Microstructure–Wear Resistance Correlation and Wear Mechanisms of Spark Plasma Sintered Cu-Pb Nanocomposites AMIT SIDDHARTH SHARMA, KRISHANU BISWAS, and BIKRAMJIT BASU The dispersion of a softer phase in a metallic matrix reduces the coefficient of friction (COF), often at the expense of an increased wear rate at the tribological contact. To address this issue, unlubricated fretting wear tests were performed on spark plasma sintered Cu-Pb nanocomposites against bearing steel. The sintering temperature and the Pb content as well as the fretting parameters were judiciously selected and varied to investigate the role of microstructure (grain size, second-phase content) on the wear resistance properties of Cu-Pb nanocomposites. A combination of the lowest wear rate (~1.5 9 106 mm3/Nm) and a modest COF (~0.4) was achieved for Cu-15 wt pct Pb nanocomposites. The lower wear rate of Cu-Pb nanocomposites with respect to unreinforced Cu is attributed to high hardness (~2 to 3.5 GPa) of the matrix, Cu2O/Fe2O3-rich oxide layer formation at tribological interface, and exuding of softer Pb particles. The wear properties are discussed in reference to the characteristics of transfer layer on worn surface as well as subsurface damage probed using focused ion beam microscopy. Interestingly, the flash temperature has been found to have insignificant effect on the observed oxidative wear, and alternative mechanisms are proposed. Importantly, the wear resistance properties of the nanocomposites reveal a weak Hall–Petch-like relationship with grain size of nanocrystalline Cu. DOI: 10.1007/s11661-013-1965-7  The Minerals, Metals & Materials Society and ASM International 2013

I.

INTRODUCTION

A tribosystem is a complex ensemble of subsystems classified on the basis of structural, operational, and interaction parameters. The structural parameters comprise the two participating triboelements, interfacial element and the environment. Similarly, operational parameters include the type of motion, applied load, velocity, test duration, etc., whereas interaction parameters involve interfacial forces, flash temperature, contact deformation modes, and stresses.[1] While assessing the performance of a tribosystem, the determination of the interaction parameters based on a set of operational parameters is quite straightforward, but on the contrary, it is the structural parameters that need to be carefully tailored to achieve a better combination of frictional and wear resistance properties. The structural parameters can be further diversified as chemical (type and amount of second phase), physical (thermal conductivity, etc.), mechanical (elastic modulus, hardness, etc.), surface topography, and microstructural (grain size, AMIT SIDDHARTH SHARMA, formerly Ph.D. Scholar, with the Department of Materials Science and Engineering, Indian Institute of Technology Kanpur, Kanpur 208016, UP, India, and also Visiting Researcher, with the Indian Institute of Science Bangalore, Bangalore 560012, Karnataka, India, is now Postdoctoral Researcher, with the Indian Institute of Science Bangalore. KRISHANU BISWAS, Associate Professor, is with the Department of Materials Science and Engineering, Indian Institute of Technology Kanpur. BIKRAMJIT BASU, Associate Professor, is with the Materials Research Center, Indian Institute of Science Bangalore. Contact e-mail: [email protected] Manuscript submitted March 31, 2013. Article published online September 5, 2013 482—VOLUME 45A, JANUARY 2014

dislocation density, etc.). From the point of view of alloy development, the structural parameters are important and the role of some of these parameters on the tribological properties of Cu-Pb system is the central theme of the current investigation. The tribological properties largely determine the performance and life of the bearing alloys. The general attributes of a bearing material are low coefficient of friction (COF) against hard materials (embeddability), ability to adapt against roughness and misalignment (conformability), reasonable hardness, and low shear strength.[1] If intended for use in a corrosive environment, then the material should possess reasonable resistance to corrosion. The metallic composites consisting of a hard matrix with one or more soft phases are considered as an important category of materials for tribological applications.[1,2] Various Cu-based bronzes, an important family of bearing alloys, are being developed with manganese, aluminum, lead, and tin as the principal alloying element. Among these alloying elements, manganese and aluminum addition imparts high strength, excellent corrosion resistance, and heat treatability, whereas Pb addition results in superior ductility, good machinability, anti-seizure, and antifrictional properties. The self-lubrication property of lead is another major advantage in self-lubricating bearings. However, the major disadvantage of using Pb stems from its inherent softness and low melting point [~600 K (~327 C)]. Therefore, the addition of a higher amount of Pb to a metallic matrix may lead to a decrease in hardness. The high lead-containing (Pb content >7 wt pct) tin bronzes, C93XXX series, can be used for general utility applications under medium loads METALLURGICAL AND MATERIALS TRANSACTIONS A

and speeds, while the aluminum-based bronzes, C95XXX series, are known as the strongest among Cu-based bearing alloys with yield and tensile strengths close to 0.5 and 0.8 GPa, respectively.[1,3] Another class of Cu-based alloys, viz., brasses with zinc as the principal alloying element has been reported to show a reduction in anti-frictional properties if zinc in excess of 4 pct is added to Cu. Therefore, the alloying addition needs to be tailored to obtain a better combination of mechanical and tribological properties. From the materials science perspective, the improvement in wear properties (quantified in terms of wear resistance) can therefore be engineered by careful alloy design approach. To this end, the optimization of the alloying additions specific for a particular application or the optimization of the manufacturing process needs to be pursued to obtain desired surface and bulk properties of the material.[4] While an increased hardness leads to a better abrasive/adhesive wear resistance, the amount and uniform distribution of softer phase, i.e., Pb, will result in superior frictional properties. It is also worthwhile to mention that conventionally processed Cu-based bronzes lack good tribological properties due to low hardness owing to the coarse-grained microstructure and the inhomogeneous distribution of the second phase. One obvious route toward harnessing superior hardness in metallic alloys is to reduce the grain size, whereas the amount of softer phase must also be optimized with uniform and homogeneous distribution in the matrix. In this perspective, the present work will demonstrate how the spark plasma sintering (SPS) route can be adopted to develop Cu-Pb nanocomposites with a better combination of tribological properties. In an earlier work,[5] the present authors had reported the optimization of SPS parameters, microstructure, and hardness of the nanocrystalline Cu and Cu-10 wt pct Pb composites. The microstructural studies revealed that it is possible to retain nanocrystalline Cu grains with fine-scale homogeneous distribution of Pb, when processed at 623 K (350 C) at 100 MPa pressure as shown in Figure 1.[5] In the present work, the fretting wear of the nanocrystalline Cu-Pb alloys with varying Pb contents is reported and discussed. Cu-Pb alloys find extensive use in automotive, aircraft, and general engineering applications. For example, high lead (with 30 pct Pb) SAE 48 alloy can be sintered and plated on a steel backing to be used as trimetal bearings in diesel engines requiring sufficient fatigue strength and seizure resistance.[1,6] As fretting is characterized by a low-amplitude oscillatory motion at low loads/sliding velocities, the intended application for these Cu-Pb alloys will be in locations near highways and small industrial equipment, where similar operating conditions are anticipated. Though fretting studies on Cu-based alloys are not available in the open literature, extensive sliding wear studies have been carried out on Cu-Pb[7–10] alloys as well as Pb-containing bronzes.[11,12] In the present study, an attempt has been made to correlate the microstructure with properties, such as hardness, friction, and wear, with a particular emphasis on the influence of Pb content/distribution and grain size of the Cu matrix. METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 1—(a) BF-TEM micrograph of Cu-12.5 wt pct Pb sample processed at sintering temperature of 623 K (350 C); (b) Cu grain size distribution. Arrows mark the Pb position in the matrix.

Keeping this aspect in mind, 10, 12.5, and 15 wt pct Pb was incorporated in the Cu matrix to strike a balance between the mechanical and tribological properties, and SPS parameters were varied to probe the influence of grain size. One of the objectives of the present study is to qualitatively deconvolute the obtained tribological properties with respect to Pb content, hardness, relative density, and formation of mechanically mixed tribolayers during fretting wear. Importantly, the subsurface damage is investigated.

II.

