Investigation of wear during dry sliding of steel against bronze

Wear -

3

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

INVESTIGATION BRONZE

K. WELLINGER,

OF WEAR DURING DRY SLIDING OF STEEL AGAINST

H. UETZ

UND

Staatliche Materialprtijungsanstalt, (Received

April a$h,

T. KOMAl

Stuttgart

(Germany)

1969)

SUMMARY

Wear during dry sliding of steel against bronze has been investigated. The effects of time of sliding, load and velocity of sliding were assessed. A wear mechanism is postulated to explain the results under the conditions of test. INTRODUCTION

As part of an OECD co-operative research programme, wear under conditions of dry-rubbing contact has been investigated in order to determine factors controlling the wear process and to establish factors influencing the mechanism of wear. TEST SET-UP

The tests were carried out on the machine developed by SIEBEL AND KEHL~, the principle of which is shown in Fig. I. The rotating test piece is held in a device at the lower end of a vertical, rotating shaft. The stationary test piece is mounted on an axially displaceable spindle and is pushed against the rotating test piece by means of a lever system. Sliding eurface

Counterpart(rotat~ng) Fixture for lever wfth torque measurement Test piece (fixed) Arrangement

for

loading

stationary (with slot) Fig.

I.

Scheme of test set-up.

Fig.

2.

rotating (without slot)

Test-specimen. Wea”, 14 (19%)

3-14

K. WELLINGER,

4

H. UETZ, 1‘. KOMAI

The annular test pieces slide face against face (see Fig. 2). The lower test piece has two slots for easy removal of abraded particles. The test pieces are fanned with compressed air to cool their sliding surfaces. MATERIALS,

WEAR

PARAMETERS,

TEST ARRANGEMENT

AND CONDITIONS

Investigations were carried out on a material combination of steel SAE 1045, normalized, (upper test piece), with HV 226 kg/mm2 hardness, against tin bronze containing 10.6% Sn, 0.6% P, remainder Cu, with HV 125 kg/mm2 hardness; this corresponds to cast tin bronze G-Sn Bz IO. The end faces were ground and polished with fine polishing paper (No. 320) and the mating surfaces cleaned with alcohol. The amount of abrasion was determined by weighing each test piece on an analytical balance with a sensitivity of I mg. From the loss of weight, the reduction of height was calculated by taking into account the specific density, the surface area of the test piece, and the sliding distances, as follows. AL = F

(pm)

roJAG us = r (pm/km) where AL is the absolute amount of abrasion, vs the specific relative amount of abrasion, G the loss of weight (g), s the sliding distance (km), F the area of the mating surfaces (cm”) (slotted test piece, F,=z.6cm2; slotless test piece, F,=3.0 cm2) and e the specific density (g/cm”) (steel es=7.8 g/cm3; bronze eB=8.8 g/cm3). The smallest measurable amount of wear is determined by the smallest TABLE

I

WEIGHT

REDUCTION

AG AND WEAR

AL AS A FUNCTION

.______. Pressure

&lcm3)

Load

Sliding distance (ml

(kg)

Velocity

(pm) 5.5 12.4 18.1

300 400 500 800 1000

54.3 70.5 84.0 122.4

23.5 28.0 40.8

165.3

55.1

100

24.9 54.0 75.0 98.7 III.0

8.3 18.0 25.0

200

26

AL ~+g)

16.5 31.2

100 200

IO

I.0 -___

B~O%Z (x

300 400 500 600 800 IO00

132.9 166.2 ‘95.7

Wea+‘, 14 (1969) 3-14

DISTANCE

(m/s)

0.07 AG

5 I3

OF SLIDING

~-

32.9 39.0 44.3 55.4 64.9

Steel ..______ AG (x Io-3g) -0.7 0.7 2.8 3.2

3.5 6.0 6.2 -1.1 2.1 3.0 4.6 6.2 6.7 8.3 10.2

-___-

___-.

AG g)

(x

-0.3 0.3 I.2

I.4 I.5 2.6

2.7 -0.5 0.9 1.3 2.0

2.7 2.9 3.6 4.5

3.9 6.9 9.3 13.2 15.6 19.5 19.8 10.8 16.5 20.4 24.3 28.2 31.8 37.2 44.4 ___._

--

-

Steel

Bronze

AG (x ro+g)

