Wear 271 (2011) 1602–1608

Contents lists available at ScienceDirect

Wear
journal homepage: www.elsevier.com/locate/wear

Dry rolling-sliding wear of austempered cast iron

G. Straffelini a,∗ , C. Giuliari a , M. Pellizzari a , E. Veneri b , M. Bronzato b
a
University of Trento, via Mesiano 77, 38100 Trento, Italy
b
Zanardi Fonderie S.p.A. via Nazionale 3, 37046 Minerbe, Vicenza, Italy

a r t i c l e i n f o a b s t r a c t

Article history: In the present investigation, the dry rolling-sliding wear of two austempered ductile irons (ADIs), charac-
Received 9 September 2010 terized by different hardness values, was investigated. The tests were carried out using disks with 40 mm
Received in revised form diameter and 10 mm height. The applied load was in the range between 50 and 500 N. Since the initial
22 December 2010
Hertzian pressures were quite large because of the line contact, a plasticity dominated wear with the for-
Accepted 23 December 2010
mation of delamination was obtained. Therefore, the treatment cycle of some specimens was conducted
with the aim of obtaining a mild wear during rolling-sliding through the formation of a surface oxide
layer. As a comparison, the dry rolling-sliding wear of a gas nitrided steel was also investigated.
Keywords:
Rolling-sliding
The results show that the wear coefficients of the specimens with the oxide layer on their surface
Austempered ductile iron was actually mild. However, the specimens produced without such a layer displayed even lower wear
Delamination coefficients. Although wear was by delamination, wear coefficients were still low because of the action
Tribo-oxidation of graphite, which was able to squeeze on the surface during sliding, thus reducing adhesional forces.
In addition, a mechanically mixed layer formed on the sliding surfaces, and this provided an additional
increase in the wear resistance. Because of this, the material with lower matrix hardness displayed a
greater wear resistance. The wear resistance of the nitrided steel was found to be determined by the
outer part of the compound layer, and it was found to be lower than that displayed by the two ADI.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction ing rolling-sliding owing to the release of the graphite nodules and
the formation of a squeeze film. This shows that such materials are
With austempering treatments on ductile iron, it is possible very promising for rolling-sliding applications.
to obtain an interesting matrix microstructure consisting of fer- In the present investigation, the dry rolling-sliding wear of two
rite and retained austenite. Such a microstructure provides a good ADIs, characterized by different matrix hardness, but roughly with
balance in the mechanical properties, including tensile strength, the same graphite content and same dimension of the graphite nod-
ductility and toughness [1]. ules, was investigated. Because of the line contact, large contact
In the literature, several investigations on the dry sliding pressures are attained in rolling-sliding conditions. Wear is then
behaviour of austempered ductile irons (ADIs) are reported. In usually by delamination and may be severe in nature [7]. In order
particular, the role of matrix hardness and graphite nodules was to obtain a mild form of wear, some specimens were thus produced
investigated [2,3]. In addition, the role of the tribological param- with a different cycle, to induce the formation of an oxide layer on
eters, such as load, sliding velocity and test duration, in mild their surfaces. For a comparison, a widely used nitrided steel was
oxidative wear and in the transition from mild to severe wear by also investigated. In fact, nitrided steels are often used in engineer-
delamination was also studied by several authors [4–6]. ing applications characterized by rolling-sliding conditions.
In order to favour the use of ADI in engineering applications, it
is clearly necessary to investigate their rolling-sliding behaviour,
since several mechanical parts, like gears, experience this con- 2. Experimental procedure
tact conditions during operation [7]. Sugisghita and Fujiyoshi [8]
investigated the role of graphite nodules on the dry rolling-sliding The study was carried out on the materials listed in Table 1.
behaviour of nodular cast iron with a ferritic–pearlitic matrix The cast irons were produced in an industrial foundry. For the ADI
microstructure. They found that graphitic films are formed dur- 800 an austempering cycle was adopted comprising austenitiza-
tion in the range 840–870 ◦ C and austempering in a salt bath at
380–400 ◦ C. For the ADI 1050, austenitization was carried out at
840–870 ◦ C and austempering at 370–390 ◦ C. The specimens for the
∗ Corresponding author. Tel.: +39 0461 882458. rolling-sliding tests were produced by machining the heat-treated
E-mail address: giovanni.straffelini@ing.unitn.it (G. Straffelini). blocks. Some specimens of ADI 1050 were produced before the

0043-1648/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.wear.2010.12.018
 G. Straffelini et al. / Wear 271 (2011) 1602–1608 1603

Table 1
Chemical composition of the materials under study (wt%) and treatment cycles.

