Effect of martensite content on friction and oxidative ... - Springer Link

Effect of Martensite Content on Friction and Oxidative Wear Behavior of 0.42 Pct Carbon Dual-Phase Steel RAJNESH TYAGI, S.K. NATH, and S. RAY In order to explore the tribological potential of the dual-phase (DP) steel as a wear-resistant material, the friction and wear characteristics have been investigated for this steel with varying amounts of martensite from 42 to 72 vol pct, developed by varying holding time at the intercritical annealing temperature of 740 ⬚C. Dry sliding wear tests have been conducted on DP steels containing 0.42 wt pct carbon using a pin-on-disk machine under different normal loads of 14.7, 19.6, 24.5, 29.4, and 34.3 N and at a constant sliding speed of 1.15 m/s. Weight loss has been measured at different time intervals on the same specimen. The variation of cumulative volume loss with sliding distance has been represented by two linear segments signifying the run-in and the steady state of wear. The mechanism of wear is primarily oxidative in nature, which has been confirmed by X-ray diffraction patterns of the wear debris generated during sliding. The wear rate varies linearly with load in both the run-in and the steady state. The wear rate decreases linearly with increasing volume fraction of martensite in DP steels reflecting the effect of hardness imparted by the increasing amount of martensite, which is a hard and load bearing phase. The average coefficient of friction also decreases linearly with increasing load as well as with increasing martensite volume fraction. In the run-in stage, the wear coefficient does not change significantly between DP1 and DP2 steels containing 42 and 51 vol pct martensite, respectively, but decreases sharply as one moves from DP2 to DP4 containing, respectively, 51 and 72 vol pct martensite. But in the steady state, the wear coefficient decreases almost linearly with increasing volume fraction of the martensite. The decrease in wear coefficient may be attributed to the decreasing wear rate dominating over the decrease in real area of contact due to increasing hardness.

I. INTRODUCTION

DUAL-PHASE (DP) steels consist of hard martensite islands embedded in a relatively soft and ductile matrix of ferrite.[1] In view of the potential of DP steel as a wearresistant material, it has already been employed as wearresistant material in the field of mineral processing, mining, and pipeline transportation of slurry.[2] Although a few studies[3,4] have been reported on the tribology of DP steels, there is a need to develop a comprehensive understanding of their tribological behavior. Wayne and Rice[3] have shown the dependence of wear on microstructure and concluded that a duplex microstructure of DP steel offers higher wear resistance than that observed in a steel with spheroidal carbides. It has also been indicated that the volume fraction of martensite and ferrite in DP steel is an important factor in wear resistance. But no systematic investigation has been undertaken to study this effect over a range of martensite content in a given steel. Sawa and Rigney[4] have found that the wear behavior of DP steel also depends strongly on its morphology, i.e., shape, size, and distribution of martensite. Basak et al.[5] have also reported that wear resistance of DP steels of different carbon contents increases with increasing volume fraction of martensite. The present study has been carried out in DP steels developed by varying the holding time of medium-carbon steel of a given carbon content at RAJNESH TYAGI, Research Scientist, S.K. NATH, Associate Professor, and S. RAY, Professor, are with the Metallurgical and Materials Engineering Department, Indian Institute of Technology Roorkee, Roorkee - 247667, India. Contact e-mail: [email protected] Manuscript submitted January 17, 2002. METALLURGICAL AND MATERIALS TRANSACTIONS A

the intercritical temperature of 740 ⬚C, in order to attain different martensite volume fractions and to examine its effect on the friction and wear behavior of these DP steels.

II. EXPERIMENTAL PROCEDURE Cylindrical pin samples (30 mm ⫻ 4.0 mm ␾ ) of commercial grade steel containing C-0.42, Mn-0.62, Si-0.15, S-0.04, and P-0.04 wt pct were used for the current investigation. All the samples were normalized at 860 ⬚C for 20 minutes. A number of these samples were intercritically annealed to develop four different DP structures of martensite and ferrite. Intercritical annealing was conducted in a vertical tube furnace at 740 ⬚C for 2.0, 2.5, 3.0, and 3.5 minutes followed by water quenching, and the resulting samples of different martensite contents are designated as DP1, DP2, DP3, and DP4, respectively; the details have been described earlier.[6] Metallographic structures were analyzed to determine volume fractions of martensite in DP steels by a point-counting method.[7] After polishing to 4/0 grade emery papers, the macrohardness of these samples was measured at a load of 30 kg (294 N) in a Brinell hardness tester, while the microhardness was measured on the etched microstructure using 0.01-kg (0.098-N) load in Vickers scale. Wear tests were conducted using pin samples having flat surfaces in the contact regions and rounded corners, polished up to 4/0 grade (⬃38 ␮m) emery paper and cleaned with acetone to remove dust and grease from the surface of the pin. Dry sliding wear tests were carried out against the counterface of a hardened and polished disk made of En32 steel having HRC 62 to 65 hardness at a relative humidity VOLUME 33A, NOVEMBER 2002—3479

