transfer equals the rate of removal, so that the rate of wear and the total amount of brass on the steel - measured at any arbitrary moment - become constant (ignoring minor statistical fluctuations). This is the steady state period.

The data obtained in the experiments have been collected in Tables III and IV.

Rate of wear

The rate of wear of the brass-expressed as decrease in weight per meter of sliding distance-differs for alloys with high or low zinc contents, if expressed as a function of s&ace roz&zaess of the steel disc, when the experiments are performed in oxygen (see Fig. 5). The alloys of low zinc content (alloys I, z and 3) show increasing wear

TABLE III RESULTS OF EXPERIMENTS PERFORMED IN OXYGEN Blcpf. Allop _- I I I 1 I 2 2 2 2 2% 2* 3 3 3 3 3 4 4 4 4 4 5 5 5 6 6 6 0.10 0.36 1.1 0.17 ““53 0.32 0.65 0.39 0.8$ Z-7 0.42 0.95 0.07 a.35 0.20 0.65 0.26 0.61 2.3 0.4” a,96 0.14 a.33 4-o 0.15 0.31 4.” 0.08 OS34 0.13 0.5.5 1.5 0.18 “47 0.20 0.47 0.41 0.73 3.7 0.10 Q”4.5 0.15 0.45 25.1 5.20 0.56 31.4 0.42 0.61 a.43 a~73 0.13 0.42 39.8 0.4” 0.39 0.43 “$45 0.10 0.32 0.36 0.45 0.12 a.34 40.0 N 2 * alby 2 in heat-treated condition x6.8 61.7 25.0 95.0 24.0 90.0 8.9 37.2 25.6 96.4 r36.0 387.” r49.0 430.0 167.5 471.0 L73.0 495.” 0.3 0.3 3-5 0.3 2.” 5.5 7.9 8.9 8.4 8.8 - 28 I 0.30 29 I 0.27 3” * 0.25 31 2 0.25 32 2 0.27 33 2 0.31 34 3 0.23 35 3 0.25 36 3 0.36 37 4 0.23 38 4 0.34 39 4 0.26 4” i 0.29 41 5 0.27 42 5 0.26 43 6 0.16 44 6 0.19 45 6 0.17 x.6 2.1 3.8 3.6 5-1 4.9 16.o 78.4 17.0 83.9 19.9 91.0 18.9 36.4 29.3 ‘44.” 26.2 Ijl.0 0.5 0.25 I.0 0.9 I.25 2.2 1.7

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON 263

with increasing surface roughness of the steel disc. An increase in roughness from 0.1 to 0.4 ,u c.1.a. causes a linear increase in the rate of wear from a mean value of 0.4 mgjm to a mean value of 0.85 mgjm. The zinc content of the alloys does nut influence this phenomenon,

Roughness of steel disc (micron c.l.a)

Fig. 5, Weight lass per meter sliding distance as a function of the roughness

oxygen. 0, allay I ; III. aiidy z ; .b I alloy 3 ; +, alloy 5 ; A, alloy 6.

of steel disc in

Fig. 6. Weight loss per meter sliding distance as a function of zinc content in

264 A. M’. J. DE GEE, J. H. ZAAl

By contrast, the rate of wear of the alloys of higher zinc content (alloys 5 and 6) is almost independent of the roughness of the steel disc. It is equal to the rate of wear

of alloys I, 2 and 3 at a surface roughness of 0.1 p c.1.a. Alloy 4 shows transitional

behaviour (see Table III, expts. 17~21). The rate of wear of alloy 2 in the heat-

treated condition appears to be 30 to 40’$,, lower than that of alloy 2 in the as-cast

condition at a comparable surface roughness of the disc (Expts. IO and II in Table III),

For reasons stated below, a separation of the wear rate ~~eystius roughness curve as

found for oxygen (Fig. 5) could not be expected for tests performed in argon. The

results obtained with experiments in argon when the surface roughness was kept

constant between 0.1 and 0.2 ,U c.1.a. are shown in Fig. 6. These results indicate

a decreasing rate of wear with increasing zinc content of the brass (probability of a decreasing slope about 990//o).

