Friction and wear of some bronzes under lubricated reciprocating sliding

Friction and wear of some bronzes under lubricated reciprocating sliding

Wear - Elsevier Sequoia S.A., Lausanne - Printed in the NetherIands FRICTION AND WEAR OF SOME BRONZES RECIPROCATING SLIDING UNDER 373 LUBRICATED ...

2MB Sizes 0 Downloads 50 Views

Wear -

Elsevier Sequoia S.A., Lausanne - Printed in the NetherIands

FRICTION AND WEAR OF SOME BRONZES RECIPROCATING SLIDING

UNDER

373

LUBRICATED

R. S. MONTGOMERY* Mugs

Research Center, Waferdiet

Amenal,

Waterdiet,

N.Y.

rzr89

(U.S.A.)

(Received March 31, 1970)

SUMMARY

Tribological behaviors of some commercial alloys were investigated in order to obtain data for the selection of bronzes for seals and bearings in reciprocating machinery. Both stationary bronze blocks and moving bronze rings were mated with hard surfaces using a Dow-Corning LFW-I friction and wear testing machine. The results were strongly influenced by operating conditions; sliding geometry and specific mating surfaces had major effects and small additions of fatty acid to the lubricant greatly reduced wear rates and coefficients of friction. Depending on test geometry, hardness was an unreliable guide to wear resistance or none at all, Furthermore, 20% Pb showed an advantage over 10% with one sliding geometry but not with the other and the superiority of the more coarse lead dispersions reported by previous workers was found with one test geometry but the opposite effect was found with the other. INTRODUCTION

Leaded tin bronzes have been important bearing materials for a great many years, but, surprisingly, there have been few data published on their friction and wear behaviors. LUNN~ has published a review of wear research on these alloys discussing the relatively few known facts and indicating those areas of greatest uncertainty. The most comprehensive experimental study is that recently published by DE GEE et nt.2. These researchers investigated the wear of a series of alloys with 6% Sn, and o, 5, IO, 15, 20, 25% Pb and another series with 15% Pb and o, 3, 6, 9, 12 and 15% Sn, lubricated as well as unlubricated. The work has the advantage that it was carried out on specially produced alloys thus allowing study of single variables and obviating difficulties arising whea the more complex commercial alloys are used. The paucity of systematic laboratory studies which would allow prediction of performance of leaded tin bronzes is probably largely a result of the large number of commercia1 alloys and their heterogeneous and complex structures. A factor which limits the usefulness of what data there is in the literature is that the tribological * Research carried out at the Ingersoll-Rand

Research Center, Princeton,

N. J. (U.S.A.).

Wew 15 (I974 373-387

behavior of a bearing bronze is critically influenced by the particular operating conditions, As pointed out by F~rs3, with these alloys friction and wear are influenced in a complex manner by run duration, load, temperature, lubricant, atmosphere and the specific mating materials. The present research was undertaken in order to obtain data for use in the selection of bearing bronzes for seals and bearings in reciprocating machinery. Wear rates and frictional behaviors of a number of different kinds of commercial alloys were studied under conditions of lubricated reciprocating sliding at a relatively light load. Furthermore, influence of the surface mated with the bronze and effect of addition of fatty acids and esters to the lubricant was also investigated. Use of commercial alloys in this study caused difficulties of interpretation of expel-inl~~tal data which had been avoided by DE GEE et al. bum allowed study of alloys of specific interest. BRONZES INVESTIGATEU

Compositions and characteristics of the leaded tin bronzes are given in Table I. These commercial alloys are representative of a number of kinds of products. Bronzes A, B, and C were typical sand-cast leaded bearing bronzes; bronze B contained IO”/~ Pb but both the others contained zo”/b(nominal). Bronzes D and E were continuously cast, 1o”/0 Pb bearing bronzes which are used where good mechanical properties are required. The continuously cast alloys were more uniform and the lead grains were appreciably smaller than were the lead grains of the sand-cast products. Bronze I;

TABLB

I

LEADED

TIN

__~

BRONZES

~.___

~

INVESTIGATED -.._-

leaded bronze

_---

-..-.

-.

Brine11

Bvo%?e

Lead grain size .- - ..-.