of 40 to 60 pct at room temperature of 25 ⬚C. A sturdy pinon-disk machine described earlier[8] was used to carry out the wear tests. Pin weight losses were measured at different intervals of time. Weight loss data were converted to volume loss using steel density of 7760 kg/m3. The procedure adopted for the wear testing has been described earlier.[8] Each test at a given load and sliding velocity was repeated 3 times with identical new samples on a fresh disk surface, and the average data for volume loss after each interval of time were used for the analysis of wear rate. Samples of all four DP steels were tested at loads of 14.7, 19.6, 24.5, 29.4, and 34.3 N and at a fixed sliding speed of 1.15 m/s. Pin wear surfaces were cleaned of wear debris and visually examined. The wear surfaces as well as the subsurfaces of the pin specimens were examined under LEO, 435 VP scanning electron microscope (SEM). The rise in temperature of the pin samples was measured during wear with a fine chromel-alumel thermocouple brazed on the pin side about 3 mm above the wear contact surface, and the temperature at the location was monitored with a millivolt meter. The wear debris was collected periodically during wear tests for X-ray diffraction analysis. Debris for the first 80 minutes (first segment, i.e., run-in) was analyzed separately from the debris of longer periods (second segment, i.e., steady state). An iron target was used for this study.

(a)

III. RESULTS AND DISCUSSION A. Characterization of Pin Samples The microstructure of the DP steel has dark and white areas when etched with 2 pct Nital, as shown in Figures 1(a) and (b). The structure of DP steel, DP1, obtained after 2 minutes of intercritical annealing has dark etching martensite and white regions of ferrite, as shown in Figure 1(a). There could be a small amount of retained austensite associated with the martensite, but it is not visible in the micrograph. At higher magnification, martensite needles in the dark areas could be resolved, as indicated by an arrow in Figure 1(b). Similar microstructural features have also been observed in DP2, DP3, and DP4 steels obtained after 2.5, 3.0, and 3.5 minutes of intercritical annealing. Microhardness of the dark and white regions in DP steels are given in Table I. The volume fractions of martensite in DP1, DP2, DP3, and DP4 steels, as estimated by a point-counting technique, are approximately 0.42, 0.51, 0.59, and 0.72, respectively. The macrohardness of DP steels with increasing volume fraction of martensite in DP1, DP2, DP3, and DP4 are 284 HB (2.78 GPa), 320 HB (3.14 GPa), 343 HB (3.4 GPa), and 373 HB (3.66 GPa), respectively. Since the microhardnesses of ferrite and martensite are not significantly different in the DP steels under investigation, increasing macrohardness may be attributed to increasing volume fraction of martensite. B. Friction and Wear Characteristics The variation of cumulative wear volume with sliding distance under different normal loads and at a fixed sliding velocity of 1.15 m/s is shown in Figures 2 through 5 respectively, for DP1, DP2, DP3, and DP4 steels. The cumulative wear volume loss with sliding distance under different normal loads of 14.7, 19.6, 24.5, 29.4, and 3480—VOLUME 33A, NOVEMBER 2002

(b) Fig. 1—Microstructure of DP1 showing dark regions of martensite and white regions of ferrite (etchant 2 pct nital).

Table I. Microhardness of Steel Phases

Steel Designation DP1 DP2 DP3 DP4

Regions

Phases Identified

white dark white dark white dark white dark

ferrite martensite ferrite martensite ferrite martensite ferrite martensite

Average Microhardness (HV)* (Standard Deviation) 190 750 195 740 202 732 210 720

(⫾7) (⫾10) (⫾10) (⫾8) (⫾8) (⫾7) (⫾7) (⫾10)

*Load: 10 g (0.01 kg).

