- Email: [email protected]
Wear map for grey cast iron
Wear 255 (2003) 401–409
Wear map for grey cast iron A.R. Riahi∗ , A.T. Alpas NSERC/General Motors of Canada Industrial Research Chair on Tribology of Lightweight Materials, Department of Mechanical, Automotive and Materials Engineering, University of Windsor, Windsor, Ont., Canada N9B 3P4
Abstract The sliding wear resistance of a A30 type grey cast iron against AISI 52100 type steel was measured within a load range of 0.3–50.0 N, and a sliding speed range of 0.2–3.0 m/s using a block-on-ring wear machine. The wear rates and the surface temperatures were measured as a function of loading conditions. A wear map was constructed to summarize the measured wear rates and mechanisms that control the wear rates. Ultra-mild, mild, and severe wear regimes were identified on the map. In the ultra-mild wear regime, the wear rates were 8 × 10−7 and 9 × 10−7 mm3 /m at 0.3 N for 0.2 and 0.5 m/s, respectively. The contact surfaces were covered by compacted iron oxide layers. In the mild wear regime, wear rates were in the range of 10−5 mm3 /m at low loading conditions, and 10−4 mm3 /m at high loading conditions. The onset of severe wear coincided with local material transfer to the steel counterface, which increased the roughness of the counterface. The wear rates in the severe wear regime were three orders of magnitude higher (e.g. 10−1 to 3.2 × 10−1 mm3 /m) compared to those of the mild regime. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Cast iron; Wear map; Wear mechanisms; Wear transitions; Microstructure
1. Introduction Grey cast iron is an inexpensive and readily available material. It is universally used for the manufacturing of piston rings and cylinder liners. Unique characteristics of cast iron include the combination of good mechanical properties, good friction and wear characteristics, and economical manufacturing processes. The excellent wear resistance of grey cast iron during dry sliding at low loading conditions is attributed to the feeding of the contact surface by graphite flakes and formation of a graphite film on the contact surface [1–3]. It is generally accepted that a pearlitic structure of the matrix and an ASTM A type graphite flake provides the best wear resistance for grey cast iron in engine cylinder bore applications [4–6]. There has been extensive work on constructing wear maps for steels [7,8], Al–Si alloys [9], and new tribological materials, such as metal matrix composites [10], and coatings [11]. However, there is no published wear map for cast iron, which is one of the oldest and cheapest tribological materials. One of the advantages of the construction of a wear map for cast iron is to provide a bench mark that can be utilized to ∗ Corresponding author. Present address: Kingston Research and Development Centre, Alcan International Limited, P.O. Box 8400, 945 Princess Street, Kingston, Ont., Canada K7L 5L9. Tel.: +1-613-541-2055; fax: +1-613-541-2134. E-mail address: [email protected] (A.R. Riahi).
assess the wear resistances of newer tribological materials, such as lightweight aluminium matrix composites, which are being developed for automotive applications. In this study, the experimental work that was performed in order to construct a wear map for ASTM A30 type grey cast iron with a pearlitic matrix is described. Specific wear mechanisms, which have not been previously reported, are discussed.
2. Materials and experimental methods Dry sliding wear tests were performed on samples made of ASTM A30 grey cast iron [12]. The grey cast iron has a pearlitic matrix with graphite flakes of 45 m in average length, and a distance of approximately 9 m between them (Fig. 1). These are type A graphite flakes that are randomly distributed and oriented throughout the matrix. Table 1 lists the measured mechanical properties and density of the material tested in this work. Wear tests were performed using a block-on-ring configuration within a load range of 0.4–100 N, and a sliding speed range of 0.2–3.0 m/s. The load was applied using dead weights. The samples were taken from the cylinder liners (for aluminium automotive engines) made of cast iron, and machined into rectangular blocks of 10 mm×10 mm×5 mm in dimension. The samples were not given any heat treatment. The narrow surfaces (10 mm × 5 mm) of the test
0043-1648/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0043-1648(03)00100-5
402
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
The mass of each specimen was measured before and after each wear test (with an accuracy of 0.1 mg). The mass difference was divided by the sliding distance to calculate the wear rate. Mass based wear rates were then converted to volumetric wear rates using the density of the material. The contact surface temperature (the temperature approximately 5 mm below the contact surface) of the specimen was measured using a commercial chromel–alumel (K type) thermocouple probe with a stainless steel sheath of 0.5 mm outer diameter. A scanning electron microscope (SEM) that is equipped with an energy dispersion spectrometer (EDS), was used to characterize compositions, morphologies of the worn surfaces, the cross-sections below the worn surfaces, and the loose debris particles that were generated during sliding. Fig. 1. Microstructure of the grey cast iron (ASTM A30) with type A graphite distribution.
3. Experimental results Table 1 The physical and mechanical properties of the grey cast iron ASTM A30 Vickers hardness at 25 g (kg/mm2 ) Density (g/cm3 ) Tensile strength (MPa)
251 ± 10 7.3 ± 0.3 238 ± 6
samples were put in contact with the counterface. The counterface ring was made of SAE (AISI) 52100 type bearing steel and had an outer diameter of 38 mm. The sliding distance for each test was 4000 m for all sliding speeds and loads. However, for the high loading conditions the total sliding distance was limited by the onset of severe wear.