Chemical composition (“/;,,I

46.5

Sand-cast

Xedium

74.03 Cu; 0.33 sn; 18.73 Pb; 0.10 Zn; 0.42 Ni; 0.045 I’. 79.98 Cu; 8.73 Sn; 10.24 Pb; 0.56 Zn. 71.28 Cu: 9,32 Sn; 18.32 Pb; 0.45 Zn. 78.39 Cu; 6.54 su; 11.35 Pb; 3.55 211; o.ooo8 P. 79.38 Cu; 9.70 Sn; 9.88 Pb; 0.56 %n; 0.42 Ni; 0.020 P. 86.79 Cu; 7.59 Sn; 1.28 Pb; 3.12 Ni. 76.44 Cu; 3.08 Sn; 10.79 Pb; 7.22 Zn; 1.94 Ni; 1.24 I’. 83.02 cu; 10.20 sn; 3.81 Pb; 2.74 Ni; 0.78 P. ~~__ .

A

~~-0-20

I3

80-rn-IO

leaded

bronze

47.5

Sand-cast

Medium

C

70.-10-20

leaded bronze

ii3.2

Sand-cast

Medium

D

80-6-10-4 containing

71.3

Continuous

Fine

B

80-10-10

91.7

Continuous

Fine

1;

88-8-1-3 leaded bronze containing 30/u Xi Patented, continuously cast zinc-containing leaded bronze 83-10-4-3 leaded bronze containing 3% Ni

98.0

Sand-cast

Coarse

105.4

Continuous

Fine

x12.8

Continuous

Fine

C H

leaded bronze 4o/o Zn leaded bronze

* Converted from liockwell JYf%r, 1.5 tr97oj 373-J87

..- -_..._

Casting method _ -.-i-_..-

FRICTION AND WEAR OF SOME BRONZES

375

was a sand-cast and bronze H a continuously cast nickel-containi~ alloy; these bronzes often have good wear resistance in certain applications. Bronze G was a new, patented alloy which can be produced only by continuous casting. The hardnesses of these bronzes ranged from 47 for A and I3 to 98, 105, and 113 Brine11 for F, G, and H, respectively. A single aluminum bronze was investigated. While not a bearing alloy, it was included in the study because aluminum bronzes have been reputed to be goodwearingl. Results obtained with this alloy will be treated separately owing to its essentially different character. EXPERIMENTAL

DETAILS

The experiments were made using a Dow-Corning LFW-I variabIe speed friction and wear testing machine. This machine is of the “Timken” type in that it employs a block sliding on a rotating ring. It was operated in the oscillating mode which corresponds with the action of reciprocating machinery. This mode should result in more rapid wear owing to loss of hydrodynamic lubrication at the end of each stroke. Speed of reciprocation was 1.4 c/see and a go” arc was used which produced a track length on the ring of 2.8 cm. A load of 27 kg was used and the lubricant was maintained at IIO’C during the experiments which was the highest control temperature feasible. The experiments were made using a purified white mineral oil the viscosity of which had been increased to 260 cSt. at IOO’F by addition of polyisobutylene. This is a lubricant commonly employed in reciprocating compressors. Pol~sobutylene is unreactive and has no significant chemical reaction with metal surfaces. In a few experiments, a small amount of reactive compound was added to the lubricant. These reactive additives were either stearic or oleic fatty acids, or the ester, methyl oleate, which would be hydrolyzed, at least partially, to oleic acid at the conditions at the sliding surfaces. Fatty acids are known to reduce friction and wear of bronzes. The coefficient of friction was obtained from the maximum instantaneous friction force during a reversal measured with a load cell. Wear was obtained by estimating the volume of bronze lost during the experiment. When the block was the bronze component, the topography of the block surface was measured near both sides and at the center in the direction of sliding with a Bendix stylus-type profilometer and wear volume estimated from cross-sectional areas lost at these locations. An effort was made to obtain similar depths of wear on the blocks so that bearing-load history would be the same for a11 bronzes tested; test duration ranged between 400 and IOOO cycles. Where the ring was the bronze component, its surface was measured near both ends and at the center of the track in a direction perpendicular to the direction of sliding and wear volume was estimated from cross-sectional areas lost at these locations. An effort was made to keep wear depths below 12 ,um by adjusting the duration of the experiments and in only a few cases did it exceed this value. The experiments ranged in length all the way from 2,000 to 20,000 cycles. With bronze rings there was negligible wear on the hard blocks so that bearing area, and consequently bearing pressure, remained constant throughout the experiment. The method of estimating wear volumes from profilometer measurements Wear, 15 (1970) 373-387