EXPERIMENTAL PROCEDURE

A. Materials For the composite preparation, the commercially available high purity nanocrystalline Cu powder (mean VOLUME 45A, JANUARY 2014—483

particle size ~97 ± 25 nm, 99.8 pct purity, SigmaAldrich) and microcrystalline Pb powder (mean particle size ~149 lm, 99.98 pct purity, Sigma-Aldrich) were used as the starting materials. Utmost care was taken at each step of processing to minimize oxidation of nanocrystalline Cu powder. B. Processing The powder mix of Cu and Pb in the ratio of 90:10, 87.5:12.5, 85:15, and 80:20 (by weight) was mechanically mixed in a high energy ball mill, using toluene as milling medium. The milling operation of the pre-mixed powders was performed in a planetary micro-mill (Fritsch Pulverisette 7) for 16 hours using agate (SiO2) balls in an agate vial at 200 rpm. The toluene level in the vial was periodically monitored and maintained in such a way that the powder mixture was completely immersed in toluene throughout the milling operation. The powder slurry for each composition was then filtered and dried in an oven at about 80 C to obtain homogenously mixed powder. SPS using Dr. Sinter 515S apparatus (SPS Syntex Inc., Kanagawa, Japan) was carried out with Cu-Pb compositions at 573 K, 623 K, 673 K, and 773 K (300 C, 350 C, 400 C, and 500 C) with a uniaxial pressure of 100 MPa. The powder mix after loading in a cylindrical graphite die was heated with the heating rate of 80 K/min to the sintering temperature and held at the sintering temperature for 5 minutes, before it was furnace cooled to room temperature. An inert atmosphere was maintained by continuous purging of high purity argon throughout the processing cycle in the sintering chamber with a flow rate of 2 L/min to insure minimal oxidation of the powders. C. Characterization 1. Density determination The density of all SPS-processed pellets was determined according to Archimedes’ principle using distilled water (Sartorius CPA 225D, Germany). The theoretical density for the sintered composition was calculated following the rule of mixtures, considering the theoretical densities of pure Cu and Pb as 8.96 and 11.68 g/cc, respectively. 2. Hardness measurement Vickers hardness of the sintered samples was measured using a Vickers microhardness tester (Bareiss Pru¨fgera¨tebau, Germany) by applying 200-g load and a dwell time of 10 seconds. The diagonals of Vickers indents were then measured using an optical microscope. At least 5 indentations for each sample were taken to report the average value with error bars, indicating the standard deviation. 3. Fretting wear test The tribological behavior of the as-sintered Cu-Pb composites was studied using a fretting tribometer (DUCOM TR281M, India), employing a ball-on-flattype configuration. Among the three modes of fretting, Mode I, which is characterized by a linear relative 484—VOLUME 45A, JANUARY 2014

tangential displacement of small amplitude between two contacting surfaces at a constant normal load, is commonly used for a laboratory-scale study.[2] Using a stepper motor, the flat sample was made to oscillate with a relative displacement of constant linear stroke (amplitude) on either side of the mean position of contact between the mating bodies with a specified frequency. A piezoelectric transducer then senses the changes in tangential frictional force and the corresponding COF was calculated by dividing tangential force with the applied normal force. The variation in COF can then be plotted with respect to number of fretting cycles or total time. For all the experiments, both the flat sample and the counterbody steel ball (Grade: SAE 52100, diameter = 10 mm, Hv ~8.3 GPa) were subjected to ultrasonic cleaning at a set frequency of 33 kHz for 30 minutes in acetone to get a reasonably clean surface, devoid of any grease or dirt. In order to investigate the wear behavior, the wear tests were performed at two different normal loads, namely, 5 and 10 N. All the tests were performed for 60,000 cycles with a relative displacement stroke between the flat and ball set to 100 lm and a frequency of 5 Hz. The temperature and the relative humidity during the wear tests were maintained at 37 ± 1 C and 40 ± 5 pct, respectively, by conducting the fretting wear tests in an environmental chamber. For good experimental repeatability, the fretting tests were iterated five times for each sample. 4. Quantification of wear parameters and examination of microstructure, wear surfaces, subsurface damage, and debris particles a. Profilometry. The geometry, i.e., extent and depth, of the wear scars was mapped using a laser surface profilometer (Perthometer PGK 120, Mahr, Germany). In this non-destructive profiling technique, an infra-red laser beam of wavelength k = 780 nm formed into a circular beam of 2 lm diameter was used to trace the dimensions of the wear scar. After completion of fretting tests, the wear debris was collected with the help of a carbon tape and then each sample was again ultrasonically cleaned to remove the wear debris particles. Several 2D profiles were taken over the entire worn surface. The area under each 2D profile was calculated and the volume of the scar region was determined by integrating the profile area of each of the 2D profiles over the distance. From the calculated wear volume, the wear rate was calculated as follows: Wear rate ¼

Wear Volume : ðLoad  No: of cycles  Stroke lengthÞ ½1

b. Electron microscopy. After the wear tests, the detailed topographical features of the wear scars were investigated using a scanning electron microscope (SEM) (Carl Zeiss SUPRA 40VP, Germany) at an METALLURGICAL AND MATERIALS TRANSACTIONS A

accelerating voltage of 20 kV. The wear debris particles formed were collected with the help of a carbon tape by slightly touching it onto the wear scar. The elemental analysis of the tribolayer formed on the sample, the steel counterbody, and wear debris was conducted with the help of an energy dispersive X-ray spectroscopy (EDS) (Oxford Instruments) module, integrated with the SEM. The finer-scale microstructures were investigated using a transmission electron microscope (TEM) (FEI UT 20) with an accelerating voltage of 200 kV. c. Ion beam microscopy. Subsurface features underneath the worn surface of the samples were investigated by focused ion beam (FIB) microscopy. The cross-sectional wear samples were prepared by utilizing an ion beam. The FIB apparatus used was FEI Strada 201, which has both electron and ion columns. A beam of gallium ions, Ga+, operating at 30 kV and 6600 pA beam current, was used to mill trenches of ~20 9 15-lm-wide and 10-lm-deep area at the center of the worn surface. The polishing of the milled surface was carried out in two steps of 2 hours duration each—the first step at a beam current of 1030 pA, followed by gentle polishing at a much lower beam current of 390 pA. To avoid milling artifacts such as curtaining while imaging, the beam current was further reduced to 85 pA with a scan time of ~22 seconds. To ascertain the topographical damage owing to the wear process, the imaging was carried out in back scattered (BSE) as well as secondary electron (SE) mode. d. Spectroscopy. The compositional determination of the wear debris particles, formed as a result of the wear process, was carried out by using a micro-Raman spectroscope (Jobin-Yvon RMS-550 micro-Raman system), employing an infra-red argon laser of wavelength k = 514 nm. The laser power on the sample was low enough (6.12 mW) to induce any further phase transitions in the wear debris, during acquisition of spectra. The particles were scanned in the wavenumber range of 80 to 1350 cm1 to obtain the probable traces of any oxide formed during the course of the fretting wear process.

III.