34.3 N has been plotted on a log-log scale, and it has demonstrated a sublinear variation, with coefficients of correlation exceeding 0.98 and 0.99 at all the normal loads for the DP steels, i.e., DP1, DP2, DP3, and DP4. However, the data can also be analyzed on a linear scale using two separate stages of wear behavior characterized by two linear segments. The change in slope has been observed after the first six experimental points (first stage run-in), with the sixth METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 2—Cumulative wear volume with sliding distance at different loads in DP1 containing 42 pct martensite.


point common between both linear segments. Both the lines have been determined by the linear least-squares fit. The wear rate is given by the slope. The procedure followed helps establish the run-in period rate separately from the long-term steady-state rate (second stage). For DP1 steel with 42 vol pct martensite (Figure 2), the first linear segment (run-in) is found to be steeper compared to the second linear segment (steady state). The transition occurs at a distance of 5.53 km. A similar variation has been observed for DP2, DP3, and DP4 steels, as shown in Figures METALLURGICAL AND MATERIALS TRANSACTIONS A



3, 4, and 5, respectively. Other researchers[3,8–12] have observed similar trends in steels. The wear rate, i.e., volume loss in wear per unit sliding distance at a given load, has been determined from the slope of the linear least-squares fit lines at different loads given in Figures 2 through 5. The variation of the wear rate with normal load corresponding to both the run-in and the steady state of wear is shown in Figure 6 for DP1, DP2, DP3, and DP4 steels. The wear rate is observed to increase more or less linearly in the range of load between 14.7 and 34.3 N for all the DP steels in both the run-in and the steady state, VOLUME 33A, NOVEMBER 2002—3481

Fig. 6—Variation of wear rates with normal load in DP steels in both the run-in stage and the steady state of wear.

Fig. 7—Variation of wear rates with martensite volume fraction in DP steels in both the run-in and the steady state of wear.

as shown in Figure 6. Figure 7 shows the variation of the wear rate with the martensite volume fraction for the runin and the steady state of wear at the minimum and maximum loads of 14.7 and 34.3 N used in the present investigation. The wear rate decreases almost linearly with increasing volume fraction of martensite in the range of 42 to 72 vol pct martensite corresponding to DP1 and DP4 steel. A similar trend has been observed at other normal loads. The wear coefficient has been estimated from the slope of the linear variation of wear rate with load, by multiplying it with the initial hardness of the corresponding pin material. Figure 8 shows the variation of the wear coefficient with 3482—VOLUME 33A, NOVEMBER 2002

Fig. 8—Variation of wear coefficient with hardness in both the run-in and the steady state of wear.

Fig. 9—Variation of sliding friction coefficient with distance of sliding at different loads in DP1 containing 42 pct martensite.

hardness corresponding to both the run-in and the steady state of wear for DP1, DP2, DP3, and DP4 steels. It is observed that the wear coefficient for the run-in stage changes relatively less in the hardness range 284 HB to 320 HB, which corresponds to DP1 and DP2 steel containing 42 and 51 vol pct martensite, respectively. However, beyond this value of hardness, there is a sharp decrease in the wear coefficient for the run-in stage with increasing hardness in the range 320 and 373 HB, as observed in DP4 steel containing 72 vol pct martensite. The wear coefficient in the steady state decreases linearly with increasing volume fraction of martensite in DP steels. Figure 9 shows the variation of the coefficient of friction METALLURGICAL AND MATERIALS TRANSACTIONS A

(a)

Fig. 11—Average friction coefficient in the steady state with martensite volume fraction in DP steels.

Fig. 12—Wear surface of the tested specimen of DP2 at a normal load of 14.7 N.

(b) Fig. 10—Average friction coefficient with load in DP steels: (a) run-in stage and (b) steady state of wear.

with sliding distance for different loads against the HRC 62-65 hardened steel disk for DP1 steel. It is noted that the friction coefficient in the run-in stage fluctuates around a mean level, then lessens and stabilizes after a certain period. This trend is similar in all the DP steels investigated in the present study. The fluctuations are relatively large in the run-in short sliding distances corresponding to the first linear segment. Figures 10(a) and (b) show the variation of average coefficient of friction over the distance of sliding corresponding to both the run-in and the steady state of wear with load in DP steels. It is observed that as the load increases, the average coefficient of friction decreases linearly in all the METALLURGICAL AND MATERIALS TRANSACTIONS A

