Dry sliding wear of nanostructured Fe30Ni20Mn20Al30

Dry sliding wear of nanostructured Fe30Ni20Mn20Al30

Intermetallics 23 (2012) 116e127 Contents lists available at SciVerse ScienceDirect Intermetallics journal homepage: www.elsevier.com/locate/interme...

3MB Sizes 2 Downloads 38 Views

Intermetallics 23 (2012) 116e127

Contents lists available at SciVerse ScienceDirect

Intermetallics journal homepage: www.elsevier.com/locate/intermet

Dry sliding wear of nanostructured Fe30Ni20Mn20Al30 Xiaolan Wu a, Ian Baker a, *, Hong Wu b, Paul R. Munroe c a

Thayer School of Engineering, Dartmouth College, Hanover NH 03755, USA State Key Laboratory of Powder Metallurgy, Central South University, Changsha 410083, PR China c Electron Microscope Unit, University of New South Wales, Sydney NSW 2052, Australia b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 3 September 2011 Received in revised form 14 December 2011 Accepted 26 December 2011 Available online 21 January 2012

The wear of both as-cast and annealed nanostructured Fe30Ni20Mn20Al30, which consists of fine alternating B2 (ordered b.c.c.) and L21 (further ordering B2) phases, was studied using pin-on-disk tribotests in three different test environments, oxygen, air and argon. The counterface in all tests was yttriastabilized zirconia. It was found that the annealed alloy showed better wear resistance than the ascast material, probably because the annealed Fe30Ni20Mn20Al30 has both a higher hardness and better ductility. The wear was reduced dramatically by the removal of the oxygen from the test environment. The zirconia counterface showed similar wear behavior in all environments. The tips of the wear pins were examined using a combination of X-ray diffractometry, scanning electron microscopy and transmission electron microscopy, the latter using specimens produced by focused ion beam milling. A considerable number of cracks and pits were present on the worn surface of pins tested in oxygencontaining environments, while a relatively smooth mixed layer and plastic flow were evident on the worn surface of pins tested in argon. Both two-body and three-body abrasive wear, as well as deformation and delamination, contributed to the wear. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: A. Nanostructured intermetallics B. Tribological properties F. Electron microscopy, transmission

1. Introduction The wear behavior of metallic materials has been widely studied and a number of wear mechanisms are well-known, such as adhesive wear, abrasive wear, delamination, mild and severe oxidation, melting, seizure, etc [1]. Wear-mechanism maps have been constructed to present wear data and models in a graphical way and describe the overall wear behavior of various materials under different sliding conditions [2e6]. Such maps help to predict the dominant mechanisms and anticipate the rates of wear for a given set of sliding conditions. Although wear-mechanism maps allocate a regime to the plasticity-dominated mechanism, the wear in that regime is complex and difficult to predict due to the interaction between mechanisms. Moreover, there is no wear-mechanism map describing the plastic deformation region for spinodal alloys. Spinodal alloys typically display high strength due to the compositional fluctuations arising from thermodynamic instability [7]. Because of their combination of attractive mechanical properties, and high corrosion resistance, some spinodal alloys have been commercially developed [8e10]. Surprisingly, there is little work on the wear of spinodal alloys.

* Corresponding author. E-mail address: [email protected] (I. Baker). 0966-9795/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.intermet.2011.12.017

Marczak and Ciach [11] studied the tribological properties of ZnAl40 alloy tested on a Skoda-Sasvin machine in lubricated conditions at loads in the range 2e20 kg. Compared with CuPb30, it was believed that the fine two-phase spinodal structure and the formation of oxide films on the surface were responsible for the good tribological properties of the alloy. Detailed research on the wear of spinodal alloys was performed on Cu-15 wt.% Ni-8 wt.% Sn, which is the most commonly-used commercial spinodal alloy in high-performance bearing applications [8,12]. Bellon’s group performed dry sliding pin-on-disc wear tests on this alloy against a martensitic stainless steel disc held at 440  C in air [13]. A transition from mild to severe wear was reported during the tests at loads of 245 N and 294 N, whereas the wear was severe throughout the tests at loads of 490 N and 980 N. The debris collected after testing in the severe wear regime, as characterized using X-ray diffraction (XRD) and transmission electron microscopy (TEM), was found to contain 20e300 nm grains of an f.c.c. phase instead of the initial two-phase spinodal microstructure [13]. Later, similar wear tests were carried out at a load of 980 N in air and argon [14]. In air, severe wear started at the beginning of the tests, as in the earlier studies, while a transition from mild to severe wear occurred during the tests in argon. The worn surface morphology suggested a mechanism of adhesive wear. Two different types of debris were observed: 5e30 mm flaky particles with a composition the same as the nominal composition

Average Mass Loss (mg)

6.00

Average Mass Loss (mg)