3.1. Wear regimes Fig. 2 shows the wear rate versus load curves measured at various sliding speeds. The mild wear regime extended over a wide range of loads (1.0–60 N) and sliding speeds (0.2–3.0 m/s). More precisely, mild wear was observed to occur under the following conditions: at 0.2 m/s between 1 and 32 N; at 0.5 m/s between 1 and 30 N; at 0.8 m/s between 1 and 12 N; at 1.2 m/s between 1 and 8 N; at 2 m/s between 1 and 5 N; and at 3 m/s between 2 and 4.5 N. In the mild wear regime, wear rates increased with the load, from
Fig. 2. Wear rate vs. load curves at various sliding speeds, for grey cast iron.
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
approximately 1 × 10−5 to 10−4 mm3 /m. The wear rates of the grey cast iron in the mild wear regime increased linearly by increasing the load from 1 to 5 N (Fig. 2) and obeyed the Archard wear equation W =K
P H
(1)
where W is the volumetric wear rate in mm3 /m, P the applied load in kg, H the Vickers hardness of cast iron in kg/mm2 , and K the wear coefficient. By substituting the values of W and P from Fig. 2, the value of K can be calculated as 5.02 × 10−3 for the mild wear regime. At low loads below 1.0 N, wear rates decreased by an order of magnitude compared to the wear rates in the mild wear regime. This revealed that another wear regime existed, namely the ultra-mild wear regime. In this wear regime, the value of K in Eq. (1) was 5.02 × 10−2 . Fig. 2 shows that a sharp increase in the wear rates can be observed at high loads where the transition from mild to severe wear occurred. This transition occurred at 37.5 N for the tests at 0.5 m/s, 16.0 N at 1.2 m/s, 10.0 N at 2.0 m/s, and 6.5 N at 3.0 m/s. The transition to severe wear was accompanied by a large scale material transfer from the surface of cast iron to the steel ring. The boundaries between ultra-mild, mild, and severe wear regimes were identified by the sudden change in the wear rates of at least one order of magnitude. Visual observations of the macroscopic changes in the worn surface morpholo-
403
gies during the tests, and the microscopic examination of the worn samples after the tests indicated that different wear mechanisms operated in each wear regime. Details of the microstructural and compositional changes that took place during wear transitions were studied by SEM and EDS examinations, and will be discussed in Section 4. 3.2. The wear map The measured wear rates, and the main wear mechanisms operating in each wear regime are summarized in a diagram (wear map) with log load versus log sliding speed axes. Fig. 3 shows the wear map of the A30 type grey cast iron. The map depicts the three major wear regimes; namely, ultra-mild, mild and severe wear regimes. The wear map indicates that the transition to ultra-mild wear occurred at loads approximately 0.5 N, and that it was not sensitive to the test speed. Wear rates in the ultra-mild wear regime ranged from 8 × 10−7 to 9 × 10−7 mm3 /m. On transition to the mild wear regime, the wear rates sharply increased by approximately an order of magnitude to 1 × 10−5 mm3 /m, then increased steadily to 10−4 mm3 /m by increasing the load and the sliding speed. In the severe wear regime, the wear rates were up to three orders of magnitude greater than those in the mild wear regime. Severe wear rates were always higher than 10−1 mm3 /m. According to the wear map of the grey cast iron shown in Fig. 3, the loads at which transition occurred between one
Fig. 3. The wear map constructed for the grey cast iron. Three major wear regimes of ultra-mild, mild, and severe wear are depicted on the map. The wear rates shown should be multiplied by 10−3 to find the actual wear rates in mm3 /m.
404
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
regime to another, Ptr , were approximately linear functions of the sliding speed used (in logarithmic scale). Therefore, the relationship between the transition loads and the transition speeds can be expressed as follows: log Ptr = C1 log V + C2
(2)
where Ptr is the transition load in N, V the sliding speed in m/s and C1 and C2 are constants that can be determined from the experimental data. For the transition between the ultra-mild and the mild wear regimes C1 = −0.27, and C2 = −0.39. For the transition between the mild and the severe wear regimes, these constants are C1 = −0.954, and C2 = 1.287. Since the value of C1 is very close to −1 for the transition from mild to severe regime, Eq. (2) becomes log Ptr = C1 log V + log C3
and (3b)
multiplying both sides of Eq. (3b) by the coefficient of friction, µ, which has small variation with speed as shown in Fig. 4, µPtr V = µC3 ≈ const. where C3 = 19.36.
As indicated in Section 3, the wear regimes depicted in the wear map (Fig. 3) were based on the wear rate information, as well as on the metallographic and compositional studies of the worn surfaces. Microscopic aspects of the wear mechanisms observed in each of the wear regimes are discussed in the following sections. 4.1. Ultra-mild wear regime
C3 V
Ptr V = C3
4. Discussion
(3a)
where log C3 = C2 then, for C1 = −1, Ptr =
In Eq. (4), the term (µPtr V) is the energy produced during sliding contact per unit time (N m/s). Therefore, the transition from mild to severe wear occurs when the term (µPtr V) reaches a certain constant value, which is equal to 9 W, as shown in Fig. 4.