In argon, the rate of wear for all alloys is about 40’$; lower than the rate of wear in oxygen at a comparable surface roughness of the steel disc.

Volume of mear particles

The wear particles, originating by removal of transferred fragments of brass from the surface of the steel disc, are metallic (metallic lustre; no oxides are detectable by means of X-ray analysis). However, their brightness directly after an experiment de- pends very much on the atmosphere in which the experiment took place; it is much greater in argon than in oxygen. This indicates the formation of thin oxide layers in

oxygen. No iron-containing particles are found.

Independently of the composition of the alloys, or the atmosphere, the particles

generated have an analogous shape, which can roughly be characterized as flattened

elliptical (width/length ratio about 0.3 and thickness/length ratio about 0.05). But

when the experiments are performed in oxygen the volume of the particles depends to a great extent on the composition of the brass. Alloys with a low zinc content

(alloys I, 2 and 3) form small particles (mean volume about 10-3 mms), whereas

alloys with a higher zinc content (alloys 5 and 6) form relatively large particles

(mean volume about 15.10~:s mm3). Alloy 4 again shows transitional behaviour. In

contrast to the behaviour in oxygen, only small particles (about 2.10-3 mm3 volume)

are formed when the experiments are performed in argon.

To obtain a reasonably accurate measurement of the particle volume, the following procedure was adopted. 40 particles of greatly divergent length were chosen and their length I, width ~1 and thickness d measured by means of a microscope. Then the functional relationship between volume V and length 1 was estimated, assuming that the volume could be expressed as

I’ = Q ?l 1 w d (1)

It appeared that in all cases this functional relationship could be characterized as

v = $0 (2)

where y and 6 are constants.

In Fig. 7 the results, derived from Expt. 18 (see Table III), are given on logarithmic

coordinates. The most probable straight line through these points corresponds to

the formula: V = 0.014 12,ifl.

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON ₂₆₅ bly to be measured, and the range of length of the particles of this sample was estimat- ed. Finally, this range of length could be translated into range of volume by using the volume-length relationship, already estimated. The cumulative curve of volume range, derived from Expt. 18, is shown in Fig. 8.

In this way it has been possible to define a 50% volume VSO~~ such that 50% of the

Fig. 7.

a001

Ql 02 Q3 0.405 1 2 345 Length Cm m)

Particle volume as a function of particle length for

100. a80- L 2 9 yJ.--- -_-_-_ _______________ 20.

Fig. 8. Cumulative curve of distribution for the volume, derived from

Expt. 18.

Particle volume (mm31

Expt. 18.

266 4. W. J. DE GEE, J. H. ZAAT

‘lat

‘lat

?lal :e 3.

Plate

Plate

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON 267

5.

Plates 1-5

Plate I. Oxidized surfaces of alloys 1-6. Oxidation temperature 300°C (specimen 7 contains

30% zn).

Plate z. Surface of scratch made in alloy 2 in heat-treated condition, showing smeared lead. (X 480)

Plate 3. Surface of a steel disc, run against alloy 2 in oxygen, showing adhering fragment of brass.

(x 200)

Plate 4. Surface of a steel disc run against alloy 6 in oxygen, showing locally thickened layer of

brass. (x 300)

Plate 5. Surface of a steel disc, run against alloy 6 in argon. Some discolouration of the surface,

as a result of subsequent oxidation after the experiment, is visible ( x 200)

total volume occupied by all particles generated during a certain experiment is formed by particles with a volume less than or equal to VSO~~ (See Fig. 8).

In a similar way, the values of V~OS:, and V~O t; are defined. The results of these measurements are shown in Fig. 9* where the values of V~O:&, as a function of the

Fig. g. Mean volume of wear particles (I ‘507;) 5s a function of percentage of zinc. ( l alloy t in

heat-treated condition.) iooot

I, I

_j

1

i /

E

P

~

/

I]

RouQhness of steel disctmicrcn c.la)

Fig. IO. Volume of wear particles of alloys I, 2 and 3 in as-cast candition as a function of rough-

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON ₂₆₉

percentage of zinc, are given for a number of experiments. The transition in these

VSoe/O values at a zinc concentration of about 8% is very clearly revealed. The results

in argon indicate an increase in volume with increasing zinc content (probability of an increasing slope about 99%).