X. Wu et al. / Intermetallics 23 (2012) 116e127

6.00

5.00 4.00

Oxygen Air Argon

3.00 2.00 1.00 0.00 As-cast 5.00 4.00

Oxygen Air Argon

3.00 2.00 1.00 0.00 Annealed

Fig. 1. Mean mass loss of Fe30Ni20Mn20Al30 pins after 1 km sliding tests with a 23N load in oxygen, air or argon. Three tests were performed in each environment. Error bars signify standard deviations.

of the alloy (Cu-15 wt.% Ni-8 wt.% Sn) and w 1 mm fine debris consisting of a mixture of (Fe, Cr)2O3 and (Cu, Ni, Sn)O oxides [14]. The subsurface microstructure of the wear pin was characterized in detail by employing various TEM analytical techniques, including high angle annular dark field imaging, electron energy loss spectroscopy and an energy dispersive spectrometry (EDS), in addition to conventional diffraction imaging and phase identification by diffraction contrast. Immediately below the surface was a 50 nm thick, porous and Cu-depleted sublayer formed by various complex reactions. Below this was a mechanically mixed layer (MML) that was several microns thick consisting of an equi-axed nanocomposite of a Cu-rich metallic phase and (Fe, Cr)2O3-based oxide particles, which was formed by compaction and mixing of wear debris. Below the MML was a severely plastically deformation layer (SPDL) consisting of elongated nanograins of a CueNieSn solid solution [15]. Recently, wear tests were performed in air at different loads on Cu-15 wt.% Ni-8 wt.% Sn in both the spinodal hardened and the solutionized softer state [16]. As the load increased, the wear rate increased, while the friction coefficient decreased. Archard’s law was obeyed, i.e. the higher the hardness, the lower the wear rate. Again, an SPDL was present, which was characterized by a significant crystallographic texture.

117

The oil-lubricated wear behavior of Cu-15 wt.% Ni-8 wt.% Sn has also been examined in air of as a function of annealing time using a block-on-ring test at a load of 1470 N. The hardness increased to a peak of w300 HV from w100 HV in the as-cast state after 2 h annealing at 400  C, then decreased to w 200 HV after further annealing. This hardness behavior resulted from changes in the microstructure, i.e. the amount of cellular structured equilibrium a and g phases, after different annealing times. The minimum wear rate occurred at the maximum hardness. In addition to abrasive wear, the dominant wear mechanism was that material was periodically extruded by the harder asperities, which then broke off and became wear debris [17]. Although the above work was performed on spinodal alloys, they were focused on the CueNieSn system and the wear behavior was not clearly correlated with changes in the microstructure. In addition, although those works showed that the wear rate was reduced by 75% in Ar [14], no further study and discussion about the effect of environment were performed. Recently, a range of nanostructured FeNiMnAl alloys were discovered that appear to have formed by spinodal decomposition. Their microstructures and mechanical properties have been studied in some detail [18e26]. The FeNiMnAl alloys have high aluminum contents, which presumably will provide excellent oxidation resistance to high temperature. Also, these alloys retain the high strength up to 500  C [26]. Because of their good strength at both room temperature and elevated temperature, those alloys could be useful in tribological applications. Thus far, there has been no research on the wear of these FeNiMnAl alloys. This study focused on the dry sliding wear of one of these FeNiMnAl alloys, i.e. Fe30Ni20Mn20Al30 (in at.%). The objective of this study was to understand how the wear behavior of a FeNiMnAl spinodal alloy varies with different environments and changes in the microstructure produced by ageing, and hence, the wear behavior of nanostructured FeNiMnAl alloys as a class of materials. 2. Material The alloy consists of B2 (ordered b.c.c.) and L21 (further ordered B2) phases, which exist in a cube-on-cube relationship. The wear behavior was examined in both the as-cast state, when the phase width was w5 nm, and after a 72 h anneal at 550  C, when the phase width had coarsened to w20 nm [26]. The Vickers hardness increased from 514  7 HV for the as-cast alloy to 547  6 HV for the 72 h annealed alloy, i.e. about a 6% increase. Both the as-cast and annealed alloy showed brittle fracture at room temperature with very similar fracture strengths in compression of 1350 MPa. At 300  C, the annealed alloy showed a yield strength of w1450 MPa and a failure stress of w2150 MPa, while the as-cast alloy still showed a brittle fracture before yielding at this temperature with a similar fracture strength found at room temperature. The room temperature fracture toughness, as determined from notched three point bend tests, was the same as that of annealed alloy [26].