AL

I.3 2.3 3.1 4.4 5.2 6.5 6.8

-2.5 -4.2 -4.8 -3.5 -3.0 0.7

- I.1 --1.8 ---2.1 ~ I.5 ~ I.3 0.2

6.7

2.9

3.6 5.5 6.8 8.1

- 1.6 -4.8 -6.5 -5.3 -6.0 -4.4 -1.9

AL IO-~ g)

(pm)

9.4 10.6 12.7 14.8 ~___

..__

~~__(pm)

3.0

-0.7 -2.1 -2.8 -2.3 -2.6 -1.9 -0.8 ~

I.3 _ ~_.... .-

DRY SLIDING OF STEEL AGAINST

BRONZE

5

measurable reduction of weight, G = 0.001 g, which for a steel test piece without slots is of the order of AL=o.4 pm. The relation between wear rate and sliding length was determined by measuring the change of weight after every IOO m of sliding at pressures of 5 kg/cm2 and IO kg/cm2 and at velocities of 0.07 m/set and 1.0 m/set (see Table I). Since the wear behaviour during dry sliding2 depends essentially on sliding velocity and load (contact pressure), the variables were tested over a sliding distance of 300 m at velocities of 0.07-1.0 m/set and at pressures of z.5--20 kg/cm2 (see Table II). Each of the tests was repeated two or three times. The properties and characteristics, which are prone to change during the wear process, such as crystalline structure, state of surface, and hardness of the sliding layers, were carefully investigated. RESULTS

Results dependent on sliding length At the low velocity of 0.07 m/set (Fig. 3(a)), the slope of the curve of the absolute amount of wear AL of bronze run at pressures of 5 and IO kg/cm2 decreases gently, that is, the amount of wear v, per km decreases with increasing sliding length. For the steel, the result is an increase of weight during the first IOO m sliding distance due to the material transfer of bronze onto the steel surface. This agglomeration of bronze on the steel is, at the beginning, greater than the wear of the steel itself.

b) v=lm/s

u

500

m

1000 0 SltdmgLength Cm)

Fig. 3. Wear AL as a function of sliding length.

Wear,

14

(1969) 3-14

6

Pvessuve

(kg/cm*)

(Load

(kg))

5 (1.7)

2.5 (6.5)

Bvonze AG (x IO-~ g)

Steel

7’d

AG

(/m/km)

(x

lirm 3l’ 10-8 g)

z’g

AG

(pm/knr)

(x

IO-“g)

Steel

“8

AG

(p/km)

(x

IO-3g)

* A negative value signifies an increase in weight

From a 200 m sliding length onwards, the steel shows a reduction of weight. As may be seen from a micrograph of the steel test piece (Fig. 4) taken after a run of IOOO m at an angle of II’ against the end face and transverse to the direction of movement, the transferred layer does not cover the entire surface of the steel test piece but accumulates mainly in the cavities. According to a qualitative investigation with an electron microprobe, this layer consists of about 62-74% copper, IO-Z~~/, iron, 5-69/o tin, and about 107; oxygen. This new surface of the steel specimen is exposed to wear, whereby steel particles are also worn away. X-ray analysis carried out by DE GEF, AND ZAAT" for tests of steel against a copper-tin alloy at a velocity of 0.77 mjsec and 2 kg/cm2 pressure has shown that this layer consists of a mixture of copper protoxide and cupric oxide and that the material transfer is less in an atmosphere argon than in an atmosphere of oxygen.

Weav, 14 (1969) 3-14

of

DRY SLIDING OF STEEL AGAINST

IO

BRONZE

(26)

7

20

(pmlkm)

AG (x 10-s g)

;;m+n,

AG (x

75 48 32

91 61 41

-2 --I --I

-3.3 -1.6 _ 1.6

21 I4 7

27 18 9

-2 -4

-3.3 -6.6 +4.9

AG (x

u.3 IO-=

g)

3

(52)

Bronze

Steel

Bronze

Steel AG

Io-ag)

ull

;m,km)

(x

76 58 40

97 74 51

-2 --2 -3

-3.3 -3.3 -4.9

23 21 II

29 27 22

-3 4 I2

-::: 19.8

IO+

gl

(pmlkm)