DP steels for both the run-in and the steady state of wear. Figure 11 shows the variation of the average coefficient of friction in steady state with martensite volume fraction at the minimum and the maximum loads of 14.7 and 34.3 N used in the present investigation. The average coefficient of friction is found to decrease linearly with increasing volume fraction of martensite from DP1 (⬃42 pct martensite) to DP4 (⬃72 pct martensite). A similar trend is observed at other loads of 19.6, 24.5, and 29.4 N. Figure 12 shows a typical SEM micrograph of the worn surface of the DP2 steel at a load of 14.7 N. It is observed that there is a highly compacted nonshining transfer layer even at the lowest load of 14.7 N used in the present investigation. The transfer material has covered the scoring marks at many places, as shown in Figure 12. At the loads used in the present investigation, the wear appears to be primarily oxidative for all the DP steels, as is evident from the transfer layer. There are a few locations indicating subsurface cracks VOLUME 33A, NOVEMBER 2002—3483

and pointing toward a possible delamination type of mechanism, but these cracks appear to form around hard martensite islands, as shown in Figure 13. An evaluation of wear particles that correspond to the run-in and the steady state, under microscope, indicates nonshining surfaces, primarily as shown in Figures 14(a) through (d) and Figures 15(a) through (d). X-ray diffraction patterns of the wear debris, as shown in Figures 16(a) and (b), corresponding to both the run-in

Fig. 13—Typical scanning electron micrograph of the subsurface of worn specimen of DP2 at a normal load of 24.5 N.

and the steady state, respectively, reveal peaks of ␣ -Fe2O3 alone without any peak from the metallic material or any other oxide. A few larger agglomerates of oxide are also visible among wear debris, which could be flaked off transfer layer. The Miller indices of planes (hkl), which have given rise to diffraction peaks, are shown in Figures 16(a) and (b). However, no peak was observed for iron. Therefore, metallic wear particles, even if present, could be below the detection limit (⬇4 to 5 pct)[13] of X-ray diffraction. Quinn et al.[14] have also observed the oxidative mechanism of wear in low alloy steels in the load range 4 to 40 N and at sliding velocities ranging from 2 to 5 m/s. They also observed a change in the nature of the oxide from ␣ Fe2O3 to Fe3O4 to FeO, at well-defined loads for every sliding velocity. Sullivan et al.[15] have also reported oxidative wear for the low alloy steels in the load range of 4 to 60 N and sliding velocities ranging from 2 to 5 m/s. They also observed a change in the nature of the oxide during sliding. However, no change in the nature of the oxide is found in the present investigation. Of the three types of oxides of iron, ␣ -Fe2O3 has been reported to be a lowtemperature oxide, which forms around temperatures less than 450 ⬚C. Fe3O4 forms between 450 ⬚C to 600 ⬚C, whereas FeO forms at temperatures greater than 600 ⬚C, as revealed by Sullivan et al. on the basis of static oxidation experiments.[15] The flash temperature, i.e., the instantaneous tem-

(a)

(b)

(c)

(d )

Fig. 14—Micrographs showing wear debris of (a) DP1, (b) DP2, (c) DP3, and (d ) DP4 steels, spread on white paper, for short sliding distances corresponding to first linear segment (run-in stage). 3484—VOLUME 33A, NOVEMBER 2002

METALLURGICAL AND MATERIALS TRANSACTIONS A

(b) (a)

(c)

(d )

Fig. 15—Micrographs showing wear debris of (a) DP1, (b) DP2, (c) DP3, and (d ) DP4 steels, spread on white paper, for long sliding distances corresponding to second linear segment (steady state).

Fig. 16—X-ray diffraction patterns of the wear debris generated at a normal load of 34.3 N, as collected from the disk for DP steels in (a) run-in and (b) steady state of wear. METALLURGICAL AND MATERIALS TRANSACTIONS A