RESULTS

A. Microstructural Characterization In order to illustrate the finer-scale microstructural features of the investigated Cu-Pb alloys, transmission Table I.

electron microscopy (TEM) has been carried out. Figure 1(a) shows a typical bright field (TEM) micrograph of the spark plasma sintered Cu-12.5 wt pct Pb sample sintered at the temperature of 623 K (350 C) with an applied pressure of 100 MPa. The Pb particles are marked by black arrows on the micrograph. The observation of a distinct Pb phase in the vicinity of the Cu grain boundaries is consistent with the immiscibility of Cu and Pb at 623 K (350 C).[13] Detailed microstructural observations for Cu-10 wt pct Pb nanocomposite have been reported in an earlier work along with an in-depth analysis of temperature-dependent variation in relative density of sintered composites.[5] Table I provides the relative density values for Cu-Pb nanocomposites (Pb wt pct = 0, 10, 12.5, 15, and 20) at sintering temperatures of 573 K, 623 K, 673 K, 773 K, and 873 K (300 C, 350 C, 400 C, 500 C, and 600 C). Figure 1(b) shows the measured grain size distribution of the Cu. It clearly indicates that a maximum number of investigated grains lie in the 50to 100-nm size range with a mean of 61 ± 26 nm. The extensive bright field TEM analysis was carried out and the grain size/distribution has been obtained for Cu10 wt pct Pb[5] and Cu-15 wt pct Pb nanocomposites. The results indicate that it is possible to retain the nanometric Cu grains by sintering at 623 K to 673 K (350 C to 400 C) at an applied pressure of 100 MPa. B. Hardness Data Figure 2 shows the hardness of the spark plasma sintered Cu-Pb composites with varying Pb contents and SPS temperatures. Pb being softer than Cu, the addition of Pb in Cu will have a degrading effect on the hardness of the composite, and therefore Pb contents of 10, 12.5, 15, and 20 wt pct have been selected to optimize the hardness values. The sintering temperature has been varied in the range of 573 K to 673 K (300 C to 400 C). To retain the intrinsic grain size in the nanocrystalline range, the upper limit of the sintering temperature has been restricted to 673 K (400 C). As reported in a previous study,[5] the grains grow abruptly at 723 K (450 C) to a microcrystalline regime, leading to a drop in the hardness values. Therefore, the investigated Cu-Pb alloys have been sintered in the temperature range of 573 K to 673 K (300 C to 400 C). As evident from Figure 2, the hardness shows a peak value at 623 K (350 C) for all compositions, except Cu-10 wt pct Pb. Moreover, the highest hardness

Relative Density of Sintered Cu-Pb Samples Using SPS as a Function of Sintering Temperature at a Pressure of 100 MPa Relative Density (Pct qtheoretical)

SPS Temp. [K (C)]

0 Wt Pct Pb

10 Wt Pct Pb

573 623 673 773 873

74.08 78.85 87.11 85.25 84.06

86.20 88.35 86.32 85.23 76.89

(300) (350) (400) (500) (600)

± ± ± ± ±

0.02 0.02 0.03 0.02 0.04

METALLURGICAL AND MATERIALS TRANSACTIONS A

± ± ± ± ±

0.06 0.05 0.10 0.05 0.04

12.5 Wt Pct Pb

15 Wt Pct Pb

20 Wt Pct Pb

72.61 ± 3.07 70.74 ± 1.99 76.42 ± 2.26 — —

62.87 ± 3.23 68.52 ± 2.96 68.70 ± 3.92 — —

81.90 ± 2.19 84.59 ± 4.36 — — —

VOLUME 45A, JANUARY 2014—485

Fig. 2—Hardness plot for Cu-Pb composites as a function of SPS temperature. SPS conditions: 100 MPa pressure, argon atmosphere. Hardness measurement conditions: Vickers hardness done at a load of 200 g and a dwell time of 10 s.

value is obtained for the Cu-15 wt pct Pb composite (average value ~3.5 GPa). Compared to the studies carried out on pure Cu samples by other investigators,[14,15] the hardness achieved in the present investigation for nanocrystalline Cu-Pb alloys was found to be higher with the inclusion of softer Pb. Zhang et al.[14,15] studied the surface hardness of pure Cu samples consolidated by surface mechanical attrition treatment (SMAT). They modified the surface layers of pure Cu by deforming at very high strain rates for 5 and 30 minutes in order to form deformation nanotwins in Cu. Though the average grain size was ~10 nm, the microhardness values achieved were close to 1.3 GPa. Srivatsan et al.[16] reported the hardness variation of the bulk Cu samples prepared by the plasma pressure compaction technique. The average values of microhardness and nanohardness were reported to be 1.9 and 2.4 GPa, respectively, for samples sintered at 1153 K (880 C) with 50 MPa pressure. In this backdrop, the achieved hardness can be considered as a significant improvement for Cu-15 wt pct Pb composites vis-a-vis Pb content and lower sintering temperature. C. Frictional Behavior The frictional behavior of the Cu-Pb composites has been studied and the results are analyzed in terms of COF variation with fretting cycles. The fretting load was selected keeping in mind the ductile reinforcement (Pb) in the composite. Figures 3(a) and (b) presents the COF evolution of the Cu-12.5 and 15 wt pct Pb composites, respectively, for a sufficiently large number of fretting cycles (60,000). It is well known that due to surface asperity contact, COF reaches a high value during the running-in period of up to 2000 cycles. During this period of fretting, hard asperities are worn away and smooth contact results in steady state friction. All CuPb (Pb wt pct = 10, 12.5, and 15) composites exhibit a 486—VOLUME 45A, JANUARY 2014

Fig. 3—The frictional behavior of (a) Cu-12.5 wt pct Pb and (b) Cu15 wt pct Pb composites, SPS at 573 K (300 C). Fretting conditions: 5 and 10 N normal load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel.

consistent COF value during steady state without any considerable fluctuation after the running-in period. From Figures 3(a) and (b), it can be seen that steady state COF increases as the load is increased from 5 to 10 N. It is also evident that in the case of samples sintered at 623 K (350 C), the average COF is higher than that of samples sintered at 573 K (300 C). This is true for both 5 and 10 N load. Figures 4(a) and (b) plot the average COF values with varying Pb contents (10, 12.5, and 15 wt pct) in the CuPb composite at the fretting loads of 5 and 10 N, respectively. It can be seen that in both the cases of the applied normal loads, the average COF values are greater for the samples sintered at 623 K (350 C) as compared to 573 K (300 C). D. Wear Data Table II summarizes the friction and wear data for all Cu-Pb (Pb wt pct = 10, 12.5, and 15) composites with respect to Pb content, SPS temperature, and fretting load. Figure 5(a) shows the variation of wear rate with Pb content. The overall wear rates for all the composites METALLURGICAL AND MATERIALS TRANSACTIONS A

Figure 5(b) shows the variation of wear resistance (defined as the inverse of the wear rate) with the grain size of nanocrystalline Cu matrix in the case of selected samples (Cu-10 wt pct Pb, Cu-12.5 wt pct Pb, and Cu15 wt pct Pb nanocomposites, all spark plasma sintered at 573 K and 623 K (300 C and 350 C)). Similar to the Hall–Petch expression which relates strength with the grain size, the data points in Figure 5(b) reveal a weak Hall–Petch-like relationship for wear resistance of CuPb nanocomposites. It can therefore be stated that the wear resistance of nanocrystalline Cu-Pb composites is influenced by the grain size, and the nanometric grains can result in higher wear resistance. A critical analysis of the COF, wear rate, and wear resistance in reference to microstructure and properties as well as the earlier literature reports will be provided in the discussion section. Also, it should be noted that the fretting wear tests have been conducted on areas of the samples which are pore free so that consistent wear data can be obtained. In fact, the microhardness measurements have also been carried out in a similar fashion so that consistency in hardness measurements can be achieved. E. Morphology of Worn Surfaces

Fig. 4—Variation of coefficient of friction (COF) of Cu-Pb composites plotted as a function of Pb content. Fretting conditions: (a) 5 and (b) 10 N load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel.

during fretting ranges from 1 9 106 to 6 3 210 9 10 mm /Nm. For Cu-10 wt pct Pb sample, an average wear rate of samples sintered at 623 K (350 C) shows higher wear rate than samples sintered at 573 K (300 C). For 573 K (300 C) sintered samples, the calculated wear rate for both the 5 and 10 N loads is ~15 9 106 and ~32 9 106 mm3/Nm, respectively, whereas in case of samples sintered at 623 K (350 C), the wear rate for 5 and 10 N loads is ~36 9 106 and ~112 9 106 mm3/ Nm, respectively. For Cu-12.5 wt pct Pb sample, unlike 10 wt pct Pb, the average wear rate of samples sintered at 573 K (300 C) is slightly higher than the wear rate of samples sintered at 623 K (350 C). The highest wear rate obtained is limited to ~25 9 106 mm3/Nm, achieved in the case of samples sintered at 573 K (300 C) with fretting load of 10 N. For Cu-15 wt pct Pb sample, all the conditions yield a converging wear rate, and notably the achieved wear rate (~1.25 9 106 mm3/Nm) is the lowest among three Pb-containing (10, 12.5, and 15 wt pct) composites studied here. METALLURGICAL AND MATERIALS TRANSACTIONS A