(4)
In the ultra-mild wear regime, the worn surfaces of cast iron were covered by continuous layers of compacted reddish iron oxide powder, and exhibited no significant evidence for plastic deformation. Fig. 5a shows the worn surface of cast iron tested at 0.2 N and 0.2 m/s, covered by a compacted fine powder of iron oxide. Similar iron oxide layers were formed on the surface of the 52100 steel counterface ring in contact with the cast iron samples. The steel counterface exhibited continuous strips of reddish iron oxide, indicating that the cast iron was separated from the steel ring by an oxide layer. In this way, metal-to-metal contact between the cast iron samples and
Fig. 4. Variation of coefficient of friction (µ), and the sliding energy rate (µPtr V) as a function of loads and speeds combinations at which the mild to severe wear transitions occurred. The continuously decreasing curve shows the change in the transition loads as a function of sliding speeds.
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
405
Fig. 5. The surface morphologies of grey cast iron in (a) the ultra-mild regime; (b) the mild regime at low loads and high sliding speeds; (c) the mild regime at low loads and sliding speeds; and (d) the severe wear regime.
the steel ring was largely prevented except during the initial stages of wear. 4.2. Mild wear regime In the mild regime, wear rates in grey cast iron were primarily controlled by surface oxidation. Detailed discussions of the oxidational wear mechanisms in ferrous materials can be found in the seminal papers of Welsh [13] and Quinn
et al. [14]. In this section, only the specific aspects of the wear mechanisms in the cast iron that were not reported in the literature will be discussed. As shown in Fig. 5c, when the applied loads and sliding speeds were modest, the oxide films that formed on the wear tracks were thin (a fraction of micrometer) and discontinuous. Under these conditions (2 N, 0.3 m/s), approximately 40% of the contact area was covered with a dark oxide film. The regions between the oxide films consisted of compacted
406
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
Fig. 6. The cross-section of the grey cast iron sample tested in the mild wear regime showing the formation of large debris as a result of the failure at the matrix material between the graphite flake and the contact surface.
fine reddish iron oxide powder. At higher loads and sliding speeds, a larger proportion of the contact surface area became covered by the oxide film. For example, at a sliding speed of 2.0 m/s and a load of 5.0 N, 80% of the wear track was covered by the dark oxide film (Fig. 5b). Due to the presence of graphite flakes in the microstructure of the grey cast iron, large-size debris was formed in the mild wear regime (and sometimes even in the ultra-mild wear regime). Fig. 6 is a cross-section of a sample tested in the mild wear regime, which shows an example of the fracture morphology formed as a result of fracture of the matrix by a mechanisms that could be described as “failed necks” between the graphite flakes. The removal of large pieces of material in this way produced large-size loose debris particles, up to 50 m in length and up to 15 m in thickness. As Fig. 1 shows, the microstructure of grey cast iron incorporated graphite flakes that were often grouped in a peculiar way, and formed a microstructural feature that could be named as the “rosette morphology”. During the sliding wear tests, the “rosette morphology” constituted the weakest link in the microstructure. As shown in Fig. 7, the rosette-grouped graphite flakes in the layers adjacent to the contact surfaces collapsed readily. The result was the formation of metallic particles, up to 50 m in length, which did not oxidize quickly due to their large size. They either left the tribo-system as loose debris or were entrapped in the interface and acted as third body elements. The particles entrapped between the contact surfaces often caused scuffing of the both contact surfaces. Therefore, two similar fracture mechanisms that operate near the contact surfaces are responsible for the formation of large chunks of loose debris particles. As indicated above, the formation of the rosette type graphite flake morphology is the primary reason for the formation of the large metallic debris. However, this type of debris can also be formed in
Fig. 7. The cross-section of the grey cast iron sample tested in the mild wear regime that shows the graphite flakes forming rosette groups close to the contact surface. This morphology creates weak points where the material near the contact surface fails easily.
locations where the graphite flakes do not form such groups. Fig. 8 schematically shows how individual graphite flakes cause formation of “tiny necks” in the material below the contact surface. A thin region of the matrix material, which is labelled as iron splat in Fig. 8a, located between a graphite flake near the surface and the asperity of the counterface, is connected to the rest of the matrix only through a very narrow region (tiny neck). The necks fail more easily than the rest of the matrix as shown in Fig. 6, and the proposed mechanism of neck failure is depicted in Fig. 8b and c. At high loading conditions, the temperature and pressure on the particles appeared to be sufficiently high to cause welding of the lose particles to the counterface as shown in Fig. 9. In the severe wear regime, oxide films were removed from the contact surfaces. There was severe mechanical damage to the contact surface (Fig. 5d) leaving no opportunity to grow stable oxide films. 4.3. Transition from mild to severe wear regime The transition to severe wear was accompanied by a significant increase in the roughness of both worn surfaces of the cast iron samples, and the steel ring. Fig. 10 shows the severity of plastic deformation at the worn surface of the grey cast iron tested at 15.0 N and 2.0 m/s. The fragmentation of severely deformed surface layers leads to the formation of loose debris. Fig. 11 shows the variation of the bulk surface temperature with the sliding speed at the onset of severe wear. At low sliding speeds, the rate of heat generation due to friction is low, and consequently the surface temperature is low (60 ◦ C at 0.2 m/s). The temperature increases with increasing the sliding speed up to 0.8 m/s. For speeds higher than 0.8 m/s, the temperature decreases as the sliding speed increases. The transition to severe wear coincided with the local welding of the large-size cast iron debris, which was transferred to the counterface. This is assumed to occur when the local temperature (i.e. the flash
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
407
Fig. 8. Schematic representation of the sequence of events leading to the failure of the necks, or narrow regions of the matrix between the graphite flakes and the contact surface, as a result of surface deformation.
temperature rather than the measured bulk surface temperature) was sufficiently high. It was also observed that the microhardness of the transferred material was higher than 900 kg/mm2 (Vickers hardness at 25 g), which represented
Fig. 9. The morphology of the steel counterface formed as a result of the material transfer from the surface of the grey cast iron.