Table 1 Summary of wear tests results from pin-on-disk wear tests on as-cast and 72 h annealed Fe30Ni20Mn20Al30 pins conducted in oxygen, air and argon. As-cast

Pin average mass loss (mg) Pin volumetric wear rate (mm3/km) Disk mean depth of the wear track (um) Disk mean volumetric wear (mm3)

4.5 0.7 3.4 4.9

72 h annealed Air

O2    

1.0 0.1 0.8 0.2

3.6 0.6 2.4 3.8

Ar    

0.6 0.1 1.5 3.1

1.4 0.2 2.6 2.6

O2    

0.8 0.1 0.2 0.2

2.8 0.4 3.8 5.0

Air    

0.3 0.1 0.5 0.3

2.3 0.3 0.2 0.3

Ar    

0.9 0.1 0.1 0.1

0.7 0.1 1.1 1.1

   

0.3 0.1 1.4 1.5

118

X. Wu et al. / Intermetallics 23 (2012) 116e127

Fig. 2. Typical profilometer traces across the wear track on zirconia disk for (a) as-cast and (b) 72 h annealed samples worn in argon, air or oxygen.

Fig. 3. Optical micrographs showing morphologies of the wear tracks on the zirconia disk after sliding performed in (a, b) oxygen and (c, d) argon. The arrow in (b) indicates the cracks. The uniform scars on both left and right sides in (a) and (c) had been left on the zirconia disk originally after polishing.

X. Wu et al. / Intermetallics 23 (2012) 116e127

Ingots of Fe30Ni20Mn20Al30 were prepared by arc-melting the constituent elements of >99.9% purity in a water-chilled copper crucible under argon. The ingots were 5 cm diameter, w50 g buttons. The ingots were flipped over and melted three times to ensure a homogeneous mixture, and cut into small pieces using a high speed abrasive saw. The pieces, whose positions were random in the button ingots, were arc-melted again in a waterchilled copper mold under argon into w9.5 mm diameter, w1 cm long cylindrical wear pins with a hemispherical tip. All the pins should have the same composition. Some were annealed at 550  C in air for 72 h, followed by air cooling. All pins were ground and polished to a mirror finish.

119

4. Results 4.1. Wear rate of Fe30Ni20Mn20Al30 pins The results of the wear tests are summarized in both Fig. 1 and Table 1. The mean mass losses after 1 km of sliding in oxygen-containing environments (i.e. air and oxygen) were about three times more than that in an oxygen-free environment (i.e. argon) for both as-cast and 72 h annealed pins. However, it should be noted that the errors bars overlap between tests in air and oxygen. Thus, the effect of water vapor in the air, which could affect the tests, is not completely clear. Compared with 72 h annealed pins, the as-cast pins lost 62% more mass for wear

3. Experimental Pin-on-disc wear tests were performed against an yttriastabilized zirconia counterface polished to a surface finish of w 0.1 Ra (w 0.0254 mm), using a home-made device [27]. All tests were conducted on the new disk surface at room temperature (w25  C) at a constant sliding speed of 1 m s1 for a total sliding distance of 1 km with a normal load of 23 N in three different environments, i.e. oxygen, air, argon. The humidity of oxygen and argon was less than 5 ppm, while the humidity of the air was w45%. Three tests were performed in each environment. Debris was collected during the wear tests using adhesive tape wrapped around the outside of the zirconia disk. The diameter of the zirconia disk was 10 cm. The diameters of all the wear tracks are listed in Appendix 3. Wear mass loss was determined by measuring the mass difference of the pins before and after using an electronic balance of 0.1 mg precision. The volumetric wear was converted from mass loss divided by the density of the alloy, 6491 kg m3, as determined using Archimedes’ law. The phases present in the debris and on the wear surface of the pins were analyzed using a Rigaku D/Max 2000 XRD with Cu Ka radiation operated at 40 kV, 300 mA. Measurements were performed by step scanning 2q from 10 to 120 with a step size of 0.02 . A count time of 1 s per step was used, giving a total scan time of w 1.5 h. The morphology of the debris and the wear surfaces of the pins were examined using both secondary electron (SE) and backscattered electron (BSE) imaging on an FEI Field emission gun XL-30 scanning electron microscope (SEM) operated at 15 kV, equipped with an EDAX Li-drifted EDS. Cross-sectional TEM specimens, used to examine the subsurface of the wear pins, were prepared using a Fei Nova 200 Nanolab FIB using the lift-out method [28]. The specimens were examined using a Philips CM200 TEM operating at 200 kV. Elemental X-ray maps were collected in a scanning transmission electron microscope (STEM) mode using EDS. The worn tracks of the zirconia disk were analyzed using a Bendix ProficorderÒ Linear Profile System with a Data TranslationÒ USB data acquisition function module attachment connected with a computer to record the profile readings. The profilometer data gave the depth readings for a cross section of the track: negative values indicate that material was removed from the surface, while positive values indicate that the surface of the wear track is built up. The mean depth of the wear track was the average depth of all the points within the width of the wear track. The cross-sectional area was obtained by multiplying the mean depth and the width of the wear track. The product of the cross-sectional area and the circumstance of the wear track was the mean volumetric wear of the zirconia disk. The zirconia counterface was examined after wear tests using a Zeiss IM 35 optical microscope.

Fig. 4. Secondary electron images of a) The debris and b) The worn surface of the ascast Fe30Ni20Mn20Al30 pin after wear tests in air. c) EDS spectrum showing the composition of the debris.

