summary - Science Direct

4 downloads 0 Views 1MB Size Report
In the case of cast aluminium-graphite ... Mica is a lamellar solid which requires 20 times more energy than graphite to be sheared over ... This method of dispersion can be adopted by a ..... Bailey Brothers and Swinfen, Folkestone, 1972, p. 124.
61

Wear, 60 (1980) 61 - 73 @ Elsevier Sequoia S.A., Lausanne - Printed in the Netherlands

WEAR CHARACTERISTICS AND ALUMINIUM-MICA PARTICULATE

BEARING PERFORMANCE COMPOSITE MATERIAL*

OF

DE0 NATH Department

of Mechanical Engineering,

Banaras Hindu University, Varanasi (India)

S. K. BISWAS and P. K. ROHATGI Department of Mechanical Engineering, Zndian Institute of Science, Bangalore and CSZR Trivandrum Complex, Trivandrum, Kerala (India) (Received October 15,1979)

summary Some tribological properties of a mica-dispersed Al-4%Cu-1.5%Mg alloy cast by a conventional foundry technique are reported. The effect of mica dispersion on the wear rate and journal bearing performance of the matrix alloy was studied under different pressures and under different interface friction conditions. The dispersion of mica was found (a) to increase the wear rate of the base alloy, (b) to decrease the temperature rise during wear and (c) to improve the ability of the alloy to resist seizure.

1. Introduction Composite materials with solid lubricant as particulate dispersoids have been found to possess good antiseizing properties and a low wear rate. Aluminium-graphite [l - 31, bronze-graphite [4] and bronze-polytetrafluoroethylene have been shown to be better bearing materials than the respective matrix alloys without solid lubricant dispersoids. This is particularly true under conditions such as high or very low environmental temperature, high or low environmental pressure and high bearing load where the liquid lubricant may be rendered ineffective. In the case of cast aluminium-graphite [ 5 - 71 particle composites with 1% graphite the graphite becomes smeared [ 81 onto the bearing interface such that the relative movement of the, journal and bearing promotes easy shear between the lamellar planes of the graphite. This ensures a low coefficient of friction and prevents bearing seizure in the absence of a liquid lubricant. In the presence of a lubricant the aluminium alloy-graphite *Paper presented at the International Conference on Wear of Materials 1979, Dearborn, Michigan, April 1979.

62

particle composite is found [ 31 to run under mixed and boundary lubrication conditions whereas the same base alloy without graphite seizes even in the hydrodynamic region. The wear rate [S] of the aluminium alloygraphite particle composite is lower than the wear rate of base alloys without graphite. Mica is a lamellar solid which requires 20 times more energy than graphite to be sheared over the lamellar planes [ 91. However, it is oxidation resistant and has been used as a grease filler for wagon axles, a bag lubricant for moulding and a self-lubricant filament tube. The abundance of mica in India and its ability to retain its lubricating properties (as compared with graphite) under high temperature and high vacuum conditions recommends its use as a dispersoid particulate in a ~~-lub~~a~g composite bearing material. It has been shown that additions of mica to silver [ 10, 111, bronze [ 121 and nickel [ 131 using powder metallurgy techniques have led to an improvement in the antifriction properties of base materials. However, there has been very little work done to produce metal-mica composites by liquid metallurgy (foundry) techniques and to establish the antifriction properties of such composite alloys. The only work reported on the dispersion of mica particles in aluminium alloys using a foundry technique is that of Sato and Mehrabian 1141. They made the composite by the compocasting method which requires specialized equipment to stir the base alloys between the solidus and the liquidus temperatures. However, they have not reported any me~u~rnen~ of the bearing ch~a~~~stics of these materials. In this paper the journal bearing characteristics and wear properties of an aluminium-mica particulate composite material made by a foundry technique in which the mica was dispersed above the liquidus temperature of the base alloy are reported. This method of dispersion can be adopted by a conventional foundry without the need for specialized equipment. Wear specimens and journal bearings made from the two types of material given in Table 1 were tested to establish the effect of mica dispersion on the bearing characteristics and wear properties of the aluminiumbase alloy. Figure l(a) shows the distribution of mica in the composite alloy on a macroscopic scale. Figures l(b) and l(c) show typical microstructures of the base alloy with and without mica respectively. TABLE 3

Base alloy Composite alloy

Al

Mg

cu

si

Mica

94% 92%

1.5% 1.5%

4% 4%

0.31% 0.31%

0 2 - 2.5%

63

(a)

(b)

Fig. 1. (a) mpical macroscopic view of Al-4Cu-1.5Mg-2 mica composite; (b) typical microscopic view of Al-4Cu-1.5Mg base alloy (magnification 225X); (c) typical microscopic view of AL4Cu-1.5Mg-2 mica composite (magnification 225X).