Fig. 10. Damage as a result of plastic deformation at the worn surface of grey cast iron in the severe regime. A large fragment of cast iron at the edge of the worn surface (marked by the arrow) is about to separate.
408
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
Fig. 11. The variation of bulk surface temperature with the sliding speed at the onset of severe wear.
Fig. 12. The change in the bulk surface temperature with the load (at a speed of 1.2 m/s).
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
an increase of almost five times compared to that of the cast iron before the wear test (Table 1). This hardness increase can be regarded to be an indication that martensitic hardening might have taken place during the rapid heating and cooling cycles accompanying the material transfer process. In addition, as a result of the material transfer process, the roughness of the counterface increased from 0.12 to about 4.0 m. In summary, according to these observations it can be suggested that transformation hardened asperities were formed on the counterface that were easily welded to the existing asperities at the tips of which high flash temperatures were maintained. The transferred material rapidly accumulated on the counterface to a maximum height of 10 m. It was further observed that the bulk temperature decreased once the transition to severe wear was complete. This can be seen in Fig. 12, which shows the variation of the bulk temperature with the load at a constant sliding speed of 1.2 m/s. It can be seen that as soon as severe wear occurred at 1.2 m/s, the bulk temperature dropped from 70 to 55 ◦ C. The hardened asperities are believed to have increased the severity of deformation that occurred on the cast iron surfaces as shown in Fig. 10.
5. Summary and conclusions Sliding wear of the A30 type grey cast iron worn against the 52100 type bearing steel, was investigated under a broad range of loading and sliding speed conditions. The wear rates were measured and the micromechanisms responsible for them were identified. The results were summarized in the form of a wear mechanism map. The main results of this work are as follows: (1) According to the wear map constructed, transition boundaries between the mild to severe wear, and the ultra-mild to mild wear regimes were linear on the log load versus log velocity scale. (2) An empirical equation was developed to show that the transition from mild to severe wear in grey cast iron occurred when the energy generated at the contact surfaces (µPtr V) reached a constant value. (3) Fracture of graphite flakes grouped to form a rosette type morphology, as well as fracture of the matrix at the “necks” formed between the graphite flakes and the
409
contact surface. This led to the formation of large-size debris in the mild wear regime, which was otherwise dominated by oxidative wear. (4) The transferred cast iron debris particles that were hardened during this process played a role in the transition to the severe wear by facilitating surface deformation.
Acknowledgements This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and General Motors of Canada Ltd. The material studied was provided by General Motors Research and Development Center, Warren, MI, USA. References [1] E. Takeuchi, The mechanisms of wear of cast iron in dry sliding, Wear 11 (1968) 201–211. [2] R.S. Montgomery, Run-in and glaze formation on grey cast iron surfaces, Wear 14 (1969) 99–105. [3] J. Sugishita, S. Fujiyoshi, The effect of cast iron graphites on friction and wear performance. 1. Graphite film formation on grey cast iron surfaces, Wear 66 (1981) 209–221. [4] C.V. White, Friction, Lubrication and Wear Technology, ASM Handbook, vol. 18, 10th ed., ASM International, Metals Park, OH, 1992, pp. 553–562. [5] Properties and selection: irons and steels, Metals Handbook, vol. 1, ninth ed., ASM International, Metals Park, OH, 1987, pp. 25–26. [6] P.W. Leech, D.W. Borland, The unlubricated wear of flake graphite cast iron, Wear 85 (1983) 257–266. [7] S. Lim, M.F. Ashby, Wear-mechanism maps, Acta Metall. 35 (1987) 1–24. [8] H. Kato, T.S. Eyre, B. Ralph, Wear mechanism map of nitrided steel, Acta Metall. Mater. 42 (5) (1994) 1703–1713. [9] R. Shivanath, P.K. Sengupta, T.S. Eyre, Wear of Al–Si alloys, Br. Foundryman 70 (1977) 349–356. [10] S. Wilson, A.T. Alpas, Wear mechanism maps for metal matrix composites, Wear 212 (1997) 41–49. [11] T.S. Eyre, R.F. Iles, D.W. Gasson, Wear characteristics of flake and nodular graphite cast iron, Wear 13 (1969) 229–245. [12] ASTM Speciality Handbook “Cast Irons”, ASM International, Materials Park, OH, 1996, p. 40. [13] N.C. Welsh, Frictional heating and its influence on the wear of steel, J. Appl. Phys. 28 (9) (1957) 960–968. [14] T.F.J. Quinn, J.L. Sullivan, D.M. Rowson, New Developments in the Oxidational Theory of the Mild Wear of Metals, American Society of Mechanical Engineering, New York, 1979, pp. 1–11.