Material Chemical composition Treatment cycle

ADI 800 3.13%C, 0.15%Mn, 2.34%Si, 0.022%P, 1.1%Ni, 0.28%Mo, 0.69%Cu Austempering and specimen machining
ADI 1050 3.84%C, 0.15%Mn, 2.34%Si, 0.017%P, 1.1%Ni, 0.24%Mo, 0.7%Cu Austempering and specimen machining
ADI 1050-Ox 3.84%C, 0.15%Mn, 2.34%Si, 0.017%P, 1.1%Ni, 0.24%Mo, 0.7%Cu Austempering after specimen machining
42CrMo4 0.39%C, 0.77%Mn, 0.25%Si, 1%Cr, 0.1%Ni, 0.15%Mo, 0.23%Cu Quenching, stress relieving, specimen machining and gas nitriding

austempering cycle. The successive heat-treatment therefore pro- diffractometry (employing Cu K␣ radiation). The relative amounts
duced an oxide layer on the surface of the specimens. This material of the phases were determined using the Rietveld analysis [10].
is codenamed ADI 1050-Ox. The wear mechanisms were investigated by microstructural
For a comparison, steel specimens submitted to a conventional investigations of the cross-sections of the worn specimens and the
gas nitriding cycle (30 h at 515 ◦ C and 40 h at 540 ◦ C) were also morphological characterization of the worn surfaces. The surface
produced. The steel is codenamed 42CrMo4, and its chemical com- and subsurface layers were also analyzed by means of microhard-
position is included in Table 1. ness measurements using a Vickers indenter and a load of 50 g.
The specimens were disks with a diameter of 40 mm and a
height of 10 mm. Dry rolling-sliding tests were carried out on an 3. Results
Amsler type tribotester. In Fig. 1a schematic representation of the
apparatus is shown. Specimens of identical material and treatment 3.1. Microstructures
ran against each other, with a 10% difference between the rota-
tion speeds of the two disks (400 rev/min for the upper disc and Fig. 2 shows the microstructure of ADI 800. The matrix is char-
360 rev/min for the lower disc). The sliding velocity was therefore acterized by the typical microstructure of austempered irons, i.e.,
0.084 m/s. On the basis of literature data and previous experience, by the presence of bainitic ferrite and retained austenite. According
the following applied loads were selected: 50, 100 and 500 N [7,9]. to X-ray diffraction the amount of retained austenite was 8%. The
The tests were carried out at room temperature and weight changes fraction of graphite nodules was 7%, the average dimension of the
of the upper disc were measured with a precision balance at inter- graphite nodules was 43.8 ␮m and the matrix microhardness was
vals of 10, 20, 50 and 80 min. The upper disc is the driven one and 350 HV 0.1. The ADI 1050 displayed a similar microstructure. The
is characterized by a larger wear damage and subsurface straining content of retained austenite was 12% and the matrix microhard-
because of the particular stress and strain field that occur on the ness was 470 HV 0.1. The fraction of graphite nodules was 8%, and
surface of the disc as a consequence of contact pressure and fric- the average dimension of the nodules was 42.8 ␮m.
tion coefficient [7,9]. The evolution of friction coefficient was also Fig. 3a shows a cross section of material ADI 1050-Ox at the sur-
monitored during each test. face. The presence of a surface oxide layer can be appreciated. Its
The microstructure of the materials was evaluated by optical thickness is in the range 8–10 ␮m, as obtained by measurements
microscopy after etching with 2% Nital solution. The amount and from the metallographic observations. It can be also noted that the
dimension of the graphite nodules were determined by means oxide is quite defective. A more compacted morphology could be
of image analysis system interfaced with the optical microscope. obtained after optimization of the treatment cycle. In addition, oxi-
The phase constitution of the materials was determined by X-ray dation involved also the interface between graphite and the matrix
in the case of the nodules emerging at the surface. This caused the
expulsion of the surface graphite nodules. Fig. 3b shows the XRD
spectrum of the surface of the specimen. The spectrum revealed the
presence of Fe3 O4 (magnetite) and Fe2 O3 (hematite), in addition to
the ausferrite and the retained austenite present in the underly-
ing matrix. The estimated fraction of hematite is 32% and that of
magnetite is 16.5%.