2. Apparatus and tests 2.1. Wear reds tance The wear test apparatus is shown in Fig. 2. Cylindrical specimens 9.5 mm in diameter and 10 mm long made from the aluminium alloys were pressed on to a steel disc of hardness 53 R, rotating at a speed of 2850 rev min-1. A spring balance was attached to the pressure spring through an axial hole in the bolt. The bolt was periodically tightened to maintain constant spring balance tension and hence constant pressure on the specimen as it wore off. SAE 40 motor oil was continuously supplied at 10 drops min-’ to the tip of the specimen in contact with the steel disc. The temperature of the specimen at a point 2 mm away from the mating surface was measured with a thermocouple. Six samples of a particular composition were tested at each pressure

64

Fig. 2. Schematic

diagram of wear test apparatus.

Fig. 3. Loading system of the hearing testing machine.

and the tests were carried out at four different pressures. Tests were run for 10 min at each pressure. Specimens were thoroughly cleaned and weighed on a single-pan electrical balance before and after each test. The weight loss was taken as a measure of wear. 2.2. Bearing characteriatics Figure 3 shows a shaft rotating in a journal bearing fitted in a housing which is itself free to rotate. The bearing presses against the shaft under a radial vertical loading. When the shaft rotates it exerts a force to overcome the friction resistance at the bearing/shaft interface. This force gives rise to a moment which tilts the floating housing about the shaft axis in the direction of shaft rotation. The force required to bring the housing back to its original position is a measure of the friction resistance. This force is measured by the spring balance. The coefficient of friction is given by p=F/W Oil was gravity fed from an overhead tank to an axial hole in the shaft and pumped to the shaft/bearing interface through a connecting radial hole. The temperature of the lubricant at the shaft/bearing interface was measured by a thermocouple. Shafts of EN 24 steel were hardened to 56 - 57 Rc. The inner surface of the bearing (1.815 in outside diameter X 1.421 in inside diameter X 1.265 in long) was machined and the shaft ground (0.2 pm centre-line aver-

66

age (c.1.a.)) to these dimensions within the Xim% given by @ib

-0,)/&t

=

0.001 + 0.0005

The bearings were run in three successive stages for friction measurement. In the first stage the bearing was run-in for 8 h at 12 rev mine1 and at a load (25 kgf) well below the seizure load. The second and third stages consisted of running the bearing at shaft speeds of 12 and 6 rev min-l respectively up to the seizure load. The frictional force and interface temperature were measured during the latter stages, At each stage the load was increased in steps of 5 kgf and at intervals of 30 min. The following series of tests was conducted to study the effect of mica dispersions on the bearing ch~c~~stics of the Al-Cu-Mg alloys. (1) Lubricated tests: five bearings of each type (with and without mica) were tested using SAE 10 oil. Two shafts were used for testing the two types of bearings. (2) Semi-dry tests: after running-in two bearings of each type with SAE 30 oil the bearings and the shaft were cleaned with acetone and dried to remove traces of oil from the surface. The bearings were now tested without any further lubrication following the procedure set out above. (3) Dry tests: five bearings of each type were tested under completely dry conditions. There was no initial running-in period with oil; however, the bearings were cleaned with acetone before testing. Two separate shafts were used for testing the two types (with and without mica) of bearings.

3. Results and discussion The effect of mica dispersion on the wear rate-bearing pressure characteristics and the contact temperature-time characteristics of the Al-Cu-Mg alloy are shown in Figs. 4 and 5 respectively, The wear rate characteristics were obtained by arithmetically averaging the results from six identical tests; the average spread +u about each point was 0.4 mg min-’ . Sato and Mehrabian [ 143 observed that at a bearing pressure of 1 X 10s2 kgf mmV2 the wear rate of extruded ~u~nium-mica composite and extruded aluminium base alloy samples increased with rubbing speed. They found that at a speed of 0.16 m s-l the wear rate (0.1 mg min-l) of the composite is almost the same as that of the base alloy whereas at a higher speed of 0.4 m s-l the composite sample yielded a higher wear rate (of 0.29 mg min-l) than that (0.22 mg min-I) of the base alloy. This trend of an increase in wear rate due to mica dispersion is corroborated by the present investigation at a higher rubbing speed of 20 m s-l. The increase is possibly due to the poor bonding of the mica particles with the matrix. This leads to the debonding of the particles during wear and their eventual removal from the matrix during testing. The weakness of bonding is inferred from the presence of voids around the mica particles as observed in the scanning electron micrograph of a fractured surface of a tensile specimen shown in