Wear map for grey cast iron A.R. Riahi∗ , A.T. Alpas NSERC/General Motors of Canada Industrial Research Chair on Tribology of Lightweight Materials, Department of Mechanical, Automotive and Materials Engineering, University of Windsor, Windsor, Ont., Canada N9B 3P4
Abstract The sliding wear resistance of a A30 type grey cast iron against AISI 52100 type steel was measured within a load range of 0.3–50.0 N, and a sliding speed range of 0.2–3.0 m/s using a block-on-ring wear machine. The wear rates and the surface temperatures were measured as a function of loading conditions. A wear map was constructed to summarize the measured wear rates and mechanisms that control the wear rates. Ultra-mild, mild, and severe wear regimes were identified on the map. In the ultra-mild wear regime, the wear rates were 8 × 10−7 and 9 × 10−7 mm3 /m at 0.3 N for 0.2 and 0.5 m/s, respectively. The contact surfaces were covered by compacted iron oxide layers. In the mild wear regime, wear rates were in the range of 10−5 mm3 /m at low loading conditions, and 10−4 mm3 /m at high loading conditions. The onset of severe wear coincided with local material transfer to the steel counterface, which increased the roughness of the counterface. The wear rates in the severe wear regime were three orders of magnitude higher (e.g. 10−1 to 3.2 × 10−1 mm3 /m) compared to those of the mild regime. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Cast iron; Wear map; Wear mechanisms; Wear transitions; Microstructure
1. Introduction Grey cast iron is an inexpensive and readily available material. It is universally used for the manufacturing of piston rings and cylinder liners. Unique characteristics of cast iron include the combination of good mechanical properties, good friction and wear characteristics, and economical manufacturing processes. The excellent wear resistance of grey cast iron during dry sliding at low loading conditions is attributed to the feeding of the contact surface by graphite flakes and formation of a graphite film on the contact surface [1–3]. It is generally accepted that a pearlitic structure of the matrix and an ASTM A type graphite flake provides the best wear resistance for grey cast iron in engine cylinder bore applications [4–6]. There has been extensive work on constructing wear maps for steels [7,8], Al–Si alloys [9], and new tribological materials, such as metal matrix composites [10], and coatings [11]. However, there is no published wear map for cast iron, which is one of the oldest and cheapest tribological materials. One of the advantages of the construction of a wear map for cast iron is to provide a bench mark that can be utilized to ∗ Corresponding author. Present address: Kingston Research and Development Centre, Alcan International Limited, P.O. Box 8400, 945 Princess Street, Kingston, Ont., Canada K7L 5L9. Tel.: +1-613-541-2055; fax: +1-613-541-2134. E-mail address: [email protected] (A.R. Riahi).
assess the wear resistances of newer tribological materials, such as lightweight aluminium matrix composites, which are being developed for automotive applications. In this study, the experimental work that was performed in order to construct a wear map for ASTM A30 type grey cast iron with a pearlitic matrix is described. Specific wear mechanisms, which have not been previously reported, are discussed.
2. Materials and experimental methods Dry sliding wear tests were performed on samples made of ASTM A30 grey cast iron [12]. The grey cast iron has a pearlitic matrix with graphite flakes of 45 m in average length, and a distance of approximately 9 m between them (Fig. 1). These are type A graphite flakes that are randomly distributed and oriented throughout the matrix. Table 1 lists the measured mechanical properties and density of the material tested in this work. Wear tests were performed using a block-on-ring configuration within a load range of 0.4–100 N, and a sliding speed range of 0.2–3.0 m/s. The load was applied using dead weights. The samples were taken from the cylinder liners (for aluminium automotive engines) made of cast iron, and machined into rectangular blocks of 10 mm×10 mm×5 mm in dimension. The samples were not given any heat treatment. The narrow surfaces (10 mm × 5 mm) of the test
0043-1648/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0043-1648(03)00100-5
402
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
The mass of each specimen was measured before and after each wear test (with an accuracy of 0.1 mg). The mass difference was divided by the sliding distance to calculate the wear rate. Mass based wear rates were then converted to volumetric wear rates using the density of the material. The contact surface temperature (the temperature approximately 5 mm below the contact surface) of the specimen was measured using a commercial chromel–alumel (K type) thermocouple probe with a stainless steel sheath of 0.5 mm outer diameter. A scanning electron microscope (SEM) that is equipped with an energy dispersion spectrometer (EDS), was used to characterize compositions, morphologies of the worn surfaces, the cross-sections below the worn surfaces, and the loose debris particles that were generated during sliding. Fig. 1. Microstructure of the grey cast iron (ASTM A30) with type A graphite distribution.
3. Experimental results Table 1 The physical and mechanical properties of the grey cast iron ASTM A30 Vickers hardness at 25 g (kg/mm2 ) Density (g/cm3 ) Tensile strength (MPa)
251 ± 10 7.3 ± 0.3 238 ± 6
samples were put in contact with the counterface. The counterface ring was made of SAE (AISI) 52100 type bearing steel and had an outer diameter of 38 mm. The sliding distance for each test was 4000 m for all sliding speeds and loads. However, for the high loading conditions the total sliding distance was limited by the onset of severe wear.