Fig. 1. Schematic representation of the testing rig. Fig. 2. . Microstructure of ADI 800.
1604 G. Straffelini et al. / Wear 271 (2011) 1602–1608

Fig. 3. (a) Morphology of the surface oxide scale in material ADI 1050-Ox; (b) XRD spectrum of the surface of material ADI 1050-Ox.

Fig. 4a shows the compound layer formed after gas nitriding on Fig. 6 shows an example of the results of wear test: mass loss
the surface of the 42CrMo4 steel. As well known, it is made of a of the upper disc is shown as a function of sliding distance (given
mixture of ␧-Fe2-3 N and Fe4 N nitrides [11]. The outer part of the by the sliding velocity, i.e., 0.084 m/s, multiplied by testing time).
compound layer is grey, whereas the internal part is white. This After a running-in distance of about 100 m, a constant slope is
means that the outer part is reacher in ε-nitride and contains some maintained by the curves (steady-state). From the slopes and con-
porosity [11]. In agreement with literature data, the microhardness sidering the density of the material (7.4 g/cm3 ), it is possible to
of the compound layer was about 550 HV 0.02 in the porous region evaluate the wear rates at the three applied loads. They turned
and 900 HV 0.02 close to the interface with the diffusion layer [12]. out to be 4.43 × 10−4 mm3 /m at 50 N, 2.64 × 10−4 mm3 /m at 100 N,
Fig. 3b shows the microhardness profile in the diffusion layer, i.e., in and 4.87 × 10−4 mm3 /m at 500 N. However, for a better comparison
the region below the compound layer where hardness is increased among the materials under study, it is useful to refer to the wear
by the diffusion of nitrogen. coefficient, Ka , calculated from the following equation [13,14]:

W
Ka = (1)
3.2. Friction and wear behaviour FN

Fig. 5a shows the evolution with time of friction coefficient in the where W is the wear rate and FN is the applied load. The rele-
first interval of testing (the first 10 min) in the case of ADI 1050. In vant values are displayed in Fig. 7. They are all lower than about
Fig. 5b the mean friction coefficient recorded during each test inter- 7 × 10−14 m2 /N, confirming a mild wear in all cases. In particular,
val is plotted as a function of testing time for all the materials under for ADI 800 and ADI 1050, the wear coefficients were particularly
study and for an applied load of 500 N. With the only exception of low, between 10−15 and 8 × 10−15 m2 /N.
the nitrided steel (and, in part, of the ADI 1050), friction coefficient
was rather constant during each test. For the nitrided steel, friction 3.3. Surface and subsurface wear damage
coefficient decreased with time and reached the same value dis-
played by the other materials in the last test interval only. In Fig. 5c In Fig. 8a and b, top views of the worn surfaces of ADI 800
the mean values of the friction coefficient recorded during the last at the end of the test at 100 N are shown. Fig. 8a is taken within
test interval are plotted as a function of the applied load for all the a secondary electron microscope (SEM) using backscattered elec-
materials under study. tron examination and clearly shows the squeezing of graphite on

Fig. 4. 42CrMo nitrided steel. (a) Compound morphology; (b) Microhardness profile of the diffusion layer.
 G. Straffelini et al. / Wear 271 (2011) 1602–1608 1605

Fig. 5. (a) Evolution of friction coefficient during the first test interval in the case of ADI 1050; (b) mean friction coefficient recorded during each test interval as a function
of testing time; (c) mean friction coefficient recorded during the last test interval as a function of the applied load.

the sliding tracks. Such squeezing was due to the wearing of the spreading of graphite to the surface can be appreciated (see arrow
graphite nodules emerging at the surface during sliding. Fig. 8b is 1). From the optical micrographs, an evaluation of the equivalent
taken with secondary electrons and better shows the occurrence strain profile could be made using the following relation [15]:
of large surface plastic deformations. Fig. 8c shows the wear frag- √
ments collected at the end of the test. These surface deformations 3
ε(z) = tan (z) (2)
and the morphology of the debris indicate that wear was by delam- 3
ination. where ε(z) is the equivalent plastic strain at a depth z, and (z) is
The consequences of plastic shear deformation due to friction the shear angle between microstructural features bent by shearing
are shown in the cross section view in Fig. 8d. It can be noted that and the normal to the worn surface. The results of this evaluation
plastic shearing produced an alignment of microstructural features are shown in Fig. 9a for ADI 800 and Fig. 9b for ADI 1050. The ADI
in the direction of sliding close to the surface. In particular, the 1050 displayed the same damage characteristics shown by the ADI

Fig. 7. Wear coefficients, Ka , as a function of the applied load for all the materials
Fig. 6. Wear curves of ADI 1050 (upper disc). under study.
1606 G. Straffelini et al. / Wear 271 (2011) 1602–1608

Fig. 8. ADI 800 tested at 100 N. (a) SEM observation with backscattered electrons of the worn surface; (b) SEM observation with secondary electrons of the worn surface. (c)
SEM observation of the wear debris. (d) Cross section observed with optical microscopy. The symbol between the two arrows is the shear angle, , between microstructural
features bent by plastic shearing and the normal to the worn surface.