Figure 6 provides representative micrographs of the unworn polished surface of the investigated Cu-Pb samples. These micrographs provide an overview of the morphology and distribution of the different phases as well as the intrinsic porosity present. Figure 6(a), corresponding to Cu-10 wt pct Pb (sintered at 573 K (300 C)), shows that some of the Pb particles are not fully melted during sintering (since melting point of Pb is 600 K (327 C)) and the distribution is not uniform. In contrast, in the case of Cu-12.5 wt pct Pb (sintered at 623 K (350 C)), the distribution of Pb particles is more uniform (Figure 6(b)). Further increasing the Pb content to 15 wt pct results in still more uniform distribution of softer Pb phase and this is relevant from the tribology point of view (see Figures 6(c) and (d)). Detailed microstructural investigation using SEM equipped with EDS was carried out on the worn surfaces to understand the mechanisms governing the fretting wear. The different elemental peaks in the EDS spectra were identified and labeled on the basis of corresponding photon energy (in KeV) as follows: CuL-1.012, PbL-10.550, PbM-2.342, CuK-8.630, FeK-6.398, FeL-0.705, and OK-0.525. In the following, the characteristic features of worn surfaces for each composite are described separately. 1. Cu-10 wt pct Pb The overall topographical features of worn scars formed on the fretted surfaces are shown in Figures 7 and 8. Figure 7(a) shows the worn surface of the sample sintered at 573 K (300 C) and at fretting load of 10 N. The accumulation of the tribolayer can be seen at the edge of the wear scar, though the region around the contact point of the ball and the flat is quite clean and free of any wear scratches. The debris collected from the scar is mainly an agglomeration of the finely kneaded Cu particles present as thin platelets. The EDS analysis VOLUME 45A, JANUARY 2014—487

Table II.

Values of Mean Coefficient of Friction (COF) and Wear Rate as a Function of Normal Fretting Load, SPS Temperature, and Pb Content

Sample Composition (Balance Cu) 0 Wt Pct Pb

SPS Temperature [K (C)]

Normal Load (N)

Mean COF

573 (300)

5 10 5 10 5 10 5 10 5 10 5 10 5 10 5 10

0.76 0.70 0.78 0.71 0.53 0.50 0.74 0.73 0.40 0.45 0.42 0.49 0.40 0.45 0.44 0.59

623 (350) 10 Wt Pct Pb

573 (300) 623 (350)

12.5 Wt Pct Pb

573 (300) 623 (350)

15 Wt Pct Pb

573 (300) 623 (350)

Wear Rate (106 mm3/Nm) 183.46 207.84 104.98 129.79 14.58 31.89 35.37 112.05 6.38 24.47 1.43 8.91 1.25 2.78 1.63 7.36

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

8.34 19.50 9.05 7.55 4.55 6.81 4.43 6.67 1.82 9.83 0.29 2.00 0.66 1.12 0.35 2.97

(Figure 7(b)) reveals minute quantities of Fe, Pb, and O. Figure 8 corresponds to the sample sintered at 623 K (350 C) and at the fretting load of 10 N. Figure 8(a) (low magnification view) reveals the segregation of Pb-rich worn particles at the wear scar region. This fact is supported by the high magnification micrograph (Figure 8(a)) in which the Pb particles, being soft, are pulled out and spread on the surface (white region). EDS data shown alongside indicate the presence of Fe and O on the worn surface. SEM examination of wear debris (Figure 8(b)) indicates the prominent presence of the equiaxed particles, unlike continuous fragments observed in the case of the sample sintered at 573 K (300 C). Raman spectroscopic analysis of the debris indicates (Figure 9) strong peaks corresponding to Fe2O3 (636.18 cm1) and Cu2O (213.57 cm1). It is quite probable that the debris particles might have oxidized due to continuously generated frictional heat and henceforth abraded the steel ball. By this reasoning, we can infer that the operative nature of the wear phenomena is partly oxidative and partly abrasive.

Fig. 5—(a) Variation of wear rate of Cu-Pb composites plotted as a function of Pb content. Fretting conditions: 5 and 10 N load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel; (b) Variation of wear resistance plotted as a function of inverse of square root of grain size of sintered Cu-Pb composites.

488—VOLUME 45A, JANUARY 2014

2. Cu-12.5 wt pct Pb Figures 10 and 11, corresponding to samples sintered at 573 K and 623 K (300 C and 350 C), respectively, illustrate the fretting-induced damage at a load of 10 N. Figure 10(a) (low magnification view) provides an overview of the worn region. The cracking at the periphery of the wear scar in a direction perpendicular to the relative motion of the ball and disk (see arrow) and the formation of deep parallel grooves on the flat (marked by arrow in high magnification micrograph) indicate abrasive wear phenomena. Since the Cu-12.5 wt pct Pb sample possesses higher hardness (2.34 GPa at 573 K (300 C) and 2.54 GPa at 623 K (350 C)) than the Cu-10 wt pct Pb sample (2.29 GPa at 573 K (300 C) and 1.93 GPa at 623 K (350 C)), the steel counterbody abrades the surface rather than deforming

METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 6—BSE–SEM micrographs of the SPS samples (a) Cu-10 wt pct Pb at 573 K (300 C), (b) Cu-12.5 wt pct Pb at 623 K (350 C), (c) Cu15 wt pct Pb at 573 K (300 C), and (d) Cu-15 wt pct Pb at 623 K (350 C). Temperature values quoted above are the sintering temperature.

it. Softer Pb particles in the vicinity of the grooves (marked with *) stay undeformed in the matrix whereas Cu grains have extracted out due to continued fretting and formed wear debris particles. Further, repeated fretting results in the transfer of the iron particles from the steel ball and thus the formation of a mechanically mixed tribolayer (MML) comprising Cu and Fe (Figure 10(b), high magnification view). Figure 10(c) indicates that the wear debris accumulated mainly consists of finer Cu particles. This finding is further corroborated by the similar finding in the case of Cu-10 wt pct Pb composite in which extensive characterization of wear debris using TEM confirms that the Cu nanograins originally present in the sintered composite have been fragmented to a still smaller size range due to fretting wear.[17] Figure 11(a), corresponding to sample sintered at 623 K (350 C), shows the formation of partial microcracking and microcavities on the worn surface (white arrow). The particle detachment over these spots can be explained by the fretting fatigue of the surface layers by the harder counterbody and subsequent fracturing of the surface. Figure 11(b) shows the size scale and elemental composition of the debris particles. After being debonded from the surface, the fragmented finer Pb particles in the further process of rubbing accumulated and smeared over the flat surface, thereby reducing the contact between the sample and counterbody. The accumulation of the Pb particles is supported by the fact that Pb present in the worn surface has much METALLURGICAL AND MATERIALS TRANSACTIONS A

finer size (Figure 11(a), high magnification micrograph) compared to Pb being collected as the debris (Figure 11(b)). The stepwise phenomena of Pb coming out, smearing over the surface, and reducing contact between the tribocouple (flat sample and spherical counterbody) manifest themselves into lowering the coefficient of friction. 3. Cu-15 wt pct Pb Figure 12(a) shows the formation of the transfer film or layer on the steel counterbody after the wear of Cu-15wt pct Pb composite (sintered at 573 K (300 C)) at a fretting load of 10 N. From the high magnification image, it is clear that the formation of the debris and the smearing of the layer have uniformly taken place over the entire contact region of the wear scar. Figure 12(b) shows the surface of the Cu-15 wt pct Pb sample, sintered at 623 K (350 C) and at fretting load of 10 N. The presence of few particles with sharp surfaces (Figure 12(b), high magnification micrograph) on the rather unworn surface can be justified by the fretting fatigue of the particles on the surface layers, subsequent loosening, and debonding from the matrix. F. Subsurface Observation of the Worn Surface The FIB analysis to investigate the subsurface damage was carried out on a selected sample. Figure 13 presents VOLUME 45A, JANUARY 2014—489

deformed Pb particles. Figure 13(d) reveals the micrograph obtained with ion beam in the SE mode. The presence of cracks (marked by white arrows) is observed. The higher magnification micrograph (right) shows elongated Pb particles. Some of the Pb particles (marked by thick black arrows) show well-defined necks probably due to localized heating during the wear process. The Cu grains cannot be clearly seen on the micrographs. However, the presence of agglomerated Cu particles is frequently observed. These results will further be discussed in the next section in view of probable temperature increase at frictional interface,[18,19] compressive stresses, and fatigue of the subsurface grains, leading to the plastic deformation of Pb particles and formation of subsurface cracks.