VOLUME 33A, NOVEMBER 2002—3485

perature at the contact points during sliding, which could be attained for a given material combination, depends upon the sliding and environmental conditions. It has also been reported by Quinn[16] that the temperature at the contacts, i.e., flash temperature during sliding, is about 200 ⬚C higher than the general surface temperature. In the present study, the temperature measured at a distance of 3 mm from the sliding surface is 70 ⬚C only, due to relatively lower loads and sliding velocity, and therefore, it is reasonable to assume that the temperature of the contact will not exceed 450 ⬚C and the formation ␣ -Fe2O3 is in good agreement with the results of static oxidation experiments. The temperature estimated from the equations given by Alpas and Ames[17] is also around 415 ⬚C. The theoretical bulk temperature at the surface of the sample calculated on the basis of equations given by Lim and Ashby[18] is found to be 65 ⬚C, which is the same as the temperature measured 3 mm away from the sliding surface. Thus, only ␣ -Fe2O3 has been observed in the debris, as evident from the X-ray diffraction patterns shown in Figures 16(a) and (b). A higher cumulative volume loss, which gives rise to a higher wear rate in the run-in stage as compared to that in the steady state in DP steels, as shown in Figures 2 to 5, may be explained on the basis of the initial surface roughness of the wearing material. When two previously unworn surfaces are first brought into contact and slid relative to one another, mechanical, thermal, chemical, and microstructural changes begin to occur in and adjacent to the contact interface.[19] It is well known that the surfaces of the engineering components are rough and have asperities. As the relative motion of sliding between the two bodies takes place, contact occurs at these asperities and the surfaces evolve to attain better conformity to each other at the end of the run-in stage. The wear in this stage occurs by the removal of high asperities, initial oxide layers, and surface contaminants. Consequently, the material loss and the wear rate are higher in the run-in stage of wear. In the run-in stage, the oxidation of the surface begins and progresses with frictional heating generated during sliding. The transfer layer of oxide may, thus, begin to form in the run-in stage and evolves to the steady state, providing an extent of cover determined by the conditions of load, sliding velocity, and environmental conditions. The second stage of wear indicates a steady state with respect to (1) the evolution of mating surfaces to better conformity, (2) the spread of oxide and compacted transfer layer, and (3) the real area of contact. The wear rate in all the DP steels increases linearly with increasing load in both the run-in and the steady state of wear (Figure 6). However, the wear rate is less in the steady state as compared to that in the run-in stage. In the run-in stage, the variation in weight losses ranged to a maximum of ⫾35 pct at the lowest load of 14.7 N and reduced to ⫾8 pct at the highest load of 34.3 N. Similarly, in the steady state, the variation in weight losses ranged to a maximum of ⫾15 pct at the lowest load of 14.7 N and reduced to ⫾7 pct at the highest load of 34.3 N. As the sliding continues, more and more debris is entrapped between the sliding surfaces and gets compacted due to repetitive sliding and forms a transfer layer of oxide over the surface. This transfer layer protects the underlying metal and the wear rate decreases. A lower wear rate in the steady state as compared to that 3486—VOLUME 33A, NOVEMBER 2002

in the run-in stage for DP steels has also been reported by Sawa and Rigney.[4] A linearly increasing pattern of wear rate with increasing applied load is observed in the present study for all the DP steels. The DP4 steel has consistently shown a lower wear rate compared to that observed in other DP steels at all the loads used in the present study. The lower wear rate in DP4 steel reflects the effect of a higher martensite volume fraction (72 pct) in this steel. However, the wear rate at a given load is observed to decrease linearly with increasing volume fraction of martensite in DP steels, as shown in Figure 7 for both the run-in and the steady state, respectively. The decreasing wear rate with increasing volume fraction of martensite may be explained on the basis of the hardness imparted by incorporation of the martensite in these steels. The steel containing higher volume fraction of martensite will have lower real area of contact as a result of the increased hardness. Since the wear rate is directly proportional to the real area of contact, the steel containing a relatively higher percentage of martensite will show a lower wear rate. The observed trend is in agreement with the observations of Basak et al.,[5] who have also reported that the wear resistance of DP steels increases with the increasing volume fraction of martensite. The increasing difference in the wear rates in DP steels at higher loads as compared to the lower load as seen from Figure 6, may be explained on the basis of better compaction and adhesion of transfer layer aided by frictional heating. It has been shown that a harder substrate is able to hold a thicker transfer layer of oxide more firmly as compared to a softer one.[20] Therefore, the DP4 steel, which is harder than other DP steels investigated, may be able to hold a transfer layer of larger critical thickness firmly before it flakes off and, hence, the difference in wear rates will be higher at higher loads. The other factor contributing to the observed behavior of wear rate may be the extent of cover provided by the transfer layer of oxide and the nature of adhesion of the compacted transfer layer to the pin surface. This layer may flake off easily from a substrate of lower hardness. In other words, the critical thickness of the oxide may be lesser in a material of relatively lower hardness as compared to a material of relatively higher hardness. Hence, there will be a higher probability of the flaking off of this layer in the DP steel containing a relatively smaller amount of martensite. Thus, a higher wear rate in materials of comparatively lower hardness may be attributed to the increase in the flaking off of the transfer layer during sliding. When a test sample is under dry sliding wear at relatively low loads, the frictional heating helps atmospheric oxidation over the sliding surface, the oxide layer gets removed by repeated and multiple contacts, and wear debris of oxide particles is generated. The wear debris gets trapped between the sliding surfaces and is compacted into a layer. The continuing process of removal of the transfer layer and its reformation and thickening results in fluctuations of the friction coefficient, as observed in Figure 9. At very short sliding distances, the fluctuations may also include the contribution resulting from the variation in contact that occurs when the sample and the counterface evolve to develop better surface conformity. The average coefficient of friction in both the run-in and the steady state is found to decrease linearly with increasing load, as shown in Figures 10(a) and (b). This may be attributed to the increased rate of oxidation and better METALLURGICAL AND MATERIALS TRANSACTIONS A