800. As expected, the equivalent plastic strain decreases as the dis- acterized by a thickness of 5 ␮m and quite a large hardness. The
tance from the contact surface is increased and it increases as the microhardness of the MML was determined by means of indenta-
contact load is increased. In addition, the equivalent plastic strain tions taken on the worn surfaces in correspondence of the metallic
is always larger in the ADI 800 with respect to the ADI 1050. Most (white) regions (see Fig. 8a). The results of such measurements are
probably this result is due to the higher hardness and therefore displayed in Fig. 10 (the experimental point for an applied load of
lower ductility of the ADI 1050 in comparison to the ADI 800. 0 N is the hardness of the unworn surface). It can be noted that the
The arrow 2 in Fig. 8c shows a second important microstructural microhardness is actually quite high showing that within the MML
modification induced by the contact plastic shear deformation, a large strain hardening took place during sliding [17]. The increase
i.e., the formation of a tribological layer, often denominated in hardness with load of the MML is responsible for the decrease in
“mechanically-mixed layer” (MML) [16–18]. Such a layer is char- the friction coefficient displayed in Fig. 5c. It is clear that delamina-

Fig. 9. Evolution of the subsurface equivalent plastic strain profile for ADI 800 (a) and ADI 1050 (b).
 G. Straffelini et al. / Wear 271 (2011) 1602–1608 1607

Fig. 10. Microhardness values of the worn surfaces acquired in correspondence of

the metallic regions, i.e., where graphite is not present.
Fig. 12. Cross-sectional microstructure of the nitrided steel after 50 min of testing at
tion wear was due to the detachment during sliding of fragments 100 N. Arrow 1 indicates a damaged region of the outer porous part of the compound
layer, and arrow 2 indicates a region where such a part was worn away.
from this tribological layer [17].
The wear track of the ADI 1050-Ox displayed different features,
as shown in Fig. 11. Contact shear stresses produced the compaction and detachment during sliding. The MML was formed once a criti-
and smoothing of the oxide layer, as shown by arrow 1. Wear was cal plastic strain was reached at the surface [15,17], and the easier
due to the brittle fragmentation of such scale, as shown by arrow 2. its formation the higher was the wear coefficient. Since the critical
In this case, the graphite nodules did not played a role in the friction strain is proportional to the ductility of the material [19], the mea-
and wear processes, since they were not present on the surface at sured plastic strains for the ADI 800 were found to be larger than
the beginning of the tests (see Fig. 2a). those of the ADI 1050, as shown in Fig. 9. This explains why the ADI
Fig. 12 shows a cross section of the nitrided steel after 50 min of 1050 was characterized by a slightly higher wear coefficient than
testing at 100 N. It can be noted that the compound layer is dam- the ADI 800 (see Fig. 8) despite its higher hardness.
aged. In particular, wear is mainly due to the fragmentation of the The attainment of plastically dominated wear during rolling-
outer porous part of the compound layer. Even at the end of the sliding is justified by the quite large contact pressures. In fact, the
test at 500 N, part of the internal part of the compound layer was initial Hertz pressures ranged from 100 to 300 MPa, and the nor-
still present on the wearing surface. This indicates that in any case malized contact pressure (given by the ratio of the Hertz pressure
wear resistance was determined mainly by the characteristics of and the hardness of the material) ranged between 0.02 and 0.08.
the compound layer, and that the diffusion layer was not directly With reference to the map of Ashby for dry sliding of steels [20],
involved in the damaging process. wear is expected to be by delamination and severe. The experimen-
tal data, however, show that wear was by delamination but mild.
4. Discussion This discrepancy can be explained by considering the high hard-
ness of the MML and, in particular, the role of graphite. As shown
In the testing conditions here considered, the ADI 800 and ADI in Fig. 8a and d, graphite squeezed on to the wear surface during
1050 nodular cast irons underwent wear by delamination during sliding, and this produced a dramatic decrease in friction (see Fig. 5)
rolling-sliding (Fig. 8). In particular, wear was caused by the forma- and in wear rate. In fact, the detachment of the wear fragments was
tion of a hard tribological layer (the MML) and by its fragmentation due to adhesion forces that are strongly reduced by the graphite
coverage [21–23].
As well known, mild wear under dry sliding conditions is usually
attained when the wear mechanism is tribo-oxidation, i.e., when
the wear surface is covered with a protective oxide layer [24–26].
Keeping this in mind, it was decided to test material ADI 1050-
Ox. The wear specimens of ADI 1050-Ox were characterized by the
presence of an Fe-oxide layer on their surface (Fig. 3). The results of
the rolling-sliding tests actually showed that wear was the brittle
fragmentation and removal of the oxide layer (Fig. 11). The resulting
wear coefficient was then typical of mild wear. Friction coefficient
was also quite low, because of the presence of a large amount of
magnetite in the oxide, that is reported to induce a solid lubricant
effect [21,22]. However, wear coefficients were lower than those
displayed by the ADI 800 and ADI 1050. This result is consistent with
the absence of the graphite coverage on the wear surface, because
the surface graphite nodules were removed during the production
cycle.
As shown in Fig. 13, there is a direct relationship between wear
coefficient and friction coefficient. But for a given friction coeffi-
cient, the wear coefficients for the ADI 800 and ADI 1050 were much
Fig. 11. Surface appearance of ADI 1050-Ox tested at 100N. Arrows 1 indicates the
formation of a compacted oxide layer, whereas arrows 2 indicates a region where lower than those of the ADI 1050-Ox. This can be explained by con-
such layer underwent wear by brittle fragmentation. sidering the different role played by the acting wear mechanisms.
1608 G. Straffelini et al. / Wear 271 (2011) 1602–1608

