Professional Documents
Culture Documents
Abstract
Steam oxidized sintered iron-based materials have been, and continue to be used for load bearing parts in sliding contacts. Although the
permanence of an oxide layer on the external surface is reported to be the main factor affecting tribological behaviour, in industrial practice
as well as in laboratory tests abrasive wear is reported to play an important role in the final stage of the process. The removal of the superficial
oxide layer, its fragmentation and transfer to the counter-body associated with metallic debris generation, induces the formation of hard
debris particles acting as abrasives on the tribological system. In this paper, abrasive wear of steam-treated sintered iron is analysed by
means of micro-abrasion tests. Sixteen specimens were produced from atomised iron powders (Ancor Steel 1000B Höganäs) of different
sizes (<65, 65–90, 90–125, >125 m). They were compacted at four different pressures (300, 400, 500 and 600 MPa), sintered for 30 min at
1120 ◦ C and then subjected to continuous steam treatments at 540 ◦ C for 2 h. The micro-abrasion tests were conducted under constant load
and speed conditions while special emphasis was given to the nature of the abrasive; two exogenous (Al2 O3 and SiO2 ) and one endogenous
abrasive (Fe2 O3 ) were used. The results show that the abrasion resistance is strongly influenced by the processing parameters, the lowest wear
coefficients were obtained by using the highest compaction pressures and the smallest powder grade. The pore size is the main micro-structural
parameter affecting the abrasive resistance. For all abrasives, interrupted tests showed that the small pores are completely filled by abrasive
particles in the initial stage of the tests while the biggest ones remains open until the completion of the tests. Wear coefficients were strongly
influenced (up to 10 times) by the abrasive nature. Scanning electron microscopy analysis showed that the abrasive wear mechanisms were
determined by the abrasive particle features.
© 2004 Elsevier B.V. All rights reserved.
0043-1648/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.wear.2004.09.042
W.M. da Silva et al. / Wear 258 (2005) 166–177 167
showed that the wear rate is low and independent of the chem- given to the nature of the abrasive; two exogenous (Al2 O3
ical composition of the specimens as long as the oxide layer and SiO2 ) and one endogenous abrasive (Fe2 O3 ) were used.
produced by the steam oxidation is able to carry out its lubri-
cating function between the sliding surfaces. After removal of
the oxide layer by adhesion, two different wear mechanisms 2. Experimental procedure
are active according to the sliding speed: a delamination wear
mechanism at low sliding speed and a delamination-oxidation Specimens were produced from Ancor Steel 1000B
mechanism at high sliding speed. In the latter case, the wear Höganäs atomised iron powder. The powder was sieved into
rate is lower, the influence of the mechanical properties of four different fractions (<65, 65–90, 90–125 and >125 m)
the matrix is reduced, and the behaviour of pure iron tends to to provide four different powder grades (denoted 1–4, re-
be comparable with that of the alloyed materials. spectively). After mixing with 0.8 wt.% zinc stearate as a lu-
de Mello and Hutchings [9] studied the behaviour of sin- bricant, the powders were compacted at four different com-
tered and oxidized iron samples worn in reciprocating sliding paction pressures (300, 400, 500, 600 MPa) with a double
tests. They showed that during the wear tests fragmentation action automatic press in such a way that the surfaces to be
of the oxide layer occurs. Besides, greatly deformed metallic tested were in contact with the moving punch.
particles are produced from both the sample and the counter- The samples were designated as Axy where x is the com-
body. These particles become adherent to the worn surface, paction pressure (in MPa) divided by 100 and y the pow-
generating a compacted layer of iron and iron oxide. The der grade. For example: A34 denotes a sample pressed at
gradual and localized degradation of this layer is the main 300 MPa, with particle size >125 m. The resulting com-
factor controlling the wear process. In the final stage of the pacts, 50 mm in length by 10 mm in width with height de-
oxide layer removing process, loose and hard oxide debris pending on compaction pressure and powder grade, were sin-
caused abrasive wear as well as localized fragmentation of tered in an industrial furnace. The sintering was conducted
the compacted layer. These authors [9,10] showed that the in a mildly reducing atmosphere (N2 + 10% H2 ) at a con-
abrasive wear mechanism is the main degradation agent in stant temperature of 1120 ◦ C for 25–30 min. After cooling to
the final stage of the process. room temperature the samples were steam-treated. The steam
Other authors [11–15] highlight the effect of the particles treatment was carried out at 540 ◦ C for 2 h in a continuous
presence on the global wear behaviour of oxidized sintered industrial furnace.