IV.

Fig. 7—(a) BSE–SEM micrographs of worn surface of Cu-10 wt pct Pb composite, SPS at 573 K (300 C); (b) the wear debris particles and corresponding EDS spectra. Fretting conditions: 10 N normal load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel. Double-ended arrow in (a) shows the fretting direction.

the FIB micrographs of worn surface of Cu-10 wt pct Pb sample, sintered at 573 K (300 C) and after fretting at a load of 5 N. In order to investigate the subsurface damage after the wear process, the worn sample was milled approximately at the central region of the worn scar, which typically experiences maximum wear damage. Figure 13(a) shows a plan view of the milled trench along with delaminated tribolayer. Some discrete debris particles are also observed, which might have formed as a result of fragmentation from MML. Figure 13(b) through (d) shows the subsurface region, viewed at an angle of 45 deg with respect to the milling direction of FIB. Figure 13(c) has been imaged using back scattered electron mode in the SEM, whereas Figure 13(d) is obtained using ion beam in FIB. Therefore, these figures reveal the presence of different phases, which are marked on the figures. One can clearly observe the presence of the elongated Pb particles formed parallel to the wear direction at a maximum distance of ~9 lm from the top. The Pb particles are found to have undergone extensive deformation by the wear process and even some of the Pb particles have been fractured (Figure 13(c), insets). The right inset in Figure 13(c) shows the formation of steps on the surfaces of the 490—VOLUME 45A, JANUARY 2014

DISCUSSION

The present investigation has clearly demonstrated that the attainment of higher hardness in Cu-Pb composites along with homogeneously distributed Pb particles in the microstructure can result in improved tribological properties, more particularly a better wear resistance. This is considered to be a major improvement as far as the inclusion of a soft phase, i.e., Pb, in a relatively hard matrix (Cu) is concerned. Moreover, the binary Cu-Pb system being an immiscible one,[13] the achievement of uniform distribution of Pb till 15 wt pct in the Cu matrix is another significant result. A summary of the results from the literature, reporting the friction and wear properties of the conventionally processed Cu-Pb alloys, is provided in Table III and a comparison with spark plasma sintered Cu-Pb nanocomposites in the present investigation is made therein. Apart from the results obtained with Cu-Pb alloys, fabricated using the chill/die casting method, the frictional properties of nanocrystalline Cu as well as electroplated Pb film on Cu substrate are also summarized in Table III. It is to be noted that these results are considered as baseline behavior. Most of the work to date on Cu-Pb alloys has been carried out using pin-on-disk configuration in sliding wear tests. Buchanan et al.[8] and Molian et al.[9] investigated the COF and wear rate variation of swirl and die-cast Cu-Pb alloys (Pb wt pct = 20, 40, and 60) in a pin-on-disk configuration at varying contact pressures (0.14, 0.28, 0.42, and 0.56 MPa). They measured steady state COF and wear rate at pressures equal to or greater than 0.28 MPa. At pressure of 0.14 MPa, COF and wear rate for the three composites are 0.4 to 0.5 and 60 to 100 9 105 mm3/Nm, respectively, whereas the values decrease down to ~0.2 and 37 to 77 9 105 mm3/Nm at pressures of 0.28 MPa or more, respectively. Pathak and Tiwari[10] investigated the sliding wear with pin-on-disk configuration of a chill-cast Cu-Pb system with varying the Pb content from 5 to 40 wt pct and reported wear rate values ranging from 5.35 9 106 to 1.27 9 106 mm3/Nm. The systematic decrease in wear rate, with increasing Pb content, and the smearing of Pb phase as well as plowing as the dominant wear mechanisms were observed. Zhang et al.[14] reported the fretting behavior of surface mechanical attritiontreated (SMAT) pure Cu; the COF was close to 0.7, METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 8—(a) BSE–SEM micrographs of worn surface of Cu-10 wt pct Pb composite, SPS at 623 K (350 C) with a high magnification image of the scar and the corresponding EDS spectra; (b) the wear debris particles and corresponding EDS spectra. Fretting conditions: 10 N normal load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel. Double-ended arrow in (a) shows the fretting direction.

whereas the wear rate varied in the range of 1 9 106 to 7.5 9 106 mm3/Nm.[15] It is to be noted that the wear rate of the spark plasma sintered Cu-15 wt pct Pb nanocomposite in the present investigation is similar to that of chill-cast Cu-15 pct Pb alloy.[10] Because of the inherent process advantages of the novel fabrication route, i.e., SPS adopted here, it has been insured that the distribution of the phases should be more uniform, free METALLURGICAL AND MATERIALS TRANSACTIONS A

from any segregation, and the nanocrystalline grain size could be retained (without any significant grain growth) in the sintered microstructure.[20] Based on the above findings, the inclusion of 15 wt pct Pb in the Cu-based composite results in the achievement of the best optimized tribological properties among all the samples studied. The possible reasons for the enhanced tribological properties of the Cu-Pb composites will now be VOLUME 45A, JANUARY 2014—491

impurities in these composites as confirmed by the X-ray analysis of the sintered composites as well as the lower grain sizes. From Figures 4(a) and (b), it can be seen that as Pb content in the Cu matrix is increased beyond 15 wt pct, the COF values show an increase for a lower sintering temperature of 573 K (300 C) and at both 5 and 10 N loads. In addition, from the SEM micrographs of Cu20 wt pct Pb, spark plasma sintered at 573 K and 623 K (300 C and 350 C) (Supplementary Figure S1), Pb shows a rather segregating tendency and appears as large flakes throughout the microstructure. The increase in COF can thus be justified on the following basis. In the course of the fretting process, the contacting surfaces become lubricated by a thin film of excess Pb extruded out of the sample. The coarser wear debris particles, thus produced, result in a surge in COF due to effect of a third body between the counterbody and sample. Fig. 9—Raman spectra showing the presence of Cu-, Pb- and Fe oxides in the debris after fretting wear of Cu-10 wt pct composite, SPS at 623 K (350 C). The exact peak positions are shown as dotted lines with corresponding values and phases formed. Fretting conditions: 10 N load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length against bearing steel.

discussed in this section in reference to the following aspects: (a) achieved high hardness: effect of matrix grain size and Pb content (b) role of Pb on the friction and wear: nature of the underlying wear phenomena—role of Pb and subsurface cracks (c) lower wear rate: mutual solubility of the constituents, size scale of the wear debris particles, and formation of MML comprising essentially of the elemental oxides (a) As reported in the previous study[5] regarding the sintering of Cu-10 wt pct Pb alloy, the mean grain sizes obtained at sintering temperatures of 573 K, 623 K, and 673 K (300 C, 350 C, and 400 C) are 82, 124, and 199 nm, respectively. The Vickers hardness values achieved for these samples are close to ~2 GPa and followed the classical Hall– Petch relationship with mean grain size. The nanometric grain size of the Cu matrix is attributed to high hardness in the Cu-Pb composites for each of the composites with different Pb contents (10, 12.5, and 15 wt pct). Since the mean grain size of the Cu matrix shows an increase beyond 673 K (400 C), there is a concomitant decrease in hardness values in sync with the Hall–Petch equation[21,22] (Figure 2). A similar trend in hardness variation with sintering temperature is evident for all the compositions, as can be seen from Figure 2. The mean hardness values for Cu-12.5 and 15 wt pct Pb at a particular sintering temperature, e.g., 623 K (350 C), are 2.5 and 3.5 GPa, respectively. The slight increase in hardness for Cu-12.5 and 15 wt pct Pb composites in comparison to Cu10 wt pct Pb composite may be attributed to the formation of some fine-scale Cu and Pb oxide 492—VOLUME 45A, JANUARY 2014