4. As expected, ADI 1050-Ox displayed a mild wear behaviour and

a low friction coefficient because of the presence of magnetite
in the oxide layer. However, wear coefficients were larger than
those of the ADI 800 and ADI 1050, because wear was mainly
due to the brittle fragmentation of the oxide layer.
5. Wear of the nitrided steel was by brittle fragmentation of the
compound layer. The resulting wear coefficients were higher
than those displayed by ADI 800 and ADI 1050. They were how-
ever lower than those displayed by ADI 1050-Ox.
6. A direct relationship between the wear coefficients and the
friction coefficients was evidenced. But for a given friction coef-
ficient, the wear coefficients for the ADI 800 and ADI 1050 were
much lower than those of the ADI 1050-Ox. This was explained
by considering the different role played by the acting wear mech-
anisms.

References
Fig. 13. Relationship between wear coefficient and friction coefficient for all the
materials under study.
[1] ASM Handbook, Properties and Selection: Irons, Steels and High Performance
Alloys, vol. 1, 1990.
In the case of the ADI 800 and ADI 1050, wear was by fragmentation [2] A. Owahdi, J. Hedjazi, P. Davami, Wear behavior of 1.5Mn asutempered ductile
iron, Mater. Sci. Technol. 14 (1998) 245–250.
of the MML (which is a strain-hardened metallic layer) induced by [3] A.S.M.A. Haseeb, Md. Aminul Islam, Md. Mohar Ali Bepari, Tribological
the contact stress due to adhesion. In the case of the ADI 1050-Ox, behaviour of quenched and tempered, and austempered ductile iron at the
wear was by brittle fragmentation of the surface oxides. Even if the same hardness level, Wear 244 (2000) 15–19.
[4] G.X. Lu, H. Zhang, Sliding wear characteristics of austempered ductile iron with
contact friction stresses are similar, the resulting wear rate in the and without laser hardening, Wear 138 (1990) 1–12.
ADI 1050-Ox was higher because of the greater brittleness of the [5] M.J. Perez, M.M. Cisneros, H.F. Lopez, Wear resistance of Cu–Ni–Mo austem-
oxide layer. pered ductile iron, Wear 260 (2006) 879–885.
[6] U.R. Kumari, P.P. Rao, Study of wear behaviour of austempered ductile iron, J.
In order to understand the wear behaviour of the ADIs under Mater. Sci. 44 (2009) 1082–1093.
study it is interesting to compare their performances to those [7] K.H. Zum Gahr, Microstructure and Wear of Materials, Elsevier, Amsterdam,
shown by a nitrided steel. The results here obtained show that the 1987.
[8] J. Sugisghita, S. Fujiyoshi, The effect of cast graphite on friction and wear per-
rolling-sliding resistance of a widely used nitrided steel (42CrMo4
formance III: the lubricating effect of graphite under rolling-sliding conditions,
steel), was lower than that displayed by ADI 800 and ADI 1050. As Wear 77 (1982) 181–193.
shown, the friction and wear behaviour of the nitrided steel was [9] G. Straffelini, A. Molinari, Wear processes in high-strength sintered alloys under
determined by the properties of the surface compound layer, since dry rolling-sliding, Wear 173 (1994) 121–128.
[10] The elaboration of the XRD-spectra by the Rietveld method has been
wear was produced by its brittle fragmentation due to local act- carried out using the program ‘Maud’ developed by L. Lutterotti.