materials. However only two works, Amsallem et al. [16] and The micro-structure of the samples used in this work
Razavizadeh and Davies [17], consider a direct approach to was comprehensively characterized in a previous paper by
abrasive wear behaviour. de Mello et al. [6]. These authors studied both superficial
In this paper, the effect of processing parameters (com- and volumetric aspects of the micro-structure. The micro-
paction pressure and powder size) on abrasive wear of steam- structure is greatly dependent on the processing parameters
treated sintered iron is analysed by means of micro-abrasion (compaction pressure and powder grade). Thickness of oxide
tests. The micro-abrasion tests were conducted under con- layer is nearly constant (3.5–4 m) and independent of the
stant load and speed conditions while special emphasis was processing conditions used. X-ray diffraction showed that the
Table 1
Micro-structural characteristics [6]
Sample CPa (MPa) PGb Porosity (%) Pore size (m) MFPc (m) Oxide content (%) HV (kgf mm−2 )
A31 300 1 42.7 ± 3.3 14.8 ± 2.8 19 ± 8 29 ± 2 152 ± 4
A32 300 2 39.8 ± 3.8 15.1 ± 4.3 22 ± 8 27 ± 2 105 ± 4
A33 300 3 37.4 ± 1.2 15.5 ± 4.9 25 ± 9 20 ± 1 87 ± 2
A34 300 4 33.1 ± 3.4 15.8 ± 5.2 32 ± 0 19 ± 1 76 ± 2
A41 400 1 36.8 ± 2.0 13.5 ± 2.6 23 ± 2 22 ± 2 141 ± 1
A42 400 2 32.4 ± 2.3 14.2 ± 3.8 29 ± 6 20 ± 1 131 ± 2
A43 400 3 29.9 ± 1.7 14.6 ± 4.1 34 ± 1 16 ± 1 103 ± 3
A44 400 4 26.6 ± 1.9 14.4 ± 4.2 39 ± 8 15 ± 2 93 ± 4
A51 500 1 29.5 ± 1.1 7.9 ± 2.2 25 ± 3 18 ± 1 124 ± 2
A52 500 2 26.6 ± 1.3 9.5 ± 2.9 32 ± 6 14 ± 2 115 ± 2
A53 500 3 22.8 ± 3.2 10.6 ± 3.5 35 ± 8 14 ± 1 105 ± 3
A54 500 4 21.7 ± 1.4 10.9 ± 3.7 39 ± 2 12 ± 1 92 ± 2
A61 600 1 25.5 ± 1.8 6.8 ± 1.8 19 ± 8 14 ± 2 116 ± 2
A62 600 2 24.1 ± 1.2 7.4 ± 2.1 27 ± 3 10 ± 1 105 ± 3
A63 600 3 20.6 ± 1.6 8.1 ± 2.7 31 ± 1 10 ± 1 102 ± 1
A64 600 4 18.9 ± 1.6 8.4 ± 3.1 36 ± 1 9±1 98 ± 1
a Compaction pressure.
b Powder grade.
c Mean free path between pores.
168 W.M. da Silva et al. / Wear 258 (2005) 166–177
πb4
k= for b << R (1)
64RSN
3. Results and discussion tends to be spherical while the alumina particles are more
elongated and sharper.
The mean size of abrasives and their particle size distri- Fig. 2 presents a typical aspect of the craters produced by
butions as well as their morphology are shown in Fig. 1. The the tests. It is clear that the test conditions impose a spherical
haematite presents an asymmetric normal distribution and the wear scar geometry, validating the use of Eq. (1) to determine
smallest mean particle size. The mean particle size increases the wear coefficient by optically measuring the diameter of
for silica and alumina, which presents the biggest particle the imposed spherical crater.
size. The same statistical distribution is presented by the sil- The evolution of the wear coefficient with test time is il-
ica particles while the alumina particles present a bi-modal lustrated in Fig. 3. This figure shows the results presented
distribution. The shape of the silica and haematite particles by samples produced in the most extremes conditions of the
W.M. da Silva et al. / Wear 258 (2005) 166–177 169
Fig. 2. Typical aspect of the craters obtained by laser interferometry. Sample A61, alumina, 360 s. (a) 3D surface; (b) trace AA. The trace of the theoretical
profile of a perfectly smooth sphere, 10 mm in diameter, is shown for comparison.
processing parameters when abraded by silica. They are typ- in thickness) had already been penetrated when the steady
ical of all the tested conditions. Wear coefficients were ob- state wear regime was established, suggesting that the micro-
tained in three sets (Kt1, Kt2 and Kt3), of 13 tests each. The abrasive wear process is governed by the substrate behaviour
mean value (M) was calculated and is also shown in the fig- in close correlation with the tribological conditions act-
ure. The averaged crater depth (d) for each test time is also ing at the final stages of the wear process of this fam-
included. ily of materials as reported by de Mello and Hutchings
The wear coefficient increases in a random manner with [9].