was sensitive and allowed reliable mcasurelllents at extremely small values, Because of the small amount of wear, geometry of the contact changed little during the experiments. Two to four individual determinations were made with each bronze ring and three to five individual determinations with each bronze block depending on scatter of the experimental values. Values in the tables are averages. IiESULTS

Brome blocks Friction and wear characteristics of the bronzes were investigated for bronze blocks sliding on carhurized steel rings. These experinlents were similar to some of those of DE GEE et al. who used &mm diam. bronze pins sliding on 1045 steel rings. However, these researchers employed a unidirectional rotation of the rings where the present results, given in Table II, were obtained using a reciprocating motion.

The lowest wear rate was obtained with the soft sand-cast 74-6-20 bronze (A) which was followed by that with the patented continuously cast bronze (G). The most rapid wearing was the 476 zinc-containing continuously cast 80-6-10-4 bronze (D); it wore over two and half times as rapidly as the SO-IO-IO continuously cast bronze (E). It is difficult to account for this extremely rapid wear. The data of 1;KBNCH AND STAPLES4 indicates that there is no significant effect of zinc up to 4::,. Perhaps the much better wearing properties of bronze E can be attributed to its o.42‘3h nickel and its higher phosphorus content. The coefficient of friction ranged from 0.12 for the patented continuously cast bronze (G) to as much as 0.2s for the sand-cast nickel bronze (1’). Roth nickel bronzes produced relatively high friction probably because of their low lead contents.

Friction and wear characteristics of the bronzes were also investigated using the opposite sliding geometry, that is, with hard blocks sliding on bronze rings. With this test geometry too, reciprocating sliding was used. In these experiments surface temperature of the bronzes doubtless remained lower because there was only momentary contact of a particular location with the mating surface rather than continuous contact as was the case with the bronze blocks described above. Furthermore, transfer of bronze to the mating surface was more important owing to the small amount of hard surface actually in contact during the experiment. With this test geometry, bearing pressure remained unchanged throughout the experiment since there was negligibie wear on the hard blocks.

FRICTION

TABLE

AND WEAR

377

OF SOME BRONZES

III

MATING

SURFACES

INVESTIGATED

48-53Rc

Ground; < 0.02 m,u peak to valley.

Hard chrome pIated

Gz-63R,*

Polished but covered with fissures; 0.1-0.5 rn,u,peak to valley.

Flame-plated tungsten carbide, O.O~O.I mm thick; Linde LWB (9% Co)

72x,*

Polished but covered with fissures; 0.1-0.2 rnp peak to valley.

4620 steel

Carburized;

59-64Rc

Nitraltoy I35 (modified) steel

Nitrided;

Ground valley. Ground valley.

4140

steel

Induction

41+3

steel

4140

St&

hardened

0.6-0.7

mm case

0.5 mm case

7oR,*

Tungsten carbide cermet ; Kennametal Kg4 (I 1 “/o Co)

; < 0.02 rnp peak to ; < 0.02 rnp peak to

Polished

9oR,

* Converted from DPH

I

0 0

1.000

I 2,000

I 3,000

I 4.000

t 3,000

I

6,OW

I

7,000

I 6.000

1

6,000

CYCLES

Fig.

I.

Erratic coefficient

of friction of bronze A as a function of sliding cycles.

