Sliding wear characteristics of non-equilibrium Cu-Pb alloys

Sliding wear characteristics of non-equilibrium Cu-Pb alloys

Wear, 146 (1991) 257 257-267 Sliding wear characteristics alloys of non-equilibrium Cu-Pb P. A. Molian”, V. E. Buchanana*, T. S. Sudarshan’ and...

997KB Sizes 0 Downloads 23 Views

Wear,

146 (1991)

257

257-267

Sliding wear characteristics alloys

of non-equilibrium

Cu-Pb

P. A. Molian”, V. E. Buchanana*, T. S. Sudarshan’ and A. Akersb “Department of Mechanical Engineering, bDepartment of Aerospace Engineering Engineering Mechanics, Iowa State University, Ames, IA 50011 (U.S.A.) “Materials [email protected] Inc., P.O. Box 4817, Falls Church, VA 22044 (U.S.A.)

and

(Received April 19, 1990; accepted November 7, 1990)

Abstract The sliding wear behavior of copper-lead alloys was studied as a function of lead content (10, 20, 40 and 60 wt.%) using a pin-on-disk test system. The specific wear rate increased with lead content up to 40 wt.% after which a drop in wear rate was observed. No correlation existed between the hardness and sliding wear of these alloys. A study of wear mechanisms shows that oxidative wear is the primary material removal mechanism at low contact pressures, while plastic deformation and adhesion are the rate-controlling mechanisms at high contact pressures.

1. Introduction In an earlier paper [I], we described the usefulness of non-equilibrium Cu-Pb alloys for various applications such as bearings. Another major area that has been proposed recently for the use of these alloys is bearing cages for space-based systems where extended life is necessary. In this application the primary requirement is sustained lubrication over periods up to lo-15 years. Since sliding wear is an important criterion for determination of such extended life, in this paper we report the sliding wear aspects of nonequilibrium alloys. Although energy is lost in overcoming friction, the most damaging effect is wear. Wear eventually leads to breakdown and loss of machines and equipment. Several models [2] have been proposed to describe the wear process occurring at the sliding surfaces in contact. Generally, the main models encountered in industrial situations include adhesion, abrasion, delamination, and erosion. The most widely used theory for predicting wear is the adhesive wear theory [ 31. Adhesive wear occurs as a result of relative sliding between two contacting surfaces under normal load. When asperities come into contact, the pressure is sufficient to cause local plastic deformation and adhesion. As the surfaces move apart, the junctions will break and particles may detach. The coefficients of friction and wear are largely dependent *Present address: College of Arts, Science and Technology, Kingston, Jamaica.

0043-1648/91/$3.50

0 Elsevier Sequoia/Printed in The Netherlands

258

on the nature of the interacting surfaces and the characteristics of the materials. Experiments have shown that compatibility of the sliding materials 641, cleanliness of the surface [ 51, crystal structure, cohesive strength [ 61, sliding temperature [ 61, and hardness change the friction and wear behavior. Archard’s wear equation [3] yields the relationship where V is the wear volume, K is a wear constant, L is the normal load, S is the sliding distance, and H is the hardness of the softer of the pair of materials. In a study of iron-lead alloys, Shahparast and Davies [ 71 reported that an iron-lead alloy, without the addition of other alloying elements, reduced both friction and wear. In addition, the alloy wear track was very smooth, whereas the track formed by pure iron displayed considerable surface damage. Alloying of copper with gray iron for cylinder liners of engines has also been observed to reduce wear intensity [ 81. This was attributed to the presence of a film of copper that diffused to the surface and promoted by elevated temperature caused by friction and the Rebinder effect (surface softening of the iron alloy because of the presence of surface-active substances in the oil). Rogers [9] also reported that scuffing in engines may be prevented by the use of copper or copper-plated rings, because copper inhibits the fo~ation of carbides. The wear behavior of Cu-Pb alloys in bearing applications has been studied [ 10, 111, but these investigations were mainly related to the determination of the wear mechanism and deposition techniques when these alloys were used as overlays, thin films, or in powdered form. For example, Krol [lo] investigated the effect of powdered Cu-30%Pb and lithium grease as a lubricant on steel bearing surfaces. He found that both friction and wear were reduced. This was at~buted to the copper deposition on the steel bearings. In another study on the surface damage of bearing alloys caused by the joint action of cavitation erosion and sliding wear, Okada and Iwai [ 111 found that wear of Cu-30%Pb is restrained under low cavitation intensity, but accelerated wear loss results at high cavitation intensity. It was observed that the resistance to damage at low intensity was dominated by the protective ability of a Iayer of soft component in the alloy on the surface subjected to cavitation damage. An extensive search of the literature revealed that no data on the wear resistance of Cu-Pb alloys containing more than 40% lead are avaiIable. Hence this work is directed toward determining wear resistance and identifying the wear mechanisms of such high-lead-containing Cu-Pb alloys under sliding conditions. Compositions are given in weight percent. 2. Experimental

details

The details of the friction and wear test rig, and the preparation of alloys, have been described earlier [ 11. The worn pins, the wear debris, and