In Fig. IO, the values of I’i~qg, V~O% and I’~o% are given as a function of surface

roughness of the steel disc for the alloys of low zinc content (alloys I, 2 and 3). The

volume of the particles appears to increase almost linearly with increasing surface roughness, this being an explanation for the scatter in the results given in Fig. 9 for

alloys I, 2 and 3.

On comparing Expts. IO and II with Expts. I and 13 in Table III, the particles

generated during the wear of alloy 2 in the heat-treated condition appear to be about

150% larger than those formed during wear of the low zinc content alloys in the as-

cast condition, the surface roughness of the steel being the same in both cases.

Finally, it appears that introduction of argon instead of oxygen during wear of

alloys I, 2 and 3 causes the particle volume to increase (cf Expt. 33 in Table IV with

Expts. I and 13 in Table III).

Total amount of transferred brass on steel

The total amount of transferred brass present on the surface of the steel discs

after termination of the experiments was determined for a number of experiments by chemical analysis. The brass was dissolved in an aqueous solution of potassium

0 10 20

% Zinc

Fig. II. Total amount of brass on the steel as a function of percentage of zinc. (e, alloy 2 in

heat-treated condition.)

dichromate, after which the copper was estimated spectrophotometrically and the

zinc polarographically. The results are given in Tables III and IV and in Fig. II

as a function of the zinc percentage.

It appears that wear of the alloys of low zinc content (I, 2 and 3) in oxygen is

accompanied by the deposit of only a slight amount of brass on the steel. Alloys 5 and Wea+‘, 5 (1962) 257~274

27o A. W. J. DE GEE, J. H. ZAAT

6, on the other hand, cause the deposit of much brass on the steel. Once more, alloy 4 shows transitional behaviour.

The amount af brass transferred during wear of alloy 2 in the heat-treated condi- tion is about IO times larger than in the as-cast condition.

fn argon, there is no sudden transition in the amount of brass transferred at about So/l zinc. The results indicate an increase in the amount of brass transferred with increasing zinc content (probability of an increasing slope exceeding ogy/,). On comparing Expt. 35 with Expts. 2, 8 and 16, it appears that the introduction of argon instead of oxygen during wear of alloys S, z and 3 causes an increase in the amount of transferred brass.

~~~~~~~0~

of

of

pias

The wear of the brass pins is accompanied by the formation of grooves in the sur- face. This causes the surface zone to be severely work-hardened. Fig. 12, a longitu-

Fig. 12. Longitudinal section of a pin of allay 5, after an experiment, showing part of work-

hardened zone. (x qao)

dinal cross-section through the pin of alloy 3 after terminatian of Expt. 23, gives an impression of this work hardening. To obtain an impression of the depth to which the work-hardening proceeds in the material, the hardness (n~~cro~~ckers, load 25 g) was measured as a function of the depth under the surface, using a prepared longitu- dinal cross-section of the material.

The results of these measurements, performed on a specimen of alloy 2, are shown in Fig. 13‘ ft appears that the work hardening proceeds to a depth of about sc~f

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON 271

microns under the surface. Extrapolation of these curves to zero depth yielded values for the surface hardness under wear conditions. These values are listed in Table V, together with the values of the hardness in bulk.

~~~

Depth under surface (micron)

Fig. 13. Hardness (microvickers, 25 g) as a function of depth under the surface for alloy z in as-

cast condition after an experiment.