With the bronze rings, the influence of the particular hard mating surface was investigated. Cemented tungsten carbide, both as a solid block and as a flameplated coating, and chromium-plated steel as well as three different hardened steels were used. These surfaces are described in Table III. The nature of the mating surface was commonly found to have a major influence on the friction and wear of the bronzes. Friction and wear data for the experiments with this test geometry are tabulated in Table IV. Wear, 15 (1970) 373-387

378

F

FRICTION AND WEAR

OF SOME BRONZES

379

The lowest wear rate was obtained with the patented continuously cast bronze (G) and the next lowest with the continuously cast nickel bronze (H). Both these ailoys wore considerably less than any of the others including the sand-cast nickel bronze (F). The two softest bronzes, A and B, wore rapidly. Judging from the erratic friction behavior (illustrated for bronze A in Fig. I.) and other indications, the rapid wear could be attributed to excessive bronze transfer to the mating surfaces. In general, the lowest coefficients of friction were obtained with the patented continuously cast bronze (G) ; they ranged from 0.09 to 0.14 depending on the mating surface. The highest values were obtained with the nickel bronzes which frequentIy gave coefficients of from 0.19 to 0.20. Influence

of the mating surface

Different hard mating surfaces produced different bronze wear rates and coefficients of friction. These effects were often different for the various bronzes. (I) The results obtained with all three steels investigated were generally similar. (2) The hard-chrome surface was tested on two of the bronzes. It resulted in more rapid wear than did steel on one of them (B) but similar wear on the other (C). Rapid wear of the relatively soft bronze B may be, at least partially, attributable to the fissured character of the chrome plate. Transfer of metal by alloy B appeared to be excessive with all the mating surfaces and this transfer might be expected to be increased by a fissured surface topography. (3) The flame-plated tungsten carbide coating was tested with four of the bronzes (A, B, C, and H) and produced more rapid wear than did steel on both the softest (A) and the hardest (II). Again, this may be, at least partially, attributable to surface fissures. (4) Wear obtained with solid tungsten carbide cermet was less than that with steel on the two softest bronzes (A and B) but similar with the others. Since the rapid wear of these soft alloys seems to be associated with excessive bronze transfer, adhesion to the cobalt surface of the cermet may be less than to steel surfaces.

The particular aluminum bronze studied was a sand-cast material with a hardness of 186 Brinell. It contained 85.6674 Cu, 3.55% Zn, ro.2% Al, and 3.620/, Fe. The aluminum content was in the center of the preferred range referred to by LUNN (8812%). It was tested both as a block against a steel ring and as a ring against hard blocks. The data is tabulated in Table V. Wear rates were low but not as low as might be expected from its extreme hardness. Almost all experiments showed erratic friction. Mating surfaces became progressively more rough throughout the experiment and a bronze color could be seen, indicating severe bronze transfer. It was concluded that aluminum bronzes are not suitable for bearing materials; they are excessively hard and therefore may cause injury to the mating surfaces and, moreover, this alloy at least, rapidly transfers to the mating surfaces resulting in progressively increasing roughness. This would doubtless eventually result in catastrophic failure. Influence

of fatty acids

Bronzes

B, C, and E were tested

with additions

of small amounts Wear2 15 (1974

of either 373-387

TABLE

V

(Wear rate in mm~/cycie

X 10s)

------._.___ Bronze

“_

t,ing matfd

Bronze block mated z&h hard rilzg __-.___.--.-.-. Wear vaate f

with havd block ___~-.I Wear vale f Mated with inductionhardened 4x40 steel

5

0.24

9

0:22-u.2‘+*

Mated with carbide-coated steel

8

0.23~“0.27’

Mated with cemented tungsten carbide

6

o.rg-0.24*

._.

Mated with carburized &to steel Mated with nitridcd Sitralloy 135 (Mod.) steel Mated with [email protected]

steel

-* Friction

erratic

INFLUENCE

OF

MATED

WITX

ACID

FATTY

CARBUKIZED

CDNCEXTRATIOX

ON

FXTCTION

--

- ..._--

(Wear rate in mmfkycle

x Io5) Brome ring mated zvith steel block

__

-.