(b) Bowden and Tabor[23] have investigated extensively the friction and wear characteristics of Cu-Pb alloys. They have compared the performances of Pb-coated Cu-20 wt pct Pb and Cu-27 wt pct Pb and reported a COF of 0.18. They have also suggested that the Pb-containing Cu alloys are better as compared to deposited Pb film on the Cu surface, since the lubrication will not be a restricted function of Pb film, and the so-called ‘‘reservoir effect’’ will be there for extended service life. Rabinowicz[24] carried out experiments on Cu-16 pct Pb alloy and reported a COF of 0.2. In contrast, Tsuya and Takagi[25] after investigating Pb films of varying thickness on the Cu substrates, reported COF values between 0.5 and 2 and concluded that the COF is a function of both the load and the Pb film thickness. Buchanan et al.[8] and Molian et al.[9] explored the feasibility of alloying large amount of Pb in Cu matrix in order to get a higher reduction in friction coefficient. They alloyed 20, 40, and 60 pct Pb into Cu matrix and observed that just increasing the Pb content does not necessarily contribute to lowering of COF and that there must be some critical value of Pb above which no further reduction in COF is possible. They attributed the change in wear mechanism from ‘‘oxidation’’ at low contact pressures (0.14 MPa) and ‘‘plastic deformation and adhesion’’ at high contact pressures (0.28, 0.42, and 0.56 MPa) to the disruption of the protective oxidative film and increased bulk temperature. In view of the above observations, it was deemed suitable to optimize the alloying content of Pb in the Cu matrix. From the tribological point of view, it has also been emphasized that ‘‘self-lubricated metallic bearings’’ should consist of a ‘‘duplex structure’’ with ‘‘continuous matrix of the harder metal with a small amount of the softer metal finely dispersed through it.’’[23] This requires the two phases to show some amount of immiscibility in the solid state. From the consideration of binary Cu-Pb equilibrium phase diagram,[13] it can be inferred that Pb is barely soluble in Cu. Therefore, it will preferentially segregate along the grain boundaries or METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 10—(a) BSE–SEM micrographs of worn surface of Cu-12.5 wt pct Pb composite, SPS at 573 K (300 C) with a high magnification image of the scar and the corresponding spot (see asterisk) EDS spectra; (b) the formation of transfer film (white arrow) on the counterbody; (c) the corresponding wear debris particles. Fretting conditions: 10 N normal load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel. Double-ended arrow in (a) shows the fretting direction.

METALLURGICAL AND MATERIALS TRANSACTIONS A

VOLUME 45A, JANUARY 2014—493

Fig. 11—(a) BSE–SEM micrographs of worn surface of Cu-12.5 wt pct Pb composite, SPS at 623 K (350 C) with two high magnification images; (b) the wear debris particles and corresponding spot EDS spectra. Fretting conditions: 10 N normal load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel. Double-ended arrow in (a) shows the fretting direction.

triple junctions of Cu grains.[5] These interfacial regions will then be the potential sites for the nucleation and propagation of cracks due to frictional heating at the fretted zone, because of the incompatibility of the thermal expansion coefficients between Cu and Pb. The subsequent sub-processes, viz., the instantaneous temperature increase and the debris generation, will induce a number of thermomechanical processes (like oxidation and work hardening) at the contacting surfaces and formation of the mechanically mixed transferred tribolayer on the counterbody/mating material. For Cu-10 wt pct Pb composite, the compositional (EDS: Figures 7(b) and 8(b)) analysis of the debris suggests the wear phenomena facilitated by the generation of an oxidized ‘‘third body’’ and indicates the presence of Pb, Cu, Fe, and O peaks. As shown in 494—VOLUME 45A, JANUARY 2014

Figure 9, the analysis of Raman spectra indicates that PbO (146.84 and 524.96 cm1) and Fe2O3 (410.48 and 636.18 cm1) are present in prominence in the debris particles, though affinity of Cu getting oxidized in the ambient atmosphere also gives a corresponding Cu2O (213.57 cm1) peak. The fragmented particles, entrapped loosely in between the contacting surfaces, may get oxidized due to temperature rise. This is indicative of a transition in the wear behavior from oxidative to slightly abrasive. The subsurface analysis of the worn surface was carried out in BSE as well as SE modes to study the behavior of Pb particles and formation of cracks. Figure 13(c) shows the plastic deformation of the subsurface Pb particles beneath the worn scar. High magnification images clearly show that discrete Pb particles have formed ‘‘steps,’’ which might METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 12—(a) BSE–SEM micrograph showing the formation of transfer film (encircled) on the steel counterbody after wear of Cu-15 wt pct Pb, SPS at 573 K (300 C) and the corresponding EDS spectra. Also shown, a magnified view of the tracks formed out of the transfer film. (b) BSE– SEM micrographs of worn surface of Cu-15 wt pct Pb, SPS at 623 K (350 C). The high magnification image shows few particles (black arrows) being detached out of the composite in the course of fretting. Fretting conditions: 10 N normal load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel. Double-ended arrow shows the fretting direction.

be due to the combined effect of frictional heating and the compressive stresses. The frictional heating can cause the thermal softening, while the compressive stresses acting simultaneously on Cu and Pb will result in the preferential deformation of the Pb particles. This METALLURGICAL AND MATERIALS TRANSACTIONS A

consequently leads to the step formation in comparatively softer and low melting point Pb grains. Figure 13(d) was taken in the SE mode to track down the behavior of the subsurface cracks. These cracks may have formed due to two reasons—first, due to dislodging VOLUME 45A, JANUARY 2014—495

Fig. 13—SE-FIB micrographs of worn surface of Cu-10 wt pct Pb composite, SPS at 573 K (300 C). (a) Plan view showing the milled trench and naı¨ ve worn surface (z-axis perpendicular to inside of the plan view), (b) the wall of the trench after polishing (white dashed line marks the boundary between surface and trench wall), (c) BSE-SEM micrographs of the polished wall. High magnification images (insets) show subsurface tribo-behavior of Pb particles and (d) SE-FIB images showing the formation and coalescing of subsurface cracks. Fretting conditions: 5 N normal load, 60,000 cycles, 5 Hz frequency, and 100 lm stroke length. Counterbody: bearing steel. Double-ended arrow in (a) shows the fretting direction.

of the Pb particles from their original microstructural sites, empty space (voids) formed therefore gets laterally deformed in the fretting wear direction. Secondly, intrinsic porosity (~12 pct, see Table I) develops cracking at the loaded tribocontact. These two subsurface processes then lead to formation of the lateral subsurface cracks beneath the worn scar region.[26] In addition, as shown in Figure 13(d) (black arrows), the bridging of the two adjacent lateral cracks and some new cracks in their embryonic stage are observed. The aforementioned reasoning is well corroborated by the fact that the size scale of the Pb particles as well as the formed lateral cracks is quite close and around 1 to 2 lm. For Cu-12.5 wt pct Pb, the formation of deep parallel grooves and microcavities (not marked) in the wear scar region (high magnification view, Figure 10(a)) with undeformed softer Pb particles (white) in the vicinity of the grooves is indicative of highly abrasive wear phenomenon. From Figure 11(a), the nature of the abrasive wear can be justified on the basis of the formation of numerous microcavities in the scar region. These microcavities are thought to form because of the removal of Pb from these sites. The detached Pb particles will remain unoxidized and therefore the wear 496—VOLUME 45A, JANUARY 2014

phenomena can be considered totally abrasive in nature. Since the highest hardness is achieved in the case of 15 wt pct Pb composites, it is quite probable that the continued fretting may have resulted in the fatigue of the surface layers, resulting in the abrasion of the Cu-Pb sample as well as the steel ball. The formation of a mechanically mixed transfer layer, formed out of Fe from the steel ball counterbody and Cu/Pb from the flat sample, has been confirmed by the EDS spectra, provided in Figure 12(a). The formation of the highly deformed wear tracks (circled in high magnification view, Figure 12(a)) in the tribolayer, comprising primarily ductile Pb, might have resulted due to the high severity and rate of straining associated with the high rate of thermal cycling. The normal load of 10 N effects an equivalent contact stress of ~0.3 GPa, which being greater than the shear strength of Cu is sufficient to induce plastic deformation of Cu, leading to formation of dislocation network and sub-cells which ultimately results in grain refinement and work hardening of the surface layers in the contact zone.[17] These two factors of fretting fatigue and work hardening of the surface layers are anticipated to be the probable reasons for Cu15 wt pct Pb to be highly resistant toward the fretting METALLURGICAL AND MATERIALS TRANSACTIONS A