66 s,_

I

100

+

AI-CCu-15Mg

+

AI-4Cu-1

5Mg

-a-

AI-4Cu-l.SMg-lhivza

+

AI-401-l

SMg-2Mlca

*

AI-CCu-l.SMg-2M~a

. . . f

:

125-

2

I I

ol 0

1

2

3

Pressure,Kg/m&10-a

Tlmo,mins

Fig. 4. Wear rate vs. pressure. Fig. 6. Temperature rise (during wear test) vs. time.

Fig. 6. Chemical analysis of the wear debris produced during the sliding of the composite alloy against steel showed the presence of mica in the debris. In the present wear tests the mica particles can therefore be expected to be located at the bearing shaft interface. The lower temperature rise during wear testing of composite alloy as compared with the base alloy is possibly

Fig. 6. SEM photograph of a fractured surface of Al-4Cu-1.5Mg-1.7

mica.

67

due ing the the

to the presence of these mica particles which reduce the frictional heatat the interface by reducing metal-to-metal contact. presence of these mica particles which reduce the frictional heating at interface by reducing metal-to-metal contact. The effect of mica dispersions on the bearing characteristics (under lubricated condition) of Al-Cu-Mg alloys is shown in Figs. 7 and 8 for 6 and 12 rev min-l respectively. The coefficients of friction plotted are the arithmetical averages of coefficients (corresponding to a particular value of ZNP) obtained from five tests involving five different bearings. A typical spread of the experimental points is shown in Fig. S(b). The following observations can be made from the results presented. (1) The friction coefficient for the composite alloy bearing is marginally lower than that corresponding to the base alloy bearing in the hydrodynamic region. (2) Beesley and Eyre [ 151 observed that under dry conditions a weak material undergoes sublayer plastic flow and surface rupture with a consequent rapid rise in wear and friction at a bearing pressure lower than that at which a stronger material starts to exhibit the same trends. Figure 9 shows that the 0.1% proof strength in compression (mean of three tests) of the aluminium alloy decreases with increasing mica content. Therefore a 0.035

0.030

0.025

*0.020

I -o-

AI-4Cu-1.5Mg

-fb

Al-4Cu-1.5Mg-2to2.5Mica

-

0.015-

0.010 -

I 5

151 I

ZNIP,

II 10

Bearing

I 15

I 20

I 25

I 30

parameter.

Fig. 7. Coefficient of friction vs. bearing parameter ZN/P. Test conditions: SAE-10 oil lubrication, 6 rev min-l.

68 0.030’ -o0.02$-

Al-PCu-1.5

Mg

-LB- AI-4Cu-l.5Mg-2tO2.5MlC~

0.020 -

0.010 -

0.00s -

4 I 5

O*

1 51’ I ‘- & 8 ta~ktydredynamic L 13; ,I, 10

, IS ZNIP,Bearing

rqion

1 I 20 25 parameter.

I 30

I 36

I 40

J 45

I 20

I 30

I 35

I 40

1 45

(a)

0.035y -o0.030 - -&-

0.025

Al-OCu-1.5Hg At -4Cu-1.5Mg-2

to2SYim

-

0.015 -

O.OlO-

0.006

0

1 5

I 10

I 15 ZNlP.

fkwin~

I 25 paranmhr

@I

Fig. 8. (a) Coefficient of friction US. bearing parameter ZN/P. Test conditions: SAE-10 oil lubrication, 12 rev min-l. (b) Coefficient of friction us. bearing parameter ZN/P. The 7 5% confidence limits of the experimental points are shown. Test conditions: SAE-10 oil lubrication, 12 rev min-l.

mechanism similar to that proposed by Beesley and Eyre can be assumed to be operative: the composite material which has a bulk strength lower than that of the base alloy shows a rapid rise in friction at the termination of the hydrodynamic region at a pressure lower than that at which the hydrodynamic region of the base alloy bearing terminates. (3) Figures 7 and 8 show that the composite alloy bearings are able to run under high friction conditions beyond the hy~odyn~ic region, whereas the base ahoy bearing without any mica seizes at the value of ZN/p which marks the end of the hydrodynamic region. (4) The ability of the composite ahoy bearing to run under conditions of high friction was further confirmed by the semi-dry and dry tests, the