compaction of the oxide caused by the enhanced frictional heating at higher loads. The junctions on the oxidized surface will require relatively less energy to shear during sliding as compared to that for metallic junctions. The average coefficient of friction in the run-in stage is also higher as compared to that in the steady state due to interlocking of high asperities in the run-in stage, which requires more energy to slide one asperity over the another. But in the steady state, the area under the cover provided by the oxide and the transfer layer of oxides is greater and the asperities have attained steady state heights, thereby reducing the friction in the steady state. The possibility of the subsurface work hardening in DP steels has been explored by hardness measurements near the subsurface on tapered section (slope ⫽ 1:10) of the specimens after the complete period of the wear test. In DP1 steel, the hardness increases from an initial value of 353 to 392, 419, and 431 HV after sliding at loads of 14.7, 24.5, and 34.3 N, respectively. The hardness of the DP2 steel increases from 412 to 431, 449, and 467 HV, respectively, after sliding under the same normal loads, whereas for the DP3 steel, the hardness is found to increase from 444 to 465, 483, and 498 HV, respectively. The hardness of the DP4 steel increases from 503 to 518, 532, and 549 HV. The subsurface work hardening, as revealed by the flow lines in the SEM micrograph of the subsurface shown in Figure 13 for the DP1 steel, may be responsible for this increase in hardness. The increased hardness of the substrate due to work hardening is expected to lower the real area of contact, when one estimates the real area of contact from considerations similar to that in indentation hardness. But the ratio of the hardness of the work-hardened surface with respect to the initial hardness of the material after sliding is less than the ratio of applied normal loads. Thus, work hardening may explain the lowering of friction with sliding distance at a given load, as shown in Figure 9, but not the observed lowering of friction with increasing load. The extent of cover provided by the compacted transfer layer on the sliding surface may account for the decrease in friction at higher loads. The increased cover of the transfer layer reduces the friction because of formation of relatively weaker junctions of lower shear strength in the oxide area. Since work hardening is expected to lower the actual real area of contact, the number of junctions should also be reduced, which may lower the friction coefficient. The wear coefficient, which may be interpreted as wear rate per unit real area of contact, does not change significantly in the run-in stage between DP1 and DP2 steels containing 42 and 51 vol pct martensite, respectively, but decreases sharply as one moves from DP2 to DP4 containing 51 and 72 vol pct martensite in the run-in stage, as shown in Figure 8. But in the steady state, the wear coefficient decreases linearly with the increasing volume fraction of the martensite. The decrease in wear coefficient may be attributed to the decreasing wear rate dominating over the decrease in the real area of contact due to increasing hardness. For the wear coefficient, because of its definition, a decrease in real area of contact becomes a disadvantage. Hence, the wear coefficient as a sensitive discriminating wear parameter may not be adequate in materials of similar property and structure, and the wear rate may be a better indicator. METALLURGICAL AND MATERIALS TRANSACTIONS A