METALLURGICAL AND MATERIALS TRANSACTIONS A

VOLUME 45A, JANUARY 2014—497

Hardness

—

0.15 to 0.20 0.15 to 0.20 0.15 to 0.20

COF

129.79 56.15 4.4 3.52

0.73 0.49 0.59

—

6.12 5.35 4.59 4.33 3.82 3.06 2.04 1.27 1 to 7.5

370 760 560

Wear Rate (9106 mm3/Nm)

0.71

0.7 ball-on-flat, fretting wear at 20 N, stroke length—2 mm, duration—60 min sliding speed—5 9 105 m/sec, 0.5 to 0.6 against copper, pressure—0.5 MPa

pin-on-disk sliding speed—1 m/s, against steel, contact pressure—0.28 MPa pin-on-disk, sliding speed—1 m/s, distance—3.6 km, load—39.2 N

Test Conditions

Hv 2.25 GPa ball-on-flat, fretting wear at Hv 1.9 GPa 10 N, against steel Hv 2.5 GPa stroke length—100 lm, Hv 3.25 GPa speed—2 9 103 m/sec, cycles—60,000

—

VHN 61.2 VHN 47.5 VHN 45.7 VHN 42.3 VHN 39.5 VHN 37.5 VHN 33.8 VHN 26.5 Hv 1.3 GPa

conventional casting BHN 50 swirling die casting BHN 38 swirling die casting BHN 25

Processing Route

delamination, mild oxidative and abrasive wear

adhesive wear

local damage and delamination of mechanically mixed layer

smearing of Pb, cracking and plowing

oxidative at low pressure and plastic deformation/adhesion at high pressures

Wear Mechanism/s

Refs.

present work

[25]

[14,15]

[10]

[8,9]

Summary of the Literature Results on Friction and Wear Rate of Cu-Based Alloys, Processed via Various Manufacturing Routes as well as Comparison with Present Results

Pure Cu chill casting Cu-5 Pb Cu-10 Pb Cu-15 Pb Cu-20 Pb Cu-25 Pb Cu-30 Pb Cu-40 Pb Pure surface Nanograined Cu mechanical attrition treatment Pb Film 0.1- to 130-lm Deposited on Cu Pb film deposited by electroplating Pure spark plasma Nanograined Cu sintering Cu-10 Pb Cu-12.5 Pb Cu-15 Pb

Cu-20 Pb Cu-40 Pb Cu-60 Pb

Material Compn. (Wt Pct)

Table III.

wear process, and hence the wear scar is also not prominently visible. (c) The low wear rate or correspondingly a high wear resistance achieved in the present study for Cu-Pb (Pb wt pct = 10, 12.5, and 15) composites is a significant improvement in light of the uniform distribution and morphology of the softer Pb phase in harder Cu matrix, nanocrystalline grain size of the Cu matrix, and the lower processing temperature as compared to conventional processing route. The severity of wear of the Cu-Pb composites can qualitatively be understood and is discussed below on the basis of the following aspects: (i) Mutual solubility of Cu, Pb, and Fe at the tribological interface It has been reasonably put forward by many investigators that the mutual solid solubility between different elements comprising the composite and also with the counterbody material can affect the extent of the debris generation as well as mutual transfer between mating materials.[27] From the phase diagram consideration of binary Cu-Pb, CuFe, or Fe-Pb systems,[13] it can be concluded that these metal couples have very little or negligible solid solubility in one another. The tribological implication of this thermodynamic aspect can practically be seen as the formation of very low amount of debris between the contacting surfaces in the present study. (ii) Size scale and morphology of the debris particles generated during fretting As evident from Figures 7(b), 8(b), and 10(c), a rough estimate about the size scale and morphology of debris particles can be made. The ultrafine debris particles (~1 to 2 lm) are characteristic of a mild wear process, as categorized by Chen and Rigney[28] and Williams.[29] At this juncture, a simple analytical calculation to supplement this discussion seems reasonable. From the Hertzian contact theory, we know that the contact radius ‘‘a’’ is given by   3WR 1=3 ; a¼ 4E

½2

where W is the normal load (N), R* the composite radius (m), and E* being the composite Young’s modulus (Pa). Now, Young’s modulus (E1) and Poisson’s ratio (m1) for the Cu-10 wt pct Pb (standard values taken from Smithells Handbook[30]) composite as obtained by applying rule of mixtures are estimated as E1 = 120.87 GPa and m1 = 0.35. Again, taking E2 = 210 GPa and m2 = 0.3 for the steel counterbody, the estimated values of composite modulus and composite radius are E* = 86.5 GPa and R* = 0.01 m, respectively. Considering these estimated values and a normal load W = 10 N in Eq. [2], we obtain Hertzian contact radius a = 95.35 lm and the corresponding nominal contact area An = pa2 = 2.85 9 108 m2. The actual pressure on the asperities on the flat, therefore, becomes ~0.3 GPa (from P = F/An, where F = 10 N), which is comparable to the shear strength of Cu. It is therefore quite likely that the asperities of the Cu-Pb nanocomposites are easily 498—VOLUME 45A, JANUARY 2014

knocked off at the contacting interface. These particles, being continuously subjected to the cyclic stresses, have been work hardened and agglomerated to form flaky sheets covering the worn scar. (iii) In situ formation of elemental oxides at the interface In view of the high thermal conductivity of the mating materials as well as due to low speed involved in fretting tests (0.002 m/s), the thermal conductivity at contacting surface is expected to be low and we believe that large shear stresses cause extensive dislocation activity on the worn surfaces. This will facilitate ‘‘pipe diffusion’’ around the dislocations, promoting the oxidation on severely worn surfaces of Cu-Pb nanocomposites. This, therefore, can explain the phase transformation of the Cu debris particles to Cu2O (as confirmed by Raman spectroscopy, Figure 9). This will hinder further the contact of the steel counterbody and fresh surface of Cu-Pb composites. The Cu2O dispersoids, being harder, may then abrade the steel counterbody and also scoop out some of the ductile Pb particles from the bulk as the debris. These in situ transformed oxide dispersoids (Cu, Fe)xOy, because of the thermomechanical processes occurring at the contact region,[27,31] will form a cloud over the scar and being harder than Cu or Pb itself will result in the reduction in COF and wear rate. This mechanically mixed layer (MML) comprising various elemental oxides present as dispersoids in the loose continuous layer of delaminated debris particles will shroud the wear scar and will be responsible for the mild oxidational wear of the composite. Apart from correlating the wear properties in terms of the observation of the surface/subsurface damage behavior and Pb distribution, the measured wear resistance (~106 Nm/mm3) for Cu-Pb nanocomposites can be related to finer microstructure (nanometric grain size) and oxides generated during the process of wear. In Figure 5(b), the wear resistance is plotted as a function of the inverse of square root of grain size in a manner analogous to the Hall–Petch expression. The probable reasons for the appearance of two different ‘‘Hall– Petch’’ slopes could be due to the following two factors: (a) The effect of the defect structure generated during fretting, i.e., twins in Cu grains as well as fragmentation of Cu nanograins to still finer nanograins (please refer to Figures 2(a) and (b) in[17]), and the corresponding strengthening effect arising out of two separate entities—the grain boundaries and interior of nanograins—could be the reasons for this dual slope behavior. The nanometric grains with large volume fraction of grain boundaries impede dislocation movement, thereby inhibiting large-scale deformation, and it can be noted that the extensive deformation can result in shear crack formation. The continued fretting over a constricted region having a diameter <200 lm (=2a where a, is the contact radius) with a frequency of 5 Hz might have resulted in the in situ generation METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 14—Schematic of the various transient wear mechanisms in spark plasma sintered Cu-Pb nanocomposites. Double-ended arrow shows the fretting direction.