69

results of which are shown in Figs. 10 and 11. Under semi-dry conditions the base alloy bearing without mica exhibited visually observable stick-slip at a load of 5 kgf. The stick-slip became violent and led to seizure at all loads beyond 5 kgf. In contrast, the composite alloy bearing did not show any sign of visually observable stick-slip under these conditions even at a load of 50 kgf which corresponded to approximately 5 h of running. Although it was difficult to measure the coefficient of friction accurately for the base alloy bearing under stick-slip conditions, Fig. 10 shows that the average level of both alloys lies between 0.15 and 0.25 under semi-dry conditions. For the composite alloy bearing this level did not change with increasing loads. Figure 11 gives the average of results from five bearing tests conducted under completely dry conditions and shows that the characteristic behaviour of the two types of bearings under dry conditions is similar to that observed under semi-dry conditions except that the level of friction coefficient for the composite alloy bearing under dry conditions is higher than that observed under semi-dry conditions. As no appreciable difference was observed between the levels of the coefficient of friction corresponding to the nonmica bearings when run under semi-dry and dry conditions it seems unlikely that this difference, when observed for the bearing with mica, could have been caused by the running-in which preceded the semi-dry runs and not the dry runs. A possible reason for this difference may be the ability of the composite bearing, once in contact with a lubricant, to absorb and store it in its pores. The absorption is such that the absorbant resists removal even if the bearing is degreased and dried. Under semi-dry conditions the absorbed oil under pressure may migrate to the bearing surface to reduce the coeffi-

0.3 r -o-

AI-4Cu-1.5

Mg

t

AI-4Cu-1.5

Mg-2M1ca

. \Artthmotlc

.

.

Test drscontwwod vlolont

I:_, 0

Fig.

2

I

o+ 0

Weight % Mica

9. 0.1% proof stress (compression)

10

mean

20

us. weight % of mica.

Fig. 10. Coefficient of friction us. load. Test condition: semi-dry.

after

stick-slip

Load,Kgs

30

40

50

Al-4Cu-l.SMq AI-LCu-l.SMg-2Mica

20 Load,

30

Rqs

Fig. 11. Coefficient of friction as. load. Test condition: dry,

cient of friction. These tests therefore indicate that the ~urni~i~rn-mica composite is a bearing material which can be used in machine tool and other applications where there are chances of a cut-off in liquid lubricant suppIy due to power failure or other reasons without running the risk of a catastrophic seizure of the moving parts. It is suggested that this improvement in the bearing property of a conventional aluminium alloy by the dispersion of mica is due to the release of mica to the mating surface by the bearing during running. Mica acts as a solid lubricant preventing metal-to-metal contact and seizure, From the wear test results (Fig. 4) it is expected that the higher the load the more mica is released with wear debris from the bearing. As shown in Fig, 12 profuse wear debris was generated under dry conditions for both alloys at even the smallest loads (10 kgf), Chemical analysis showed that the composite bearing produced wear debris which contained 9.88% silicon as compared with the base alloy debris which contained 0.31% silicon. The comparatively high percentage of silicon in the composite ahoy bearing wear debris is apparently due to the mica particles removed during testing. Comparison of the surface finishes of two shafts each run with only one type of bearing under lubricated conditions yields indirect evidence of the presence of mica particles at the bearing/shaft interface. Table 2 shows the improvement in the shaft surface finish (c.1.a.) when run with composite alloy bearings. Figure 13 shows the mirror polish imparted to the shaft surface when it is run with the composite alloy bearing under lubricated eonditions. Mica slurry is a known lapping agent [16] and hence this improvement in the surface finish of the shaft is assumed to be caused by the lapping

71

Fig. 12. Bearing run under mica.

TABLE

dry conditions:

(a) Al-4Cu-1.5Mg;

(b) Al-4Cu-1.5Mg-2.5

2

Details of the shaft used for friction

testing Remarks

Mica content of bearing (%I

C.i.o. uafue of the shaft (Pm)

56

2 - 2.5

0.32

0.1

Shaft run with three composite bearings under SAE 30 oil lubrication

55

2 - 2.6

0.18

0.08

Shaft run with five composite bearings under SAE 10 oil lubrication

56

0

0.10

0.11

Shaft run with three matrix alloy bearings under SAE 30 oil lubrication

57

0

0.20

0.18

Shaft run with five matrix alloy bearings under SAE 10 oil lubrication

Gardner &

Before test After test

action of a slurry which consists of mica particles (released during the test) and lubricating oil.