3.1. Wear regimes Fig. 2 shows the wear rate versus load curves measured at various sliding speeds. The mild wear regime extended over a wide range of loads (1.0–60 N) and sliding speeds (0.2–3.0 m/s). More precisely, mild wear was observed to occur under the following conditions: at 0.2 m/s between 1 and 32 N; at 0.5 m/s between 1 and 30 N; at 0.8 m/s between 1 and 12 N; at 1.2 m/s between 1 and 8 N; at 2 m/s between 1 and 5 N; and at 3 m/s between 2 and 4.5 N. In the mild wear regime, wear rates increased with the load, from
Fig. 2. Wear rate vs. load curves at various sliding speeds, for grey cast iron.
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
approximately 1 × 10−5 to 10−4 mm3 /m. The wear rates of the grey cast iron in the mild wear regime increased linearly by increasing the load from 1 to 5 N (Fig. 2) and obeyed the Archard wear equation W =K
P H
(1)
where W is the volumetric wear rate in mm3 /m, P the applied load in kg, H the Vickers hardness of cast iron in kg/mm2 , and K the wear coefficient. By substituting the values of W and P from Fig. 2, the value of K can be calculated as 5.02 × 10−3 for the mild wear regime. At low loads below 1.0 N, wear rates decreased by an order of magnitude compared to the wear rates in the mild wear regime. This revealed that another wear regime existed, namely the ultra-mild wear regime. In this wear regime, the value of K in Eq. (1) was 5.02 × 10−2 . Fig. 2 shows that a sharp increase in the wear rates can be observed at high loads where the transition from mild to severe wear occurred. This transition occurred at 37.5 N for the tests at 0.5 m/s, 16.0 N at 1.2 m/s, 10.0 N at 2.0 m/s, and 6.5 N at 3.0 m/s. The transition to severe wear was accompanied by a large scale material transfer from the surface of cast iron to the steel ring. The boundaries between ultra-mild, mild, and severe wear regimes were identified by the sudden change in the wear rates of at least one order of magnitude. Visual observations of the macroscopic changes in the worn surface morpholo-
403
gies during the tests, and the microscopic examination of the worn samples after the tests indicated that different wear mechanisms operated in each wear regime. Details of the microstructural and compositional changes that took place during wear transitions were studied by SEM and EDS examinations, and will be discussed in Section 4. 3.2. The wear map The measured wear rates, and the main wear mechanisms operating in each wear regime are summarized in a diagram (wear map) with log load versus log sliding speed axes. Fig. 3 shows the wear map of the A30 type grey cast iron. The map depicts the three major wear regimes; namely, ultra-mild, mild and severe wear regimes. The wear map indicates that the transition to ultra-mild wear occurred at loads approximately 0.5 N, and that it was not sensitive to the test speed. Wear rates in the ultra-mild wear regime ranged from 8 × 10−7 to 9 × 10−7 mm3 /m. On transition to the mild wear regime, the wear rates sharply increased by approximately an order of magnitude to 1 × 10−5 mm3 /m, then increased steadily to 10−4 mm3 /m by increasing the load and the sliding speed. In the severe wear regime, the wear rates were up to three orders of magnitude greater than those in the mild wear regime. Severe wear rates were always higher than 10−1 mm3 /m. According to the wear map of the grey cast iron shown in Fig. 3, the loads at which transition occurred between one
Fig. 3. The wear map constructed for the grey cast iron. Three major wear regimes of ultra-mild, mild, and severe wear are depicted on the map. The wear rates shown should be multiplied by 10−3 to find the actual wear rates in mm3 /m.
404
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
regime to another, Ptr , were approximately linear functions of the sliding speed used (in logarithmic scale). Therefore, the relationship between the transition loads and the transition speeds can be expressed as follows: log Ptr = C1 log V + C2
(2)
where Ptr is the transition load in N, V the sliding speed in m/s and C1 and C2 are constants that can be determined from the experimental data. For the transition between the ultra-mild and the mild wear regimes C1 = −0.27, and C2 = −0.39. For the transition between the mild and the severe wear regimes, these constants are C1 = −0.954, and C2 = 1.287. Since the value of C1 is very close to −1 for the transition from mild to severe regime, Eq. (2) becomes log Ptr = C1 log V + log C3
and (3b)
multiplying both sides of Eq. (3b) by the coefficient of friction, µ, which has small variation with speed as shown in Fig. 4, µPtr V = µC3 ≈ const. where C3 = 19.36.
As indicated in Section 3, the wear regimes depicted in the wear map (Fig. 3) were based on the wear rate information, as well as on the metallographic and compositional studies of the worn surfaces. Microscopic aspects of the wear mechanisms observed in each of the wear regimes are discussed in the following sections. 4.1. Ultra-mild wear regime
C3 V
Ptr V = C3
4. Discussion
(3a)
where log C3 = C2 then, for C1 = −1, Ptr =
In Eq. (4), the term (µPtr V) is the energy produced during sliding contact per unit time (N m/s). Therefore, the transition from mild to severe wear occurs when the term (µPtr V) reaches a certain constant value, which is equal to 9 W, as shown in Fig. 4.