of a profuse nanotwinned structure in the initially processed nanometric Cu grains. These nanotwin lamellae, which typically can have an average band thickness of ~20 nm, have sevenfold rate sensitivity of hardness, as reported by Lu et al.[32] This extraordinary work hardening will induce a high density of nanoscale twins in fcc-Cu, resulting in unusual hardness of surface layers restricted to few microns from the top surface.[15,33] The mechanistic interaction of dislocations with nanotwins in the nanograined Cu and their subsequent arrest will thus be responsible for the enhanced wear resistance of the currently investigated Cu-Pb nanocomposites. (b) Secondly, the generation of in situ oxides (viz., Cu2O, Fe2O3, PbO) in the wear debris (please refer to Figure 2(c) in[17]) may also result in the reduction of wear rate due to formation of an oxide dispersion-strengthened (ODS) layer resulting in enhanced wear resistance of these nanocomposites. As the precise estimation of these in situ oxides and their cumulative effect on the quantification of the strengthening effect of surface layers on the wear process is complex and outside the scope of the present study, the results are presented and discussed only in a qualitative fashion. In accordance with the above discussion, Figure 14 shows the schematic of the wear mechanisms corresponding to surface processes for individual Cu-10 wt pct Pb, Cu-12.5 wt pct Pb, and Cu-15 wt pct Pb composites based on the variation in their hardness values. The process highlight for each of the composite has been mentioned below the individual figure.

respectively, at a fretting load of 5 N. The optimal combination of COF and wear rate can be explained by the interplay of high hardness and the deformation of the Pb during the fretting wear. (b) The formation of the MML comprising oxides of Cu, Pb, and Fe inhibits the continuous contact of the steel counterbody and Cu-Pb surface and thus is responsible for the lower wear rate in the present study. (c) Subsurface damage evaluation using FIB reveals the formation of lateral cracks close to the wear surface. The bridging of these cracks and propagation toward the surface are responsible for the formation of the delaminated tribolayer. (d) Importantly, the wear resistance reveals a weak Hall–Petch-type correlation with grain size and therefore indicates the possibility of achieving better wear resistance in ultrafine-grained or nanostructured Cu-based nanocomposites.

ACKNOWLEDGMENTS We would like to acknowledge the reviewers for constructive criticism, in-depth analysis, and fruitful recommendation. Overall, the authors are extremely impressed with the reviewers. It was a pleasure to go through the reviewers’ comments. The authors would also like to acknowledge the funding agencies, the Department of Science and Technology (DST), Govt. of India, and CARE Grant, IIT Kanpur, for procuring the SPS facility at IIT Kanpur. AFMM, IISc Bangalore is thanked for facilitating the FIB characterization. ELECTRONIC SUPPLEMENTARY MATERIAL

V.

CONCLUSIONS

The present work has demonstrated that it is possible to achieve a low wear rate (of the order of 106 mm3/Nm) for Cu-Pb composites by consolidation of nanometric powders and by tailoring SPS conditions. In particular, the following major conclusions can be drawn: (a) Cu-15 wt pct Pb composite, spark plasma sintered at 573 K (300 C), exhibited the lowest COF and wear rate of 0.4 and 1.25 ± 0.66 9 106 mm3/Nm, METALLURGICAL AND MATERIALS TRANSACTIONS A

The online version of this article (doi: 10.1007/s11661-013-1965-7) contains supplementary material, which is available to authorized users.

REFERENCES 1. ASM Handbook: Friction, Wear and Lubrication Technology, vol. 18, ASM, Materials Park, OH, 1992.

VOLUME 45A, JANUARY 2014—499

2. B. Bhushan: Principles and Applications of Tribology, 1st ed., John Wiley and Sons, New York, NY, 1999. 3. W.A. Glaeser: J. Met., 1983, vol. 35, pp. 50–55. 4. B. Basu and M. Kalin: Tribology of Ceramics and Composites: Materials Science Perspective, 1st ed., John Wiley Publications, New York, 2011. 5. A.S. Sharma, K. Biswas, B. Basu, and D. Chakravarty: Metall. Mater. Trans. A, 2011, vol. 42A, pp. 2072–84. 6. G.C. Pratt: Int. Metall. Rev., 1973, vol. 18, pp. 62–88. 7. F.P. Bowden and D. Tabor: J. Appl. Phys., 1943, vol. 14, pp. 141– 51. 8. V.E. Buchanan, P.A. Molian, T.S. Sudershan, and A. Akers: Wear, 1991, vol. 146, pp. 241–56. 9. P.A. Molian, V.E. Buchanan, T.S. Sudershan, and A. Akers: Wear, 1991, vol. 146, pp. 257–67. 10. J.P. Pathak and S.N. Tiwari: Wear, 1992, vol. 155, pp. 37–47. 11. B.K. Prasad: Wear, 2004, vol. 257, pp. 110–23. 12. T. Kimura, K. Shimizu, and K. Terada: Wear, 2007, vol. 263, pp. 586–91. 13. T.B. Massalski, J.L. Murray, and L.H. Bennett, eds.: Binary Alloy Phase Diagrams, vol. 1, ASM International, OH, 1986, pp. 944–47. 14. Y.S. Zhang, Z. Han, K. Wang, and K. Lu: Wear, 2006, vol. 260, pp. 942–48. 15. Y.S. Zhang, K. Wang, Z. Han, and G. Liu: Wear, 2007, vol. 262, pp. 1463–70.

500—VOLUME 45A, JANUARY 2014

16. T.S. Srivatsan, B.G. Ravi, A.S. Naruka, L. Riester, S. Yoo, and T.S. Sudarshan: Mater. Sci. Eng. A, 2001, vol. 311, pp. 22–27. 17. A.S. Sharma, K. Biswas, and B. Basu: J. Nanoparticle Res., 2013, vol. 15, pp. 1–12. 18. D.A. Rigney and J.P. Hirth: Wear, 1979, vol. 53, pp. 345–70. 19. P. Heilmann and D.A. Rigney: Wear, 1981, vol. 72, pp. 195–281. 20. Z.A. Munir, U.A. Tamburini, and M. Ohyanagi: J. Mater. Sci., 2006, vol. 41, pp. 763–77. 21. E.O. Hall: Proc. Phys. Soc., 1951, vol. B64, pp. 747–53. 22. N.J. Petch: J. Iron Steel Inst., 1953, vol. 174, pp. 25–28. 23. F.P. Bowden and D. Tabor: The Friction and Lubrication of Solids, Clarendon, Oxford, 1950. 24. E. Rabinowicz: J. Lubr. Technol., 1975, vol. 97, pp. 217–49. 25. Y. Tsuya and R. Takagi: Wear, 1964, vol. 7, pp. 131–43. 26. N.P. Suh: Wear, 1973, vol. 25, pp. 111–24. 27. B.K. Prasad, A.K. Patwardhan, and A.H. Yegneswaran: Mater. Sci. Technol., 1996, vol. 12, pp. 427–35. 28. L.H. Chen and D.A. Rigney: Wear, 1985, vol. 105, pp. 47–61. 29. J.A. Williams: Tribol. Int., 2005, vol. 38, pp. 863–70. 30. E.A. Brandes and G.B. Brook, eds: Smithells Metals Reference Book, 7th ed., Butterworth-Heinemann, Oxford, 1992, pp. 13-1-13-119. 31. F.E. Kennedy, Jr: Wear, 1984, vol. 100, pp. 453–76. 32. L. Lu, R. Schwaiger, Z.W. Shan, M. Dao, K. Lu, and S. Suresh: Acta Mater., 2005, vol. 53, pp. 2169–79. 33. Y.S. Zhang, Z. Han, and K. Lu: Wear, 2008, vol. 265, pp. 396–401.

METALLURGICAL AND MATERIALS TRANSACTIONS A