259

the track surface were examined in a 20 kV scanning electron microscope (JEOL 840-A) to identify the mechanisms of material removal and to facilitate the interpretation of test results. A number of scanning electron micrographs were taken to demonstrate the most important features. The sliding direction is indicated by an arrow in these micrographs.

3. Results and discussion The dry wear characteristics of the tested materials are presented graphically in Pig. 1, where the weight loss of the pin is plotted as a function of sliding distance. With the exception of hardness values, the results confirm the trends predicted by Archard’s wear equation. Specifically, the wear volume increased with either an increase in sliding distance or an increase in load. However, the observed wear characteristics were not directly related to the hardness because of the extrude,d lead particles coalescing around the edges of the pin. High wear, with the exception of pure copper, was observed in Cu-Pb alloys for the test conditions used. In all cases, there was no evidence

,“:::c 0.09

0 0 o v

0.14 0.29 0.42 0.56

MPI MPa MPa MPI

0.06 y

4

0.07 0.06

il

0.05

iz

0.04 0.03 0.02 0.01 0.00

If-b 2

!Z 4

c

; 6

0

Sliding Distance, km

(a>

0 0.42 D 0.56

m Ia-

0.25

4

0.20

2 9u

-

MPa MPa

0.15 0.10

0.00 0 @I

2

4

6

Sliding Distance, km

(continued)

260

Sliding Distance, km

0

2

4

6

Sliding Distance, km Cd)

Fig. 1. Wear

loss as a function

of sliding

distance:

(a) Cu; (b) Cu-20%Pb;

(c) Cu-40%Pb;

(d) Cu-GO%Pb.

of “running-in” wear in the initial stages. No attempts were made to evaluate the compositions of wear debris. Since the wear of the materials had a linear relationship with sliding distance, steady-state wear was established. The wear rates, in grams per meter, were calculated and plotted as a function of pressure and lead content. These wear rates are shown in Fig. 2. Although the wear rate increased with load, a transition stage can be seen at high loads. It is of interest to note that reduction in the coefficient of friction occurred at the same transition stage (see results in ref. 2). This may be attributed to the contribution from the size and volume fraction of lead distributed in the matrix. Figure 2 also indicates that pure copper exhibited a low wear rate and was relatively unaffected by the load. Figure 3 shows the variation of the average coefficient of friction with specific wear rate. Although a decisive relationship between friction and wear is generally not obtained, possibly because of the presence of surface contaminants, the data suggest some dependence of the wear on the coefficient of friction. For pure copper, an increase in wear rate reduces the friction.

261

Load,MPa Fig. 2. Wear rate as a function of load.

0 A o D

cu Cu-ZO%Pb CudO%Pb Cu-60%Pb

Specific Wear Rate

Fig. 3. Effect of the coeflicient of friction on the specific wear rate (mm3 Nm-’ X lo-‘).

The stable wear debris appears to reduce the friction significantIy. For Cu-Pb alloys, the trend is more or less opposite. Increased wear rate increased the friction. This may be attributed to the removal of unstable lead debris during the wear process. The wear resistance of the alloys is si~ific~tly affected by the addition of lead. At pressures of 0.28-0.56 MPa, Cu-40%Pb undergoes a higher wear rate than Cu-GO%Pb, even though the latter has a lower hardness value (see ref. 1). Examination of the wear pins showed excessive upsetting of the contacting surface of the Cu-GO%Pb pin with the formation of curIed extrusions around the edges (Fig. 4). The profile of the pin edge is mostly due to localized melting wear that occurs during the extrusion of the transfer material and results in agglomerations. Such agglomeration of particles (Fig. 5) was frequently observed in the wear debris. Instead of weight loss, a me~~ement of height loss indicates a higher wear rate as the percentage of lead increases in all cases. In Table 1, the specific wear rates of the pins at different loads are summarized. It can be seen that the specific wear rates are maximum in aII

Fig. 4. Formation

of curled cxtrus~ons

Fig. 5. Typical aggregate

of particles

around

edges of pin.

by adhesion.