(4)
In the ultra-mild wear regime, the worn surfaces of cast iron were covered by continuous layers of compacted reddish iron oxide powder, and exhibited no significant evidence for plastic deformation. Fig. 5a shows the worn surface of cast iron tested at 0.2 N and 0.2 m/s, covered by a compacted fine powder of iron oxide. Similar iron oxide layers were formed on the surface of the 52100 steel counterface ring in contact with the cast iron samples. The steel counterface exhibited continuous strips of reddish iron oxide, indicating that the cast iron was separated from the steel ring by an oxide layer. In this way, metal-to-metal contact between the cast iron samples and
Fig. 4. Variation of coefficient of friction (µ), and the sliding energy rate (µPtr V) as a function of loads and speeds combinations at which the mild to severe wear transitions occurred. The continuously decreasing curve shows the change in the transition loads as a function of sliding speeds.
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
405
Fig. 5. The surface morphologies of grey cast iron in (a) the ultra-mild regime; (b) the mild regime at low loads and high sliding speeds; (c) the mild regime at low loads and sliding speeds; and (d) the severe wear regime.
the steel ring was largely prevented except during the initial stages of wear. 4.2. Mild wear regime In the mild regime, wear rates in grey cast iron were primarily controlled by surface oxidation. Detailed discussions of the oxidational wear mechanisms in ferrous materials can be found in the seminal papers of Welsh [13] and Quinn
et al. [14]. In this section, only the specific aspects of the wear mechanisms in the cast iron that were not reported in the literature will be discussed. As shown in Fig. 5c, when the applied loads and sliding speeds were modest, the oxide films that formed on the wear tracks were thin (a fraction of micrometer) and discontinuous. Under these conditions (2 N, 0.3 m/s), approximately 40% of the contact area was covered with a dark oxide film. The regions between the oxide films consisted of compacted
406
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
Fig. 6. The cross-section of the grey cast iron sample tested in the mild wear regime showing the formation of large debris as a result of the failure at the matrix material between the graphite flake and the contact surface.
fine reddish iron oxide powder. At higher loads and sliding speeds, a larger proportion of the contact surface area became covered by the oxide film. For example, at a sliding speed of 2.0 m/s and a load of 5.0 N, 80% of the wear track was covered by the dark oxide film (Fig. 5b). Due to the presence of graphite flakes in the microstructure of the grey cast iron, large-size debris was formed in the mild wear regime (and sometimes even in the ultra-mild wear regime). Fig. 6 is a cross-section of a sample tested in the mild wear regime, which shows an example of the fracture morphology formed as a result of fracture of the matrix by a mechanisms that could be described as “failed necks” between the graphite flakes. The removal of large pieces of material in this way produced large-size loose debris particles, up to 50 m in length and up to 15 m in thickness. As Fig. 1 shows, the microstructure of grey cast iron incorporated graphite flakes that were often grouped in a peculiar way, and formed a microstructural feature that could be named as the “rosette morphology”. During the sliding wear tests, the “rosette morphology” constituted the weakest link in the microstructure. As shown in Fig. 7, the rosette-grouped graphite flakes in the layers adjacent to the contact surfaces collapsed readily. The result was the formation of metallic particles, up to 50 m in length, which did not oxidize quickly due to their large size. They either left the tribo-system as loose debris or were entrapped in the interface and acted as third body elements. The particles entrapped between the contact surfaces often caused scuffing of the both contact surfaces. Therefore, two similar fracture mechanisms that operate near the contact surfaces are responsible for the formation of large chunks of loose debris particles. As indicated above, the formation of the rosette type graphite flake morphology is the primary reason for the formation of the large metallic debris. However, this type of debris can also be formed in
Fig. 7. The cross-section of the grey cast iron sample tested in the mild wear regime that shows the graphite flakes forming rosette groups close to the contact surface. This morphology creates weak points where the material near the contact surface fails easily.
locations where the graphite flakes do not form such groups. Fig. 8 schematically shows how individual graphite flakes cause formation of “tiny necks” in the material below the contact surface. A thin region of the matrix material, which is labelled as iron splat in Fig. 8a, located between a graphite flake near the surface and the asperity of the counterface, is connected to the rest of the matrix only through a very narrow region (tiny neck). The necks fail more easily than the rest of the matrix as shown in Fig. 6, and the proposed mechanism of neck failure is depicted in Fig. 8b and c. At high loading conditions, the temperature and pressure on the particles appeared to be sufficiently high to cause welding of the lose particles to the counterface as shown in Fig. 9. In the severe wear regime, oxide films were removed from the contact surfaces. There was severe mechanical damage to the contact surface (Fig. 5d) leaving no opportunity to grow stable oxide films. 4.3. Transition from mild to severe wear regime The transition to severe wear was accompanied by a significant increase in the roughness of both worn surfaces of the cast iron samples, and the steel ring. Fig. 10 shows the severity of plastic deformation at the worn surface of the grey cast iron tested at 15.0 N and 2.0 m/s. The fragmentation of severely deformed surface layers leads to the formation of loose debris. Fig. 11 shows the variation of the bulk surface temperature with the sliding speed at the onset of severe wear. At low sliding speeds, the rate of heat generation due to friction is low, and consequently the surface temperature is low (60 ◦ C at 0.2 m/s). The temperature increases with increasing the sliding speed up to 0.8 m/s. For speeds higher than 0.8 m/s, the temperature decreases as the sliding speed increases. The transition to severe wear coincided with the local welding of the large-size cast iron debris, which was transferred to the counterface. This is assumed to occur when the local temperature (i.e. the flash
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
407
Fig. 8. Schematic representation of the sequence of events leading to the failure of the necks, or narrow regions of the matrix between the graphite flakes and the contact surface, as a result of surface deformation.
temperature rather than the measured bulk surface temperature) was sufficiently high. It was also observed that the microhardness of the transferred material was higher than 900 kg/mm2 (Vickers hardness at 25 g), which represented
Fig. 9. The morphology of the steel counterface formed as a result of the material transfer from the surface of the grey cast iron.