TABLE V

HARDNESS IN BULK AND ON THE SURFACE

Matnial Hardness in bulk Hardness 011 the

fkslmm’) srrface(k&wn*) I 60 150 2 60 150 2* 65 I45 3 65 160 4 85 175 5 100 250 6 II0 270 * heat-treated

Condition. of su7jace of the steeE discs during the stea$y-state period

As has already been stated with regard to the experiments performed in oxygen, many of the measured quantities show abrupt transitions when a zinc percentage of 8% is exceeded. This same discontinuity appears to determine the appearance of the steel discs after termination of the experiments.

The wear of the alloys of low zinc content (alloys I, 2 and 3) is initiated by the for-

mation of localized adhering fragments, directly bonded to the steel. In Plate 3 (p. 266) such a transferred fragment of alloy 2 is shown. Each fragment is partly pressed

into the steel surface, thus causing an indentation in the surface, accompanied 1~~. tlrc formation of “walls” protruding above the surface of the steel. These indentations

must originate from plastic deformation, since no iron-containing particles are found.

After removal of transferred brass as a wear particle, the indentation more or

less filled with brass residue - remains. The piled up parts of such a damaged site act as nuclei for repeated brass transfer; this causes some of these transfer patches to grow in the course of the process. Consequently, during the steady state period there is a fixed and limited amount of relatively large active transfer patches still able to

participate in metal transfer (order of magnitude: 5 per cm”). Figure 14 shows a

cross-section through the steel at an indentation, on the edge of which some brass

has accumulated.

Apart from the localized damage the surface of the steel remains quite unattacked; this is proved by the undisturbed appearance of the original grinding grooves in the steel, In contrast to this behaviour, the wear of alloys of higher zinc content (alloys

5 and 6) is initiated by the formation of a thin semi-continuous Iayer of transferred

brass. This covers a large part of the steel surface, including all protruding parts and most of the grooves. During the steady state period this layer thickens locally to large adhering fragments, which are removed as wear particles after some time. No indentation of the disc surface occurs. Plate 4 (p. 267) shows a magnified detail of a steel disc run against alloy 6 in oxygen. Alloy 4 again shows transitional behaviour.

In argon, wear of all alloys investigated proceeds via the mechanism of localized

transfer. Plate 5 (p. 267) shows magnified detail of the surface of a steel disc run against alloy 6 in argon.

Fig. 14. Section of a steel disc run against a specimen of alloy 2. (

x

WEAR OF COPPER ALLOYS IN OXYGEN AND ARGON ₂₇₃ DISCUSSION OF RESULTS

Imfluence of compositiolz of oxide layers 012 wear process

Close examination of the results already discussed indicates the separate action of two different mechanisms of adhesive wear.

Mechanism I (viz. the “mechanism of local attack”) governs the wear of all the investigated alloys in argon and of the low zinc content alloys (alloys 1, z and 3) in oxygen, It is characterized by the following features.

(I) Wear is initiated by the formation of highly localized adhering fragments, causing plastic deformation, which results in indentations in the steel.

(2) During the steady-state period, a limited number of transfer patches on the, otherwise undamaged, steel surface are active sites of metal transfer.

(3) The amount of brass on the steel surface is relatively low. (4) The wear particles have a relatively small volume.

(5) The surface roughness of the steel disc exerts a pronounced influence 04 the volume of the wear particles and the rate of wear, both increasing with increasing roughness.

Mechanism II (viz. the “mechanism of semi-continuous layer fo~ation”) governs the wear of the alloys of higher zinc content (5 and 6) in oxygen. It is characterized as follows.

(I) Wear is initiated by the formation of a semi-colztinuozcs layer of transferred brass, which covers a large part of the surface of the steel.

(z) During the steady-state period, wear takes place by the formation, and subse- quent removal, of transferred particles of brass owing to local growth of the brass layer in a direction perpendicular to the steel surface. No indentations in the steel occur.

(3) The amount of brass on the steel is relatively high. (4) The wear particles have a relatively large volume.

(5) The surface roughness of the steel disc exerts no detectable influence on the wear process.

The available information indicates that the essence of the second mechanism is the formation of a thin semi-continuous layer of brass.