TABLE 1 Spccifk

wear rate (mm” Nm--‘X 10m5) for metals sliding in contact

Metal

Cu Cu-2O%Pb Cu-4O%Pb Cu-60%Pb

at different

loads

Load (MPa) 0.14

0.28

0.42

0.56

1 60 95 97

14 37 76 56

13 45 77 52

14 42 77 52

cases for the Cu-Pb alloys at 0.14 MPa but remain fairly constant at all other loads. This confirms earlier results [ 121 that smooth surfaces under low loads tend to wear faster than at high loads. When the non-dimensional Archard wear coefficient [ 31 k was computed, for Cu-Pb alloys, it was in the range 6.0 X lo-” to 1.0X 10B3, while for copper it was in the range 3.0X lop5 to 4.0X 10w4. The value obtained by Rabinowicz [ 131 for Cu-16%Pb closely agrees with the findings of this study, but the coefficient for copper was much lower in this test. 3.1. Wear analysis The sliding wear process of Cu-Pb alloys is associated with a metallic film transfer from the alloy to steel, owing to the adhesion mechanism. In this test, a layer of transfer material from the pin surface to the steel disk was established in a relativefy short period and remained strongly attached to the disk. The track surface indicated good adhesion between steel and Cu-Pb as seen by the transfer material completely masking the disk surface

263

locations (Fig. 6). This prevented direct contact between pin and disk. As a result, the wear and frictional behavior depended on the mechanism between the pin and transfer layer with the disk mainly acting as support for the load. Most of the transfer film in the high-lead-containing alloys possessed a dark grayish color, which indicated subsequent oxidation because of the high flash temperature at the sliding interface. The surface morphology of the pins, after sliding against the steel, showed diierent topographical features. Further examination also revealed that each surface was subjected simultaneously to more than one mode of material removal. Cu-Pb alloys exhibited a mixed abrasive-plastic deformation mechanism (Figs. 7, 8). Abrasion is evidenced by the fine scoring grooves visible in the sliding direction, as shown in Fig. 7. The scored grooves are presumably due to the action of entrapped oxide particles or wear-hardened deposits on the disk track. However, this mechanism does not appear to be the dominant mode of material removal because the material removed from the fine grooves would be small and more likely to be curled. Flat ridges of deformed metal proliferated on the Cu-Pb surface were generally found to extend from one end of the pin surface to the other (Fig. 8). Higher magnification of the flat ridges revealed extensive areas of plastic deformation with cracks being formed (Figs. 9, 10). In Pig. 9, a particle is in the process of detaching from the surface to form wear debris that may either become entrapped and cause abrasion, be transferred to the disk, or adhere to the edge of the pin, forming a large composite particle (Fig. 5) before dropping off. The visible cracks may have been initiated below the surface by a combination of fatigue and plastic flow. As more cracks intersected, the process of adhesion and deformation work hardened the interface so that the shear strength of the deformed area increased sufficiently to result in shear fracture occurring in the bulk material of the pm. in most

Fig. 6. Typical diik covered with transfer material.

Fig. 7. Fine wear grooves caused due to abrasion. Fig. 8. Flat ridges of deformed metal observed in all tests.

Fig. 9. Particle in the process of detaching from the surface to form wear debris. Fig. 10. Plastic flow and fracture to form debris.

Figure 11 shows a micrograph of the typical debris of the Cu-Pb ahoys recovered after the wear experiment. Two types of wear particles are evident: large, irregular laminar particles (Pig. 11(a)) and small equiaxed particles (Fig. 11 (b)). The small equiaxed particles are frequently found adhering to the larger particles, forming an agglomeration. Although it has been reported [14] that wear particle size could be used as an index of wear, there is insufficient evidence from this test to support such a statement. Figure 12 shows a large, plate-shaped particle that appears to be partially fractured with very small particles clinging to the surface. Its surface characteristics

265

Fig. 11. Typical wear debris recovered from test specimens.

Fig. 12. Wear particle Fig. 13. Oxidative

demonst~t~g

wear associated

repeated with plastic

transfer. deformation.

suggest that the particle, after being detached from the pm, was transferred to the disk. Subsequent passes of the pin resulted in more transferred particles that were compressed to produce layers of shear tongues visible on the particle. When the film reached a critical thickness, subsurface cracking and shearing of the laminate particle eventually took place. Oxidative wear (Figs. 13, 14) appears to be a major wear mechanism for pure copper and Cu-Pb alloys at low contact pressures. A layer of oxide interspersed with an oxide-free surface was clearly visible (Pigs. 14, 15). Figure 15 also indicates some evidence of plastic defo~ation in areas, free of oxides. The low specific wear rate obtained for all test specimens at

266

Fig. 14. Oxidativc debris attached to the surface of the pin. Fig. 15. Fatigue cracks on oxide-coated surface.