120

X. Wu et al. / Intermetallics 23 (2012) 116e127

Fig. 5. X-ray diffraction pattern of debris collected during a wear test of an annealed Fe30Ni20Mn20Al30 pin tested in oxygen.

Fig. 6. X-ray diffraction patterns of the worn surfaces of as-cast Fe30Ni20Mn20Al30 pins tested in different environments. The pattern from the worn surface of the pin tested in oxygen shows stronger zirconia peaks than that from the pin tested in air, which in turn shows stronger peaks than the pin tested under argon.

X. Wu et al. / Intermetallics 23 (2012) 116e127

tests performed in oxygen, 59% more in air, and 91% more in argon. The as-cast Fe30Ni20Mn20Al30 showed comparable wear resistance to as-cast AleSi tested under the same conditions, whose mass losses were 2.2  0.3 mg in argon, 5.5  0.5 mg in air and 3.3  1.7 mg in oxygen [26]. And the 72 h annealed Fe30Ni20Mn20Al30 showed better wear resistance than as-cast AleSi.

4.2. Examination of zirconia counterface Profilometer results of the wear tracks on the zirconia disk are shown in Fig. 2 and Table 1. After tests of the 72 h annealed samples conducted in either air or argon the zirconia disk showed much less wear, compared with the tests of the as-cast samples in the same environments. In contrast, the wear tracks on the zirconia disk showed significant wear after the tests conducted in oxygen for both the as-cast and 72 h annealed samples. The morphologies of the wear tracks on the zirconia disk were examined using optical microscopy. The wear tracks after tests of the as-cast pins tested in oxygen exhibit intensive wear and numbers of cracks w 50 mm in length, which are perpendicular to the sliding directions, as shown in Fig. 3(a) and (b). However, a transfer layer, without any observable cracks, was generated on the zirconia disk after the tests conducted in argon, as shown in Fig. 3(c) and (d). Because of the size limitation of the SEM stage, it was impossible to check the chemistry of this layer using EDS or the morphologies at higher magnification.

121

4.4. Tribolayer A “Tribolayer” or “mechanically mixed layer” underneath the worn surface after sliding wear is usually associated with deformation, compaction, and fragmentation [29,30]. SE images of cross-sections produced by FIB from the pins worn in air are shown in Fig. 9. The as-cast specimen (Fig. 9(a)) showed three layers starting from the surface, i.e. a porous layer produced by debris compaction, the thickness of which various from place to place (labeled A); a severely deformed layer, w 500 nm in the thickness, in which debris are embedded and grain refinement may have occurred (labeled B); a slightly deformed layer, which is mainly the original specimen material (labeled C). In contrast, the annealed specimen in Fig. 9(b) shows a layer similar to Layer A in Fig. 9(a) followed by a relatively homogeneous layer beneath. A bright field TEM image and corresponding selected-area diffraction (SAD) patterns of the as-cast pin worn in air are shown in Fig. 10. The outermost layer mainly consisted of ZrO2 (Fig. 10(a)) while the layer below consists of severely deformed B2 and L21 phases that show grain refinement (Fig. 10(a)). Fig. 11 show X-ray elemental maps and the corresponding STEM image of the as-cast pin worn in argon. ZrO2 particles were embedded well below the worn surface, indicating the particles shed from the disk have been mechanically mixed into the tribolayer. This feature was observed

4.3. Debris and the wear surface The debris from the wear tests consists of irregular flakes mixed with small broken pieces, as shown in Fig. 4 (a). The size of debris (w50 mm) matches with the width of the long grooves generated by severe plowing in Fig. 4 (b), indicating the existence of either abrasive wear or delamination during the sliding. The composition of the debris is zirconia mixed with pin material, as indicated in the EDS spectrum in Fig. 4 (c). An XRD pattern of the debris collected during a wear test on an annealed pin tested under oxygen is shown in Fig. 5. All the clearly identifiable peaks are from zirconia, which is consistent with the EDS analysis. Similar results were obtained for all tests in air and oxygen. It was not possible to perform a similar analysis for the tests of both as-cast and annealed pins tested under argon due to the limited quantity of debris. XRD patterns of the worn surfaces of the pins after sliding tests performed in different environments are shown in Fig. 6. For the pin worn under argon, most of the peaks are from the B2 and L21 phases of Fe30Ni20Mn20Al30 alloy. Only a few very weak ZrO2 peaks are identifiable. Peaks from both the B2 and L21 phases and from ZrO2 are found in the patterns of the pins worn in oxygencontaining environments (i.e. air and oxygen). The morphologies of the worn surfaces of the pins worn in air show features of abrasive wear, e.g. long parallel grooves induced by plowing, cracks, and wear pits from material pullout, indicated as A, B, and C in Fig. 7. There is also evidence of plastic deformation in Fig. 7(a), which could result from adhesive mechanisms. The wear scars on the pin tested in argon (Fig. 8 (a)) were much smaller than those tested in air (Fig. 8 (c)) and oxygen (Fig. 8 (e)), which is again consistent with the wear loss results. Wear pits and plastic deformation can be observed in the higher magnification images (Fig. 8 (b), (d) and (f)) of the worn surfaces after tests in all three environments.