Fig. 4. Micrograph of steel specimen cut at an angle of I I’ w.ith I-espect to the end face. (a) Overall view (x 100); (b) detail of (a) (x 500). Sliding length IOOO m Pressure IO kg/cm2 Velocity 1 m/set Wea+‘, =4 (w69)

3-14

8

K. WELLIn‘GER, H. (JET%, ‘I.. KOBIAI

At a higher velocity of I nljsec (Fig. 3(b)), the amount of wear is only a fraction of that observed at the lower velocity, i.e., for the range of up to 500 m sliding length, it is about one-third to a quarter. The trend of the wear curve is the same as at the lower velocity but more pronounced. The weight increase of the steel specimen and, hence, the material transfer from the bronze onto the steel surface, is greater and can be observed for a longer sliding distance. Up to that point (apart from the very beginning), bronze has really been moving against bronze, i.e., one layer against another, but not steel against bronze. The beginning of the wear of the steel specimen (that is, the deposited layer) and the sudden change in the wear

Fig. 5. Micrograph of the bronze specimen cut at an angle of I IO with respect to the end face through the bearing place. (a) General view (x 50); (b) detail of (a) ( x 100). Sliding length 1000 m 10 kg/cm2 Pressure Velocity 1 mjsec J+‘ear, ‘4 (IV%) 3-14

DRY SLIDING OF STEEL AGAINST BRONZE TABLE

III

RESULTS

OF HARDNESS

Matevial

9

TEST

Place of hardness measurement

-

Hardness (kg/mma)HV IO

HV 0.05 11+130 187-201 359-567 290 270 256

Bronze

Virgin material Sliding surface* without oxide layer Sliding surface with oxide layer Oxide layer at different depth*

122-127

Steel

Virgin material (1) perlitic structure (2) ferritic structure Sliding surface** without layer Sliding surface with layer

216-232 240-250 165-175 382-395 24”

* According to the test 300 m of sliding length, 1.0 m/set, IO kg/cm2. ** According to the test 300 m of sliding length 0.07 m/set, IO kg/cm2.

60

ii <

40

5 ? j

20

0 20

0 -10

0.05

0.l

0.3 Velocity (m/s&

0.5

Fig. 6. Wear o. as a function of velocity. Wear, r4 (1969) 3-14

IO

Ii. WELLISGEI<,

Fig. 7. Sliding surfaces ofpom(x IO). Wear,

r$ (1969) 3-14

of the material

combination

steel SAIC ro45-bronze

H. 1

after

;t sliding

length

DRY SLIDING

OF STEEL

AGAINST

II

BRONZE

Steel SAE 1045: (a), 0.07 mjsec, 5 kgjcml; (b), 0.5 m/set, Bronze: (d), 0.07 m/set, 5 kg/cm2; (e), 0.5 m/set, 5 kg/cm2;

5 kg/cm2; (c), I.om l/set, (f), 1.0 m/s’-, Io kg/ ‘cmz. ~'t'at',I4

IO

kg/c:mz.

(1969)