IV. CONCLUSIONS 1. The mechanism of wear is primarily oxidative in the range of loads and sliding speed used in the present study for all the DP steels investigated, as evident from the wear debris and the transfer layers. 2. The wear debris of oxides, confirmed by X-ray diffraction patterns, consists of ␣ -Fe2O3 in the wear debris; no metallic particles or other oxides could be detected by Xray diffraction. 3. For a given load, the cumulative wear volume increases linearly with increasing sliding distance both in the runin stage and in the steady state for all the DP steels. 4. The onset of the steady state has been attributed to the attainment of a steady state with respect to the real area of contact, oxidation, and the formation of a well-compacted transfer layer of the oxide for a given condition of load and sliding velocity. 5. The wear rate increases linearly with increasing load in both the run-in and the steady state of wear for all the DP steels investigated. The wear rate is less in the steadystate stage as compared to that in the run-in stage of wear. For a particular load, the wear rate in the DP steels decreases linearly with the increasing volume fraction of martensite. The lowest wear rate is observed in the DP4 steel containing 72 pct martensite. 6. The friction coefficient decreases with increasing normal load in both the run-in and the steady state of wear for all the DP steels, and, at a given load, the average coefficient of friction decreases with the increasing hardness of the materials containing an increasing amount of martensite. 7. The wear coefficients are similar for DP1 and DP2 in the run-in stage of wear despite the difference in the wear rates, but it decreases sharply for DP3 and DP4 containing a higher amount of martensite, indicating a higher decrease in wear rate compared to the decrease in the real area of contact. However, in the steady state, the wear coefficient decreases linearly with the increasing volume fraction of martensite. REFERENCES 1. M.S. Rashid: “A Unique High-Strength Sheet Steel with Superior Formability,” SAE Preprint No. 760206, Sect. 2, 1976, vol. 85, pp. 938-49. 2. X. Sui, X. Zhao, X. Li, W. Yao, Q. Wang, H. Ying, and W. Bi: Proc. Conf. on Low Carbon Silicon-Niobium Dual Phase Steel Wires for Wear Resistant Screens, HSLA Steels: Processing, Properties and Applications, Beijing, China, Oct. 28–Nov. 2, 1990, TMS, Warrendale, PA, 1992, pp. 483-88. 3. S.F. Wayne and S.L. Rice: Wear, 1983, vol. 85, pp. 93-106. 4. M. Sawa and D.A. Rigney: Wear, 1987, vol. 119, pp. 369-90. 5. A. Basak, D.C. Reddy, and D.V.K. Kanth: Mater. Sci. Technol., 1998, vol. 14, pp. 776-82. 6. S.K. Nath: Ph.D. Thesis, University of Roorkee, Roorkee, 1989, p. 91. 7. R.T. Dehoff and F.N. Rhines: Quantitative Microscopy, Materials Science and Engineering Series, McGraw-Hill Book Company, New York, NY, 1968, p. 52. 8. R. Tyagi, S.K. Nath, and S. Ray: Metall. Mater. Trans. A, 2001, vol. 32A, pp. 359-67. 9. P. Clayton: Wear, 1980, vol. 60, pp. 75-93. 10. A.F. Smith: Wear, 1986, vol. 110, pp. 151-68. 11. A.F. Smith: Wear, 1988, vol. 123, pp. 313-31. 12. A. Iwabuchi, K. Hori, and H. Kubosawa: Wear, 1988, vol. 128, pp. 123-37. VOLUME 33A, NOVEMBER 2002—3487

13. B.D. Cullity: Elements of X-Ray Diffraction, Addison-Wesley Publishing Company, Inc., Reading, MA, p. 395. 14. T.F.J. Quinn, D.M. Rowson, and J.M. Sullivan: Wear, 1980, vol. 65, pp. 1-20. 15. J.L. Sullivan, T.F.J. Quinn, and D.M. Rowson: Trib. Int., 1980, vol. 13(4), pp. 153-58. 16. T.F.J. Quinn: Trib. Int., 1983, vol. 16(5), pp. 257-71.

3488—VOLUME 33A, NOVEMBER 2002

17. W. Ames and A.T. Alpas: Metall. Mater. Trans. A, 1995, vol. 26A, pp. 85-98. 18. S.C. Lim and M.F. Ashby: Acta Metall., 1987, vol. 35(1), pp. 1-24. 19. P.J. Blau: Wear, 1981, vol. 72, pp. 55-66. 20. N. Saka, J.J. Pamies-Teixeira, and N.P. Suh: Wear, 1977, vol. 44, pp. 77-86.

METALLURGICAL AND MATERIALS TRANSACTIONS A