Fig. 7. Back-scattered electron image of the worn surface of an annealed Fe30Ni20Mn20Al30 pin after a wear test in a) air and b) oxygen, showing A) plowing grooves, B) cracks, and C) Wear pits.

122

X. Wu et al. / Intermetallics 23 (2012) 116e127

Fig. 8. Back-scattered electron images of the worn surface of the as-cast Fe30Ni20Mn20Al30 pins after wear tests in a, b) argon, c, d) air and e, f) oxygen. Higher magnification images b), d) and f) show plastic deformation and wear pits.

after dry-sliding wear of AleSi [31]. The tribolayers of all the worn pins showed a similar structure and composition. 5. Discussion 5.1. The Effects of environment on the wear mechanisms In spite of the differences in microstructure and mechanical properties [26], the wear processes for both the as-cast and the

annealed Fe30Ni20Mn20Al30 wear pins against zirconia appears to be very much dependent on the environment, especially the existence of oxygen, in which the dry sliding tests were carried out. Wear in oxygen-containing environments, i.e. air and oxygen, was much more severe than that in an oxygen-free environment, i.e. argon. This result is consistent with the results from Bellon’s group, who found the same behavior for dry sliding wear of Cu-15wt% Ni8wt% Sn tested in air or argon [14]. The explanation for this phenomenon could be that the oxide surface film is continuously

X. Wu et al. / Intermetallics 23 (2012) 116e127

breaking and forming, thus leading to a larger wear rate in oxygencontaining environments. During sliding in all three environments, the initial asperities experienced a large shear stress. Highly localized shearing occurs in conjunction with the nucleation and propagation of cracks along which the asperities detach. Shear fracture of the asperities generate the debris, which could have acted as third bodies in the wear process. Delamination may also play a role in the sliding process. This involves the nucleation of subsurface cracks and their propagation parallel to the surface. The cracks originate within the highly deformed material beneath the surface, possibly from voids or defects, and eventually grow up to the free surface [32]. The material detached from the pin transfers to the counterface and mechanically mixes to form the outermost layer of the tribolayers. The tribolayer certainly show heavily-deformed regions, including grain refinement. The morphological features of the worn surfaces of the pins, such as plowing, cracking and wear pits (Figs. 7 and 8), all confirm that abrasive wear is the dominant wear mechanism in all three different environments. It is worth noting that plastic deformation was also observed in the wear, which indicates that adhesive wear also happened. However, in oxygen-containing environments, the wear process is abetted by oxidation, the rate of which is strongly dependent on temperature. Due to the frictional heating, the local temperature at the sliding interface is substantially higher than the ambient temperature and may be enhanced at the asperity contacts by transient “flashes” or “hot spots”. An approximate analysis of the contact temperature present at Fe30Ni20Mn20Al30 sliding interface was performed assuming a Hertzian contact area [31]. As calculated in the Appendix 1, because of the low thermal conductivity of the Fe30Ni20Mn20Al30 the contact temperature could easily reach 478  C, and localized “flash” temperatures may rise even higher in concentrated asperity contacts; these temperatures would be sufficient to cause oxidation. The hard oxides on the pin surface could act as abrasive asperities and abrade the zirconia counterface, producing zirconia particles in the debris. At the same time, the brittle and hard oxides break off from the pin surface and tumble in the space between the surfaces. Both zirconia and other oxides particles become abrasive third bodies and keep removing material from both the pin and the zirconia disk. Since the third bodies in oxygen-containing environments (i.e. zirconia and other metal oxides) were harder than those in an oxygen-free environment (i.e. Fe30Ni20Mn20Al30 particles fractured from the pins), the wear losses in oxygen-containing environments were larger than those in an oxygen-free environment.. Moisture in the test environment can also be an important factor affecting wear behavior [33,34]. The effect of moisture on the wear of Fe30Ni20Mn20Al30 is not clear. Although the relative humidity in air was considerably higher than that in dry oxygen, it is hard to differentiate the wear rate of the pins tested in oxygen and air, due to the large scatter in the results. Thus, it is unclear whether the wear of Fe30Ni20Mn20Al30 is sensitive to moisture.