0% stearic acid

Sand-cast

B

bronze

Continuously bronze E

WEAR

hED

BEHAVIOR

OF

BRONZES

STEEL

0.21;

TVelw rate __.__

f

61

0.18-0.24”

20

a.18

cast -

. ,...-_

09/o stearic

0.2”;

acid --_--IYear rafe

acid

~--stpavic

acid -l_-~ wmr rate

f

16

0.14

I6

0.19

5

o.r_i

7

0.14

31

O.ZO

8

0.18

_. ~._--_-

._

stearic

___.~ Bvcvaze block mated with steel ring __-..-..

.--- _.~_.__.

f

..-__--

WPnr r& vatr

__l_______--_..

.---.-l--

f

-

~--.---

“.~

* Friction erratic

INFLUENCE BRONZE

OF

C BLOCK

FATTY

ACID

MATBD

WITH

(Wear rate in mm3/cycle Com7entration of vleic acid (%)

CONCENTRATION CARBURIZED

ON STEEL

x Io6) 0

0.25

0.50

FRICTION RING

AND

WEAR

BEH.AVIOR

OF

SAND-CAST

FRICTION AND WEAR

OF SOME BRONZES

BRONZE STEEL

llL4CK

BRONZE CARBIDE O‘EIC BRONZE

ON CARG”RIZE4

RlNG - OLEIC RING

CERMET

CERMET AC,D

METWL

-

RfNG ON TUNGSTEN

STEAR1C

CARSIOE

GLOCK

ACID

C*,RSIOE

BRONZE

ACID

ON TUNGSTEN

RiNO

SLOW

-

ON TUNGSTEN

CERMET

@LOCK

-

OLEATE

3.0

Fig. L. Influence

of lubricant

additive

concentration

on wear of bronze C.

stearic or oleic acids or of the ester, methyl oleate, to the lubricaat. (Methyl oleate tends to hydrolyze at the conditions present at the sliding surfaces with production of oleic acid.) The influence of stearic acid on friction and wear behavior of the sandcast bronze B and on the continuously cast bronze E when mated with carburized steel is shown in Table VI; the influence of oleic acid on the behavior of blocks of the sand-cast bronze C when mated with carburized steel is shown in Table VII; and the influence of both of the fatty acids and of the ester on the friction and wear behavior of the sand-cast bronze C rings when mated with tungsten carbide cermet blocks is shown in Table VIII. The wear rates from these latter two tables are plotted in Fig. z to illustrate the influence of concentration. The most effective additive tested was stearic acid. When bronzes B and E were mated with carburized steel, a concentration of 0.2% reduced wear rates by 65-75 y. and frictions by 8-25 “/o. However, when bronze C was mated with tungsten carbide cermet, there was less effect on wear rate; it was only reduced by 33% by a concentration of 0.25 %. Infhcence of bronze hardness It would be expected that the harder alloys would have the better wear resistances. This was, indeed, generally found to be the case when the bronzes were tested in the form of rings sliding on a stationary hard block, but there were a number of exceptions. Bronze A generally had considerably better wear resistance Wear, 15 (1970) 373-387

MATING

01

0

I

I.*00

I 2,000

SURFACE

I

I

I

I

I

I

3,000

4,000

5,000

6,000

7,000

6,000

I

9,000

CYCLES

Fig. 3. Coefficient

of friction

of bronze E as a function

of sliding cycles

than did bronze B although it was slightly (one Brine11 point) softer and bronze C had considerably better wear resistance than would be expected on a basis of its hardness. Furthermore, bronze G had better wear resistance than would be expected on a basis of its hardness. Furthermore, bronze G had better wear resistance than did bronze I3 although it was seven Brine11 points softer. On the other hand, when the bronzes were tested as blocks sliding on a moving steel riug there was no effect of hardness apparent. In this case, the best wearing bronze was the softest (A) and the second best wearing was the second hardest (G). With bronze rings, the bronze was sliding entirely on a film of transferred metal owing to the small area of the hard surface actually in contact. With the other test geometry, transferred films would not be expected to coat the mating steel surfaces so thoroughly. With the bronze specimens in the form of rings, the two softest alloys, A and B, showed erratic friction behaviors (illustrated for A in Fig. I). These erratic behaviors appeared to be caused by excessive transfer of bronze to the mating surfaces and Weas, 15 (1970) 373-387

FRICTION AND WEAR OF SOME BRONZES

383

Fig. 4. Section through bronze C block after run-in showing surface coating of lead. Surface was protected by nickel electroplate before sectioning. x 500

relatively

infrequent sluffing off of a portion of the transferred metal probably by a mechanism analogous to the “prow” formation described by ANTLERS and others. The test geometry would correspond to a hard slider on a soft flat with ANTLER’s pin-on-disk machine where he found prow formation to occur with dissimilar metals provided that the unworn fiat is not excessively harder than the unworn rider. With the harder bronzes, friction generally showed an initial rise with a subsequent drop to the steady-state value (Fig. 3) and erratic behavior was not obtained.