0.14 MPa is indicative of the presence of oxides at the sliding interface. At high pressures and with increased bulk temperature, which softens the surface of the pin, oxidative wear becomes less important since disruption of the protective film is easily accomplished. In such cases, adhesion and plastic deformation can become the dominant wear mechanisms. Tomlinson and Foulkes [ 151 reported that the early stages of wear of copper on dry nitrided steel occurred mainly by abrasion that resulted in extensive ploughing on the wear surface of copper. Rabinowicz and Tabor [ 161 observed that when a copper pin slid on a mild steel disk there was more than three times the metal transfer than for a steel pin sliding on a copper disk. It has also been shown that the choice of material for the pin VS. disk can influence the outcome of the results [ 171. 4. Conclusions This study has demonstrated the following. (1) The influence of lead content on the wear characteristics is a critical function of the size, shape and volume fraction of lead distributed in the matrix. Friction and wear phenomena at the interface during actual sliding are difficult to observe or detect. (2) Further, wear is not a materials property but depends on many influencing parameters such as load, speed, temperature, or en~onment, which may interrelate with each other. This is substantiated by the observation that the wear rate is increased as the pressure is increased for all the compositions studied in this work. (3) Increased wear rates increased the friction in Cu-Pb alloys. This was rationalized by the removal of the unstable lead debris during the wear process.

267

(4) Oxidative wear is the dominant mode at low contact pressures, while adhesion and plastic deformation appear to dominate at higher pressures.

Acknowledgments

The authors would like to thank Applied Metals Technology Inc., Naples, FL, for the supply of the material used in these studies. We also acknowledge the Engineering Research Institute of Iowa State University for assistance in the preparation of this document.

References 1 V. Buchanan, P. A. Molian, T. S. Sudarshan and A. Akers, Frictional behavior of nonequilibrium Cu-Pb alloys, Wear, 146 (1991) 241-256. 2 D. A. Rigney (ed.), Fundam.entals of Friction and Wear, 1980 ASM Materials Science Seminar, American Society for Metals, Metals Park, Ohio, 1981. 3 J. F. Archard, Contact and rubbing of flat surfaces, J. Appl. Phys., 24 (1953) 981-988. 4 H. Ernst and M. E. Merchant, Surface friction between metals-a basic factor in the metal cutting process, Proc. Special Summer Con$ onFriction and Surface Finish, Massachusetts Institute of Technology, Cambridge, MA, 1940. 5 F. P. Bowden and D. Tabor, The Friction and Lubrication of Solids, Clarendon, Oxford, 1950. 6 E. Rabinowicz, Friction and Wear of Materials, Wiley, New York, 1965. 7 F. Shahparast and L. B. Davies, Astudy of the potential of sintered iron-lead and iron-lead-tin alloys as bearing materials, Wear, 50 (1978) 145-153. 8 M. A. Medvedev and V. A. Komarov, Influence of copper on the wear resistance of gray cast irons, Sov. J. Frict. Wear, 1 (5) (1980) 128-130. 9 M. Rogers, The use of copper to prevent scuffing in diesel engines, Wear, 22 (1972) 245-257. 10 M. Krol, Studies of frictional copper plating of steel surfaces with copper-containing lubricants, Trybologia, 17 (3) (1986) 18-20. 11 T. Okada and Y. Iwai, Resistance to the damage caused by the joint action of cavitation erosion and sliding wear of bearings alloys, Lulrr. Eng., 44 (1) (1988) 61-68. 12 S. Jahanmir and N. P. Suh, Wear topography and integrity effects on sliding distance, Wear, 44 (1977) 87-89. 13 E. Rabinowicz, Friction and wear of self lubricating metallic materials, J. L&-T-. Technol., 97 (1975) 217-220, 249. 14 E. Rabinowicz, A qualitative study of the wear process, Proc. Phys. Sot. Lmukm, Ser. B, 66 (1953) 929-936. 15 W. J. Tomlinson and S. N. Foulkes, Early stages of wear of copper, steel, and electroless nickel against a nitrided steel surface, Tribal. Int., 22 (3) (1988) 149-153. 16 E. Rabinowicz and D. Tabor, Metallic transfer between sliding metals: an autoradiographic study, Proc. R. Sot. London, Ser. A, 208 (1951) 455-475. 17 C. L. Goodzeit, The seizure of metal pairs during boundary lubrication, Proc. Symp. on Friction and Wear, Elsevier, Amsterdam, 1957, p. 67.