As already stated, alloys 5 and 6 form zinc oxide, if oxidized in air, whereas alloys I, 2 and 3 form copper oxides. This, in addition to the fact that in argon the formation of a smeared layer of brass is inhibited, makes it probable that the presence and contin- uous formation of a very thin oxide layer of suitable composition (in this case zinc oxide, or zinc oxide with minor quantities of lead or lead oxide) is essential for the formation of this layer of brass, although the mechanism by which the oxide layer exerts its influence is not clearly understood; it is still being studied.

~~f~~e~&e of surface ~o~~h~ess of steel

When mechanism II governs the wear process, the initial surface roughness of the disc exerts no influence whatever, since the thin layer of brass to a large extent masks the original roughness pattern (see Plate 4, p. 267).

When mechanism I predominates, increasing surface roughness of the steel is Wear, 5 (1962) 257-274

274 A. W. J. L)E GEE, J. H. %AA’l

accompanied by an increasing particle volume and an increasing rate of wear, but the total amount of brass on the steel remains constant within the analytical limits (cf. Expts. 2, 8 and 16 in Table III).

The increase in jmrticle volume is due to the fact 1 that with increasing surfact,

roughness of the steel, the number of areas where brass and steel touch decreases, while the extent of these areas increases. This in turn leads to a decreasing number of enlarging transfer patches, which implies larger particles.

The increase in the rate ofwear with increasing surface roughness might be explained by assuming that the increase in the average rate of growth of the individual adhering fragments is more pronounced than the drop in the average number of transfer

patches actually participating in metal transfer. This is probably due to an appre-

ciable shortening of the average time for which the transferred fragments adhere to

the disc. Close examination by means of high-speed photography of the events

occurring on the surface of the steel disc is now in progress.

Strength of brass -steel junctions

No explicit information about the absolute value of the strength of the brass-steel

junction has been gained. However, some information about the influence of some other variables, at constant roughness, on the adhesive force between the brass and the steel was obtained for the experiments in which the wear process was governed by the mechanism of local attack. As was reported in RESULTS, the particle volume and the total amount of brass transferred to the steel increase very markedly, whereas the rate of wear decreases as a result of the following systematic variations;

(I) increase of the zinc content for the experiments performed in argon,

(2) introduction of argon instead of oxygen for the experiments performed with

alloys I, 2 and 3,

(3) heat treatment of alloy 2.

The resulting observations would seem to imply a considerable increase in the

average time for which the transferred fragments adhere to the surface of the disc, as a result of each of these variations. Since the roughness is kept constant, increasing adherence time means increasing strength of the junctions between the brass and the steel. The influence of the zinc content indicates that the zinc atoms play a

dominating role in causing the brass-steel junction. The influence of the atmosphere

reveals an adhesion-inhibiting action of the copper oxides present on the surfaces of

alloys I, 2 and 3 in oxygen. The influence of the heat treatment of alloy z is not clear.

It seems not unlikely, however, that the size of the crystallites affects the adhesion.

REFEREXCES

1 F. I’. BOWDEN AND D. TABOR, The Friction and Lubrication

of

IgjO.

2 ICI. KERRIDGE AND J. K. LANCASTER, Pvoc. Roy. Sot. (Loxdon), A 236 (I9j6) 250.

3 E. RABINOWICZ, Wear, z (1958) 4.

4 R. E'. STEIJN, J. Rasic Eng., (Trans. ASME, Ser. D), 81 (1959) 56.

5 \I;. HIRST AP;D J, K.LANCASTER, Pvor. Roy.Soc.(London),A 2.59 (1960)228. 8 X.E.ROACH,C.L.GOODZEIT AND R. P.HUN~ICUTT,T~U~S. ASME, 78 (1956) 1659.

7 R. DAVIES, in Handbook of Mechanical Weav, edited by C. TAPSON AND I_. 1’. COLWELL, The

University of Michigan Press, 1961, p. 7.

* M. B. PETERSON AND J. J. FLOREK, in Handbook of Mechalzical Wear, edited by C. LIPSON AND

I,. V. COLLWELL, The University of Michigan Press, 1961, p. 16.