The complex commercial alloys used in this study did not allow a real evaluation of the effect of tin. However, the fact that the 74-6-20 bronze (A) had significantly better wear resistance than the 70-10-20 bronze (C) when tested as blocks is surprising in view of results of DE GEE et al. which show that higher tin contents result in better intrinsic wear resistance. The test alloys of these researchers contained 15% rather than 20 $Lolead but a more important difference is probably that A contained 0.42% Ni and 0.045% P which C did not. DE GEE et al. attributed the effect of tin to two things. First, they said, addition of tin to copper-lead increases the tendency of the bearing surface to adsorb polar compounds and, second, tin increases the hardness of the matrix. As pointed out by ROWE 6, the latter is certainly true but it would seem doubtful that Wear, 15 (1970) 373-387

I~zfl~teme

of lead coded

and distviDd.iou

The presence of lead in bronze bearing materials is believed to be essentialzs” because of the protection it affords to the mating shafts during periods of inadequate lubrication. The lead inclusions smear over the bronze and transfer to the hard mating surfaces as well (Fig. 5). The smeared lead reduces friction and scoring tendency and so is especially beneficial at high sliding speeds. When tested as blocks sliding on steel rings, the wear rate of SO-IO-IO bronze (B) was identical with that of 70.-10-20 bronze (C) although the latter produced a somewhat lower coefficient of friction. This corroborates results of 1)~ GEE et al. who found, with the same test geometry, that lower wear rates are not obtained with

FRICTION AND WEAR OF SOME BRONZES

0.1

’ 60

,

60

I 70

I 80

HARDNESS.

1 90

I

I

100

110

ERINELL

Fig. 6. Influence of test geometry on wear rates of leaded tin bronzes as a function of hardness.

Fig. 7. Section through bronze E block after run-in showing surface coating of lead. Suriace was protected by nickel electroplate before sectioning. x 500

lead additions of greater than IO:‘, and also agrees with the earlier work of FRENCH AND STAPLES who concluded that there is no advantage to be gained by additions greater than 15”,b. However, with the opposite test geometry, that is, with a hard block sliding on the bronze ring, the 2036 Pb alloy C was considerably more wearresistant than was the xo”/bPb alloy B and the latter alloy showed excessive metal transfer to the mating surfaces as well. Bronze A, the 74-G-20 alloy, also wore less than did bronze B although it, too, showed excessive transfer. Alloys with larger lead grains had better wear resistance than the smaller lead-grained alloys when tested as blocks sliding on steel rings as was found by DE GEE et rcl. The sand-cast SO-IO-IO bronze (B) had a wear rate of just over half of that of the smaller lead-grained, continuously cast SO-IO-IO bronze (E) although the latter contained 0.42% Ni and 0.020~~ P which might be expected to improve its wear resistance. On the other hand, with hard blocks on bronze rings, the effect Wear, 15 (1970) 373-387

was exactly

opposite.

\T:ear rate

of tlre smaller

lead-grained,

continuously

ca5t

SO-IO--IO bronze

\vas generally only about a third of that of the larger, lc:&$ine~l, 80--IO--IO bronze. I~urthermore, thy continuously cast nickel bronze (H)

sand-cast

also showed considerably better Lvear resistancck than did the sand-cast nickel bronze (1;) although there were significant differences in composition in this cast: 2~s ~~11. The reversal of the effect of lead distribution

with the different

test geometries

could be an effect of hardness. When the ratio of the wear rates obtained with the two test geometries was plotted as a function of hardness, a reasonably smooth curve resulted which is strongly inflected at both its ends (Fig. 6.). Bronzes with hardnesses below about 60 Brine11 wore appreciably more as moving rings while those above about

rag Brine11 wore appreciably

less. Between

these

two values,

relative

wear

rates were much less sensitive to test geometry. The finer distribution was associated with an increase in hardness in both cases where effect of lead grain size was investigated