123

However, the wear of the 72 h annealed material is about 50% less than the as-cast material while the hardness of 72 h annealed material is only about 5% greater than the as-cast material. Thus, Archard’s law cannot fully explain all the observed phenomena. In earlier work, the as-cast alloy was shown to have a twophase B2/L21 microstructure with each phase having a width of w5 nm. After a 72 h anneal, the phases were shown to have coarsened to w20 nm [27]. Thus, the number of interfaces between the phases decreased from the as-cast alloy to the annealed alloy. Interfaces are often where defects, such as cracks, initiate. A greater number of interfaces could lead to more crack nucleation and, possibly, easier propagation. Since cracking plays an important role in the wear processes, the coarser phases and fewer interfaces in the annealed alloy could be the origin of the greater wear resistance. It is also worth noting that the 72 h annealed alloy shows better ductility than the as-cast alloy at

5.2. The Effects of microstructure on the wear mechanisms The results here show that the 72 h annealed Fe30Ni20Mn20Al30 shows greater wear resistance than the as-cast material in all environments. It is well-established that wear behavior of metals usually follow Archard’s law, i.e. the total wear volume has an inverse relationship with a material’s hardness [35]. Thus, the hardness increase of the annealed alloy contributes to the better wear resistance.

Fig. 9. Secondary electron images from pits in a) as-cast and b) annealed specimens wear-tested under air from which TEM specimen was removed using FIB. The Pt stripe was deposited on the top to protect the surface while machining with the FIB. A, B, C indicate different layers.

124

X. Wu et al. / Intermetallics 23 (2012) 116e127

Fig. 10. (a) Bright field TEM image of tribolayer in an as-cast Fe30Ni20Mn20Al30 pin worn in air; (b) Selected-area diffraction pattern from layer A showing ZrO2; (c) Selected-area diffraction pattern from layer B showing the matrix B2 and L21 structure.

elevated temperatures [27]. Although both alloys show the brittle fracture at room temperature, the material at the contact area could behave in a ductile fashion since the contact temperature is raised due to the frictional heating. 5.3. Tribolayer During the dry sliding of Fe30Ni20Mn20Al30 alloy, a complex tribolayer has formed, as shown in the Results section. The outmost layer is the so-called “debris layer” [14], which is formed as a capping layer on the pin surface during wear. This layer is formed very early, but continuously destroyed and reformed. It transmits

not only the load, but also the materials to the next layer, a mechanically mixed layer. The observation of embedded zirconia particles suggested an intermittent transfer mechanism, rather than a continuous one, i.e. similar to observations on Cu-15 wt%Ni8%Sn [14]. Two possible mechanisms have been suggested for the formation of the deformed mechanically mixed layer observed after the dry sliding of as-cast Fe30Ni20Mn20Al30 alloy: nucleationcontrolled crystallization and plastic deformation [35]. Since all the wear tests were carried out at room temperature, the temperature is still too low for recyrstallization, even taking the contact temperature into account. Thus, plastic deformation may be the suitable explanation in this case.

X. Wu et al. / Intermetallics 23 (2012) 116e127

125

Fig. 11. X-ray maps using Zr, O, Fe, Ni, Mn and Al peaks from the region shown in the bright field STEM image of an as-cast Fe30Ni20Mn20Al30 pin worn in argon.

6. Conclusions The dry sliding wear of both as-cast and annealed nanostructured two-phase Fe30Ni20Mn20Al30 was studied using pin-ondisk tribotests at a constant sliding speed of 1 m s1 for a total

sliding distance of 1 km with a normal load of 23 N against a zirconia disk in three different environments, viz., oxygen, air and dry argon, and the resulting mass losses from the pins and disk were determined. In order to elucidate the wear mechanisms, the wear pins and debris were examined using a combination of XRD,

126

X. Wu et al. / Intermetallics 23 (2012) 116e127

an SEM and a (S)TEM including EDS, and the worn disk was examined with both optical microscopy and a profilometer. It was found that: 1. The wear rate was more sensitive to oxygen than to moisture in the environment. The wear in an oxygen-free environment (i.e. argon) was less than that in the oxygen-containing environments (i.e. air and oxygen). 2. 72 h annealed material showed improved wear behavior compared to the as-cast alloy, because of the higher hardness and the better ductility resulting from the coarser annealed microstructure. 3. Both two-body and three-body abrasive wear, as well as deformation and delamination, take place in the wear process. In an oxygen-free environment, the third bodies were Fe30Ni20Mn20Al30 materials fractured from the pins. However, in oxygen-containing environments, the third bodies were both zirconia and other oxides particles, which would be harder than Fe30Ni20Mn20Al30 particles. Thus, more wear loss happened in oxygen-containing environments than in an oxygen-free environment. 4. Zirconia particles were embedded in the tribolayer, suggesting a transfer mechanism probably caused by plastic deformation. Acknowledgements This research was supported by U.S. Department of Energy (DOE), Office of Basic Energy Science Award #DE-FG02-10ER46392. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing official policies, either expressed or implied of the DOE or the U.S. Government. The authors would like to acknowledge the assistance from Evan Zeitchick, an undergraduate at Dartmouth College, and Dr. Charles Daghlian, Director of the Dartmouth Electron Microscope Facility. Appendix 1. Calculation of contact area and contact temperature