S-14

12

K. WELLIXGER,

H. LIETZ, 1‘.KOMAI

diagram of the bronze is probably due to the fact that, from that point on, the process is increasingly influenced by the appearance of oxide particles on the bronze specimen. Apparently the layer previously transferred from the bronze onto the steel is readily rubbed off by the relatively hard oxides of the bronze, which at the same time increases the wear of the steel Thereby, the bronze layer which was spread on the steel surface may be transferred back onto the surface of the bronze specimen by a rewelding process of the separated particles. Figure 5 shows a layer locally formed in such a manner, the structure of which points to the occurrence of plastic deformations. Its bearing uea clearly projects out from the surface. It is likely that steel particles are also included in this layer. The micro-Vickers-hardness HV 0.05 of this layer (Table III) reaching a maximum value of about 550 kg/mm2 is greater than that of bronze and of steel. This explains why the surface of the steel specimen is attacked and worn away at a relatively high rate. These tests show that during the wearing process the materials themselves change through deformation, oxidation, materials transfer, etc. The type of modification is also influenced by the length of the sliding distance so that for different sliding distances, different wear rates may be expected. This has been clearly confirmed. Results dependent OHvelocity The wear curves (Fig. 6) show a velocity dependence different for material and mating material which, particularly in the case of steel, is influenced by load or contact pressure. Typical photographs of the sliding surfaces are given in Fig. T(a)-(f). At the lower loads of 2.5 and 5 kg/cm2, the weight increase of the steel specimen is small up to a velocity of 0.5 m/set. In this range, the layer transferred from the bronze specimen was uniformly distributed onto the steel surface (Fig. 8). At a velocity of I m/set, the material transfer to the steel specimen increases with pressure; apparently, under these conditions, the layer transferred from the bronze to the steel becomes thicker and more resistant (Fig. 7(c)). In the velocity range of up to 0.5 m/set, no oxides can be seen on the sliding surface (Fig. 7(d)) and the wear of the bronze specimen decreases slightly. Above 0.5 m/set, the wear rate drops more sharply and traces of oxides become visible on the sliding surface. Beyond this velocity, more and more oxides are formed on the bronze specimen (Fig. 7(e) and (f)), which probably accounts for the reduction of wear. The weight increase and the abrasion products of the steel specimen probably cause only a small modification of the sliding surface which hardly influences the oxides layer on the bronze. At the critical velocity of 0.5 m/set mentioned above, the weight of the steel specimen also begins to increase more rapidly. At the high loads of IO and 20 kg/cm”, the wear rate of the bronze specimen decreases with the velocity, over the range tested. At a velocity of 1.0 m/set, the formation of oxide on the surface can readily be seen; thereby the slope of the wear rate curve of the steel becomes considerably steeper, as has already been observed for the dependence on the sliding distance (see Fig. 3). This wear rate develops essentially through the effect of the hard oxide layer of the bronze, which scratches grooves into the steel surface (similar to those in Fig. 7(c)). Owing to the rough surface thus produced and to the wear particles formed between the surfaces, the Wear, I# (1969) 3-14

I3

DRY SLIDING OF STEEL AGAINST BRONZE

oxide layer of the bronze specimen is disturbed and the wear of the bronze specimen does not decrease as quickly as might be expected. The transition to severe wear of the steel at the higher pressure of 20 kg/cm2 occurs at a lower velocity than it would at the lower pressure of IO kg/cm2. Table IV summa&es the appearance of characteristic surfaces as a function of pressure and velocity. Here again, it shows the close correlation between conditions of velocity and load which lead to the formation of oxides.

Fig. 5. Bronze layer deposited on the sliding surface of steel

TABLE SYNOPSIS

Velocity (mlsec)

0.07-0.44

zoo).

IV OF CHARACTERISTIC

Pressure

SURFACE

STATES

AS A FUNCTION

OF PRESSURE

AND

VELOCITY

(kg/cm=)

2.5

5

IO

::

2

::

0.50 1.0

(X

20 X

0 n

A

o

0

x Steel, evenly covering layer; bronze, no formation of oxide. 0 Steel, partly covering layer; bronze, very little formation of oxide. o Steel, ground surface: bronze, positive formation of oxide.

The transfer of bronze probably takes place first in the ferritic structure, the softer part of the steel (see Fig. 8), which after the test has a slightly polished appearance. The embedded layer is more deeply impressed into the ferritic structure than into the Perlite. wea+‘. I4 (1969) j-14

K. WELLINGEK,

14

H. I~lil‘%, T. KOkl.~l

CONCLUSIONS

The formation of oxides on the surface of the bronze occurs at high velocity and high load. Velocity and load have a combined influence on the temperature increase of the surface. Frictional heat enables the surface of the bronze specimen to react chemically, i.e., the copper, as an alloying element of bronze, reacts quite quickly with the oxygen of the atmosphere at the high temperatures produced. However, the formation of oxides requires a certain time or a corresponding sliding length. As soon as oxidation takes place, the wear process changes rapidly, as shown in Fig. 4. For this reason, as already mentioned, it must be taken into account that the wear rate may be different for different sliding distances (that is, different from those mentioned in Fig. 6). REFERENCES I E. SIEBEL AND B. KEHL, Investigation

of wear during sliding of metals, Arch.

(1936) 563-570. 2 H. UETZ, Contribution todryslidingbetweenmetals,.%nderh 3 A. W. J. DE GEE AND J. H. ZAAT, Wear, 5 (1962) 257-274. Wear,

14 (1969)

3-14

MPA,

Stuttgart,

Eisenhiittenw., 1964,~~.

59-66.

9