(comparison

such that the increase

of H with E and of I; with H). The pairs fell on the curve

in hardness

could halve been the important

factor rather than

the lead distribution,

per SP. 1)E GEE et al. attributed the better performance which they found for the bronzes with coarse lead distributions to less effect of the lead in these cases in reducing formation of a boundary lubricant film. They reasoned that the more coarse the lead distribution is, the smaller becomes the total circumference of the inclusions at each individual cross section of the material and, accordingly, the amount of smeared lead at the surface would decrease with increasing coarseness of the lead distribution. This was not true in did, in fact, affect thickness of smeared lead opposite to that conjectured. Bronze blocks purified mineral oil were slid on steel rings geometry

of DE GEE

et al.

the present work. The lead distribution on the bronze but it was in the direction of both B and E alloys lubricated with and then sectioned. (This was the test

and the one which produced

the lower wear rate for the

coarse lead distributions.) The lead coating on the larger lead-grained B alloy (Fig. 4) was much thicker than that on the small lead-grained E alloy (Fig. 7). In any case, the lead surface films appeared to be essentially continuous in both caseh. The results of the experiments where fatty acids were added to the lubricant also cast doubt on the conjecture of DE (;EE et al. With bronze blocks sliding on steel rings, the addition of stearic acid to the neutral oil decreased rather than increased the relative superiority of the larger lead-grained alloy. Furthermore, addition of stearic acid did not make the larger lead-grained alloy superior to the other when the reverse sliding geometry was used. It does not appear that differences in boundary film adsorption play much part in determining the superiorities of the different lead distributions.

Tribological behaviors of the bearing bronzes are strongly influenced by conditions of operation but not necessarily in the same way for each bronze. Sliding geometry and specific mating surface were found to have major effects. Furthermore, addition of a small amount of fatty acid to the lubricant greatly reduced wear rates and coefficients of friction. Where the bronze specimen was a moving ring sliding on a stationary hard M’FaJ,.15 (1970) 373-387

FRICTION AND WEAR

OF SOME BRONZES

387

block, there was a general trend toward lower wear rates with increasing hardness but hardness was not a completely reliable guide to wear resistance. Where the bronze specimen was a stationary block sliding on a hard steel ring, there was no apparent effect of hardness; the lowest wear rate was obtained with the softest bronze. With bronze blocks, equivalent wear rates were obtained with both IO and 20% Pb-containing alloys although the higher lead content resulted in a somewhat lower coefficient of friction. With the opposite sliding geometry, on the other hand, the 20% Pb showed a definite advantage. Furthermore, alloys with a more coarse lead distribution showed the lower wear rates reported by previous workers when tested as stationary bronze blocks but showed exactly the opposite effect when tested as moving bronze rings. This reversai of the effect of lead distribution may well be a consequence of the increased hardness of specimens with fine distributions. The superior wear resistance of stationary bronze blocks with coarse distributions is not a result of decreased adsorption of a lubricant boundary film on fine leadgrained alloys owing to increased coverage of the surface with lead as conjectured by previous workers. Aluminum bronzes are excessively hard for bearing applications and the example tested showed large amounts of metal transfer to the mating surfaces. REFERENCES B. LUNN, The wear resistance of tin bronzes and related alloys, wear, 8 (Ig65j 401. A. W. J. DE C&E, G. H. G. VAESSEN _&NDA. BEGELINGER, The influence of composition and structure on the sliding wear of copper-tin-lead alloys, ASLE Trans., 1.z (1969) 44. R. S. FEIN, Discussion of ref. (z), ASLE Trans., 12 (1969) 53; Discussion in Proc. NilSA-

Sponsored SymF., ~~te~d~sc~~l~nar~ Approach to Friction and wear, P. M. Ku, Ed., U.S. Government Printing Office, Washington, 1968, p. 321. H. J. FRENCH AND E. M. STAPLES, Bearing bronzes with and without zinc, J. Res. Natl. Rur. Std., .z (1929) 1017. M. ANTLER, The lubrication of gold, Wear, 6 (1963) 44; Sliding noble metal electric contacts, Proc. Inst. Mech. Engrs. (London), 182 Pt. 3A (1967-68) 355. C. N. ROWE, Discussion of ref. (2), ASLE Tram. 12 (1969) 52. Wear, r5 (rg70) 373-387