1:31ampV DTmax ¼ pffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 453 p K1 1:2344 þ Pe1 þ K2 1:2344 þ Pe2 where Peclet number Pe1 ¼ V1 ar1 C1 =2K1 ¼ 116, Pe2 ¼ V2 ar2 C2 =2K2 ¼ 0. Given that the tests were conducted at room temperature (25  C), the surface temperature at the center of the contact area was 478  C, which is a relatively conservative estimate. The contact region of the stationary pins would remain at an elevated temperature for a substantial period of time, which gives oxides sufficient time to form [38]. Note that sparking happened frequently during the wear tests, probably because of encountering concentrated asperities in the contact region. The flash temperature rise could be considerably higher for a short period of time. Appendix 2. Measurement of material properties of as-cast Fe30Ni20Mn20Al30 Hardness H was determined using a Leitz MINIload tester at a load of 200 g and a drop time of approximately 13 s followed by a settle time of approximately 60 s. Reported values are the average of 10 measurements. The modulus of elasticity E was measured using a Hysitron Uni1 nanoindenter with a Berkovich-type diamond indenter at a load of 1 mN. The themal conductivity l ¼ 418.68  a  CP  r, where a is thermal diffusivity 0.024 cm2 s1; CP is specific heat capacity 0.128 cal g1 K1; and r is density 6.491 g cm3. Both thermal diffusivity and specific heat capacity were measured by using a JR-3 laser flash diffusivity instrument. The density was determined based on Archimedes’ principle. An A&D HR200 precision analytical balance scale was used to measure the mass of the pins and the mass of the water the pins displaced. Appendix 3. The diameters of the wear tracks for tests in different environments and for different states of the material.

Contact area of stationary wear pins with moving zirconia disk.

Operating conditions: Normal load w ¼ 23 N Average friction coefficient (measured) m ¼ 0.2

H Hardness (GPa) E Modulus of Elasticity (GPa) r Density (kg/m3) U Poisson’s ratio K Thermal Conductivity (W/m K) C Specific heat (J/kg K)

Sliding velocity V ¼ 1 m/s Pin radius r ¼ 4.59 mm Zirconia (material 1)

As-cast Fe30Ni20Mn20Al30 (material 2)

12.7 290 6100 0.24 1.8 630

5a 160a 6491a 0.33 8.3a 538a

Material properties. a Measured in Appendix 2, remainder from [36], subscript 1 and 2 mean material 1 (Zirconia) and material 2 (As-cast Fe30Ni20Mn20Al30), respectively.

Contact geometry (assuming Hertzian contact [1]). Radius of contact circle a ¼ ð3wr=4E0 Þ1=3 . Where 1=E0 ¼ 1  v21 =E1 þ 1  v22 =E2 . Contact temperature rise due to frictional heating for the Hertzian elastic contact (following methodology of [37]). Assume moving flat Zirconia disk (material 1) and stationary pin (material 2).

Environment

State of material

O2 Air Argon

5.0 5.0 5.2

As-cast (cm) 6.4 7.5 5.7

72 h Annealed (cm) 8.4 8.6 6.2

6.2 4.4 4.6

6.9 8.0 5.4

7.3 9.2 8.2

References [1] Hutchings IM. Tribology, friction and wear of engineering materials. Boca Raton, FL: CRC Press; 1992. p. 14e16. [2] Lim SC, Ashby MF. Overview No. 55 wear-mechanism maps. Acta Metallurgica 1987;35:1e24. [3] Lim SC, Ashby MF, Brunton JH. Wear-rate transitions and their relationship to wear mechanisms. Acta Metallurgica 1987;35:1343e8. [4] Ashby MF. The development of wear-mechanism maps, the modeling of wear in energy conversion and utilization technologies international workshop on wear modeling. Argonne National Lab; 1989:25e59. [5] Ashby MF, Lim SC. Wear-mechanisms maps. Scripta Metallurgica et Materialia 1990;24:805e10. [6] Lim SC. Recent developments in wear-mechanism maps. Tribology International 1998;31:87e97. [7] Cahn JW. Hardening by spinodal decomposition. Acta Metallurgica 1963;11: 1275e82. [8] Cribb WR, Ratka JO. Copper spinodal alloys. Advanced Materials and Processes 2002;11:1e4.