4. Conclusions Mica particles dispersed by liquid metallurgy and casting in an aluminium copper alloy have the following main effects on the antifriction properties of the base alloy. (1) The wear rate increases with mica content as well as with bearing pressure. This may be due to the loss of mica particles which are loosely bonded to the matrix.

^__....._ _ .

.._

^

.-..-

(b) Fig. 13. Shaft run under lubricated (b) Al-4Cu-1.5Mg-2.5 mica.

conditions

against a bearing of (a) Al-4Cu-1.5Mg

and

(2) The Al-Cu-mica composite alloy bearing can run under (a) boundary lubrication, (b) semi-dry and (c) dry conditions where the micafree base alloy bearing seizes. The main reason for this is probably the presence of loose mica particles released by bearing pressure at the bearing/ shaft interface. In the absence of any liquid lubricant film at this interface the mica acts as a solid lubricant and diminishes metal-to-metal contact and thus prevents seizure. (3) These preliminary results indicate that aluminium-mica particle composites may be good candidate materials for bearings required to run under conditions where extensive metal-to-metal contact may be expected.

Acknowledgments The authors are grateful to the CSIR for providing the grant for carrying out this investigation and to the authorities of the Indian Institute of Science for providing the necessary facilities. The authors are also grateful to Mr N. Raman of the Department of Mechanical Engineering for useful suggestions. Nomenclature Dib Ds F L N

bearing internal diameter shaft diameter friction& force (spring balance reading) bearing length bearing speed (rev mind’)

73

P W z g u

W/LDib, bearing pressure bearing load (normal) viscosity of the lubricant after correction coefficient of friction standard deviation

for temperature

rise at the bearing surface

References 1 B. C. Pai and P. K. Rohatgi, Graphite aluminium - A potential bearing alloy, nuns. Ind. Inst. Met., 27 (1974) 97. 2 V. G. Gorbunov, V. 13. Parshin and V. V. Panin, Aluminium-~aphite antifriction alloys, Russ. Cast. Prod., 8 (August 1974) 346. 3 F. A. Badia and P. K. Rohatgi, Gail resistance of cast graphitic aluminium alloys, SAE Itans., 76 (1969) 1200. 4 G. C. Pratt, Graphite-metaI composite for dry and sparsely lubricated bearing appfications, Tribology, 6 (6) (1973) 259. 5 F. A. Badia and P. K. Rohatgi, Dispersion of graphite particles in aluminium castings through injection of the melt, ?‘rans. Am. Foundrymen’s Sot., 77 (1969) 402. 6 F. A. Badia, Dispersion of oxides and carbides in aluminium and zinc alloy castings, Trans. Am. Foundrymen’s Sot., 79 (1971) 347. 7 F. A. Badia, D. F. MacdonaId and J. R. Pearson, Graphitic aluminium - A new method of production and some foundry characteristics, nans. Am. Foundrymen’s Sot., 79 (1971) 265. 8 B. C. Pai, P. K. Rohatgi and S. Venkatesh, Wear resistance of cast-graphite aluminium alloys, Weor, 30 (1974) 117. 9 E. P. Bowden and D. Tabor, Friction and Lubrication in Solids, Oxford Univ. Press: Ciarendon Press, Oxford, 1964, p. 199. 10 V. F. Afanas’ev, M. A. Parkhomenko, N. I. Semenyuk, V. B. Vishnevskii, M. K. Kovpak and L. V. Zabolotnayi, New mate&Is based on silver synthetic micas, Fiz.Khim. Mekh. Muter., 5 (6) (1970) 680. 11 U.S.S.R. Patent 209,765 (Jan. 26,1968), to V. F. Afanas’ev, V. B. Vishevskii,M. K. Kovpak, M. A. Parkhomenko, S. G. Tresvyatskii and N. I. Chernavskya. 12 French Patent, 2,031,667 (1970), to J. Rollet. 13 V. N. Pavlikov, A. V. Thachenko, A. D. Kondratenko and S. G. Tresvyatakii, Nickel synthetic mica base cermets, Sov. Powder Metall. Met. Cerum., 13 (10) (Oct. 1974) 806. 14 A. Sato and R. Mehrabian, Aiuminium matrix composites: fabrication and properties, Meta~l. mans., 78 (Sept. 1976) 443. 15 C. Beesley and T. S. Eyre, Friction and wear of ahrminium alloys containing copper and zinc, !Z’riboZ.Znt., 9 (2) (1976) 63. 16 R. C. Gunther, Lubrication. Bailey Brothers and Swinfen, Folkestone, 1972, p. 124.