ing friction stresses. Therefore, the higher wear coefficient of the http://www.ing.unitn.it/luttero/maud.
nitrided steel can be attributed to the relatively higher brittleness [11] K.E. Thelnig, Steel and its Treatment, Butterworths, London, 1984.
[12] A. Molinari, G. Straffelini, M. Pellizzari, M. Pirovano, Wear behaviour of diffusion
of the outer porous part of the compound layer. The relatively low and compound layers in nitrided steels, Surf. Eng. 14 (1998) 489–495.
values of the friction coefficient is due to the compound layer. It is [13] I.M. Hutchings, Tribology: Friction and Wear of Engineering Materials, Edward
made by a mixture of nitrides that induce a low friction coefficient Arnold, London, 1992.
[14] M.F. Asbhy, Materials Selection in Mechanical Design, Pergamon Press, Oxford,
[13], and it was present in the contact region for the whole duration 1993.
of the tests. [15] G. Straffelini, D. Trabucco, A. Molinari, Sliding wear of austenitic and
austenitic–ferritic stainless steels, Metall. Mater. Trans. 33A (2002) 613–624.
[16] D.A. Rigney, L.H. Chen, M.G.S. Naylor, A.R. Rosenfield, Wear processes in sliding
5. Conclusions systems, Wear 100 (1984) 195–219.
[17] B. Venkataraman, G. Sundararajan, The sliding wear behaviour of Al–SiC par-
In this study, the dry rolling-sliding behaviour of two austem- ticulate composites. 2. The characterization of subsurface deformation and
correlation with wear behaviour, Acta Mater. 44 (1996) 461–473.
pered ductile irons (ADI 800 and ADI 1050) was investigated. Their [18] X.Y. Li, K.N. Tandon, Mechanical mixing induced by sliding wear of an Al–Si
behaviour was also compared to that of a nitrided steel and an alloy against M2 steel, Wear 225–229 (1999) 640–648.
ADI covered by an oxide layer (ADI 1050-Ox; the oxide layer was [19] G. Straffelini, F. Bonollo, A. Molinari, A. Tiziani, Influence of matrix hardness
on the dry sliding behaviour of 20 vol.% Al3 O3 -particulate-reinforced 6061 Al
produced during the treatment cycle). The following points can be
metal matrix composite, Wear 211 (1997) 192–197.
highlighted. [20] S.C. Lim, M.F. Ashby, Wear mechanism maps, Acta Mater. 35 (1987) 1–24.
[21] G.W. Stachowiak, A.W. Batchelor, Engineering Tribology, Tribology Series 24,
1. Both ADI 800 and ADI 1050 displayed a very low friction coeffi- Elsevier, Amsterdam, 1993.
[22] E. Rabinowicz, Friction and Wear of Materials, Wiley, New York, 1965.
cient because of the action of graphite that squeezed on to the [23] G. Straffelini, A simplified approach to the adhesive theory of friction, Wear
wear surface during rolling-sliding. 249 (2001) 79–85.
2. The wear coefficients of ADI 800 and ADI 1050 were therefore [24] T.F.J. Quinn, J.L. Sullivan, D.M. Rowson, Origins an development of oxidational
wear at low ambient temperatures, Wear 94 (1984) 175–191.
very low, notwithstanding that wear was by delamination. [25] F.H. Stott, The role of oxidation in the wear of metals, Tribol. Int. 31 (1998)
3. ADI 800 displayed a slightly lower wear coefficient, because of 61–71.
its higher matrix ductility that made the formation of the tribo- [26] G. Straffelini, D. Trabucco, A. Molinari, Oxidative wear of heat-treated steels,
Wear 250 (2001) 485–491.
logical layer during rolling-sliding more difficult.