Fig. 10. Damage as a result of plastic deformation at the worn surface of grey cast iron in the severe regime. A large fragment of cast iron at the edge of the worn surface (marked by the arrow) is about to separate.
408
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
Fig. 11. The variation of bulk surface temperature with the sliding speed at the onset of severe wear.
Fig. 12. The change in the bulk surface temperature with the load (at a speed of 1.2 m/s).
A.R. Riahi, A.T. Alpas / Wear 255 (2003) 401–409
an increase of almost five times compared to that of the cast iron before the wear test (Table 1). This hardness increase can be regarded to be an indication that martensitic hardening might have taken place during the rapid heating and cooling cycles accompanying the material transfer process. In addition, as a result of the material transfer process, the roughness of the counterface increased from 0.12 to about 4.0 m. In summary, according to these observations it can be suggested that transformation hardened asperities were formed on the counterface that were easily welded to the existing asperities at the tips of which high flash temperatures were maintained. The transferred material rapidly accumulated on the counterface to a maximum height of 10 m. It was further observed that the bulk temperature decreased once the transition to severe wear was complete. This can be seen in Fig. 12, which shows the variation of the bulk temperature with the load at a constant sliding speed of 1.2 m/s. It can be seen that as soon as severe wear occurred at 1.2 m/s, the bulk temperature dropped from 70 to 55 ◦ C. The hardened asperities are believed to have increased the severity of deformation that occurred on the cast iron surfaces as shown in Fig. 10.
5. Summary and conclusions Sliding wear of the A30 type grey cast iron worn against the 52100 type bearing steel, was investigated under a broad range of loading and sliding speed conditions. The wear rates were measured and the micromechanisms responsible for them were identified. The results were summarized in the form of a wear mechanism map. The main results of this work are as follows: (1) According to the wear map constructed, transition boundaries between the mild to severe wear, and the ultra-mild to mild wear regimes were linear on the log load versus log velocity scale. (2) An empirical equation was developed to show that the transition from mild to severe wear in grey cast iron occurred when the energy generated at the contact surfaces (µPtr V) reached a constant value. (3) Fracture of graphite flakes grouped to form a rosette type morphology, as well as fracture of the matrix at the “necks” formed between the graphite flakes and the
409
contact surface. This led to the formation of large-size debris in the mild wear regime, which was otherwise dominated by oxidative wear. (4) The transferred cast iron debris particles that were hardened during this process played a role in the transition to the severe wear by facilitating surface deformation.
Acknowledgements This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and General Motors of Canada Ltd. The material studied was provided by General Motors Research and Development Center, Warren, MI, USA. References [1] E. Takeuchi, The mechanisms of wear of cast iron in dry sliding, Wear 11 (1968) 201–211. [2] R.S. Montgomery, Run-in and glaze formation on grey cast iron surfaces, Wear 14 (1969) 99–105. [3] J. Sugishita, S. Fujiyoshi, The effect of cast iron graphites on friction and wear performance. 1. Graphite film formation on grey cast iron surfaces, Wear 66 (1981) 209–221. [4] C.V. White, Friction, Lubrication and Wear Technology, ASM Handbook, vol. 18, 10th ed., ASM International, Metals Park, OH, 1992, pp. 553–562. [5] Properties and selection: irons and steels, Metals Handbook, vol. 1, ninth ed., ASM International, Metals Park, OH, 1987, pp. 25–26. [6] P.W. Leech, D.W. Borland, The unlubricated wear of flake graphite cast iron, Wear 85 (1983) 257–266. [7] S. Lim, M.F. Ashby, Wear-mechanism maps, Acta Metall. 35 (1987) 1–24. [8] H. Kato, T.S. Eyre, B. Ralph, Wear mechanism map of nitrided steel, Acta Metall. Mater. 42 (5) (1994) 1703–1713. [9] R. Shivanath, P.K. Sengupta, T.S. Eyre, Wear of Al–Si alloys, Br. Foundryman 70 (1977) 349–356. [10] S. Wilson, A.T. Alpas, Wear mechanism maps for metal matrix composites, Wear 212 (1997) 41–49. [11] T.S. Eyre, R.F. Iles, D.W. Gasson, Wear characteristics of flake and nodular graphite cast iron, Wear 13 (1969) 229–245. [12] ASTM Speciality Handbook “Cast Irons”, ASM International, Materials Park, OH, 1996, p. 40. [13] N.C. Welsh, Frictional heating and its influence on the wear of steel, J. Appl. Phys. 28 (9) (1957) 960–968. [14] T.F.J. Quinn, J.L. Sullivan, D.M. Rowson, New Developments in the Oxidational Theory of the Mild Wear of Metals, American Society of Mechanical Engineering, New York, 1979, pp. 1–11.