X. Wu et al. / Intermetallics 23 (2012) 116e127 [9] Rudy E. Spinodal carbonitride alloys for tool and wear applications, United States Patent 3971656. 1974. [10] Rudy E, Worcester S, Elkington W. Modified spinodal alloys for tool and wear applications, Plansee Seminar 8th, Reutte Austria, May 27e30, 1974. [11] Marczak RJ, Ciach R. Tribological properties of the concentrated Al-Zn alloys, First European Tribology Congress. 1973, p. 223e36. [12] Kar NJ. Copper-based spinodal alloy, United States Patent 4641976. 1987. [13] Weninger J, Bares JA, Tyler T, Ye J, Bellon P. Microstructural evolution in Cu-spinodal alloy driven by high-energy ball milling and dry sliding wear. In: Howe JM, Laughlin DE, Lee JK, Dahmen U, Soffa WA, editors. Proceedings of an international conference on solid-to-solid phase transformation in inorganic materials. Diffusional transformation, vol. 1. TMS; 2005. p. 887e91. [14] Singh JB, Cai W, Bellon P. Dry sliding of Cu-15wt%Ni-8wt%Sn bronze: wear behavior and microstructures. Wear 2007;263:830e41. [15] Singh JB, Wen JG, Bellon P. Nanoscale characterization of the transfer layer formed during dry sliding of Cu-15wt%Ni-8wt%Sn bronze alloy. Acta Materialia 2008;56:3053e64. [16] Cai W, Mabon J, Bellon P. Crystallographic textures and texture transitions induced by sliding wear in bronze and nickel. Wear 2009;267:485e95. [17] Zhang SZ, Jiang BH, Ding WJ. Wear of Cu-15Ni-8Sn spinodal alloy. Wear 2008; 264:199e203. [18] Hanna JA, Baker I, Wittman MW, Munroe PR. A new ultra-high-strength spinodal alloy. Journal of Materials Research 2005;20:791e5. [19] Baker I, Hanna JA, Wittmann MW, Munroe PR. Microstructure of a spinodal Fe-Ni-Mn-Al alloy. Microscopy and Microanalysis 2005a;11:1864e5. [20] Baker I, Hanna JA, Wittmann MW, Munroe PR. The mechanical properties of Fe30Ni20Mn25Al25, processing and fabrication of advanced materials XIV with frontiers in materials science: innovative materials and manufacturing techniques. 2005b, p. 237e48 [21] Wittmann MW, Baker I, Hanna JA, Munroe PR. Microstructure and mechanical properties of Fe-Ni-Mn-Al alloys, integrative and interdisciplinary aspects of intermetallics. Material Research Society Symposium Proceedings 2005;842: S5e17. p. 35e40. [22] Saxey DW, Hanna J, Zheng R, Marceau RKW, Baker I, Ringer SP. Nanostructural analysis of advanced alloys in a local electrode atom probe. Microscopy and Microanalysis 2005;11:872e3.

127

[23] Loudis JA, Baker I. Dislocation identification and in situ straining in the spinodal Fe30Ni20Mn25Al25 alloy. Microscopy Research and Technique 2008;71: 489e96. [24] Baker I, Zheng RK, Saxey DW, Kuwano S, Wittmann MW, Loudis JA, et al. Microstructural evolution in the spinodal alloy Fe35Ni15Mn25Al25. Intermetallics 2009;17:886e93. [25] Wu X, Baker I, Miller MK, More K. Microstructure and mechanical behavior in spinodal Fe35Ni15Mn25Al25 alloy. Microscopy and Microanalysis 2009;15: 116e7. [26] Wu X, Baker I and Wu H, Microstructure and mechanical behavior of twophase Fe30Ni20Mn25Al25 alloy, in preparation. [27] Johnson BJ, Kennedy FE, Baker I. Dry sliding wear of NiAl. Wear 1996;192: 241e7. [28] Giannuzzi LA, Stevie FA. A review of focused ion beam milling techniques for TEM specimen preparation. Micron 1999;30:197e204. [29] Rigney DA. Transfer, mixing and associated chemical and mechanical processes during the sliding of ductile materials. Wear 2000;245:1e9. [30] Li J, Elmadagli M, Gertsman VY, Lo J, Alpas AT. FIB and TEM characterization of subsurfaces of an AleSi alloy (A390) subjected to sliding wear. Materials Science and Engineering A 2006;421:317e27. [31] Baker I, Sun Y, Kennedy FF, Munroe PR. Dry sliding wear of eutectic Al-Si. Journal of Materials Science 2010;45:969e78. [32] Hutchings IM. Tribology, friction and wear of engineering materials. Boca Raton, FL: CRC Press; 1992. p. 98e103. [33] Yen BK, Ishihara T. Effect of humidity on friction and wear of AleSi eutectic alloy and AleSi alloy-graphite composites. Wear 1996;198:169e75. [34] Gwaze M, Baker I, Kennedy FE, Munre PR. Effect of sliding environment on dry sliding wear of as-cast eutectic AleSi. Journal of Materials Science 2010;45: 6849e52. [35] Archard JF. Contact and rubbing of flat surfaces. Journal of Applied Physics 1953;24:981e8. [36] Bhushan B, Gupta BK. Handbook of tribology. New York: McGraw-Hill; 1991. p. 4.57. [37] Kennedy FE. In: Bhushan B, editor. CRC handbook of modern tribology. Boca Raton, FL: CRC Press; 2001. p. 235e72. [38] Razavizadeh K, Eyre TS. Oxidative wear of aluminium alloys. Wear 1982;79: 325e33.