the test time up to 80 s. After that time the dispersion de- Fig. 4 presents the effect of processing parameters on the
creases and the evolution of the wear coefficient becomes wear coefficient. For all abrasives, the smallest abrasive wear
roughly constant. For the low compaction pressure, samples coefficients were obtained by the combination of a high com-
A31 and A34, Fig. 3a and b, the initial transient is more accen- paction pressure with a fine powder grade. In general, for the
tuated and the wear coefficient is stable for test times above same powder grade, an increase in compaction pressure led to
180 s. For the high compaction pressure, samples A61 and a decrease in the wear coefficient. On the other hand, for the
A64, Fig. 3c and d, the stabilization of the wear coefficient is same compaction pressure an increase in powder size gave
reached after 100 s of test time. The permanent wear regime, rise to a slight decrease in the wear coefficient.
i.e. the stability of the wear coefficient was reached for all The effect of porosity on abrasive wear resistance is quite
tested conditions (sample/abrasive pair) after 180 s [20]. In complex as illustrated by Fig. 5. In general, for a fixed powder
this paper, all the analyses were done in the permanent wear grade, increasing porosity always caused an increase in the
regime. This regime is represented by an averaged wear co- wear coefficient. On the other hand, at constant compaction
efficient (K), obtained by the averaging of the results relative pressure, an increase of porosity produced a decrease in the
to the test times 180, 240, 300 and 360 s, e.g. this coefficient wear coefficient.
is representative of an average of 12 tests carried out in each Other micro-structural parameters like the mean free path
sample/abrasive condition [20]. between pores and the amount of oxide formed on the in-
The crater depth increases logarithmically with test time terconnected porosity showed a complex relationship with
until the permanent wear regime is reached. In this regime, abrasive wear resistance [20,21].
the evolution of the crater depth with time is linear. For all tri- It is also noticeable in Fig. 4 that the nature of the abra-
bological conditions, the superficial oxide layer (3.5–4.0 m sive has a strong influence on the abrasive wear behaviour
170 W.M. da Silva et al. / Wear 258 (2005) 166–177
Fig. 3. Evolution of the wear coefficient and crater depth with test time.
Silica. (a) A31; (b) A34; (c) A61; (d) A64.
Fig. 4. Effect of processing parameters on wear coefficient. (a) Haematite;
(b) silica; (c) alumina.
of steam-treated sintered iron. The highest wear coefficients
were produced by alumina while abrasion by haematite pro-
duced the lowest wear coefficient. The general behaviour A34. For sample A61, wear coefficient decreases from alu-
can be more easily understood by analysing the influence mina to silica and then shows a slight decrease from silica to
of the nature of abrasive on the wear coefficient of the sam- haematite. On the other hand, sample A64 presents an approx-
ples produced when using the extreme processing parame- imately constant wear coefficient for silica and haematite.
ters, e.g. A31, A34, A61 and A64, Fig. 6. It can be seen in This fact suggests that the samples produced with small com-
Fig. 6 that the wear coefficient for sample A31 decreases in paction pressure (300 MPa) are more sensitive to the type
the following sequence: alumina, silica and haematite. Sim- of abrasive than the samples produced with higher com-
ilar behaviour is observed for the wear coefficient of sample paction pressure (600 MPa). In all the studied cases, an in-
W.M. da Silva et al. / Wear 258 (2005) 166–177 171
Fig. 5. Effect of porosity on wear coefficient. (a) Haematite; (b) silica; (c)
alumina.
Fig. 7. Wear mechanisms. SEM. Centre of the crater. (a) Sample A64
abraded by alumina; (b) sample A61 abraded by silica; (c) sample A64
abraded by haematite.
alumina (11.5–27) to silica (4.8–17) and then to haematite abrasive size the greater the wear of the samples. Besides,
(6.4–14). This decrease in severity correlates very well with the individual load per abrasive particle is one of the most
the wear coefficients presented by the studied samples e.g. important parameter controlling the wear mechanism acting
the higher the relative hardness (and as a consequence the on the micro-abrasion process [25].
severity of the abrasive process) the higher the wear co- Fig. 7 shows typical aspects of the abraded surfaces and il-
efficient. In addition, Fig. 2 shows that alumina particles lustrates the wear mechanism acting according to the nature
present a sharp, angular, elongated geometry while the sil- of the abrasive. When using alumina, the dominant mech-
ica and haematite grits are blunted and spheroidal in shape. anism was like that of a three-body process, in which the
It has been reported that particles with an angular shape abrasive particles roll between the two surfaces producing
cause higher wear than spherical particles [22] thus induc- a heavily deformed, multiply indented wear surface with no
ing an even more severe wear regime when using alumina evident directionality, Fig. 7a. On the other hand, a grooving
grits. wear mechanism similar to a two-body mechanism was found
The severity of the wear system is also related to the abra- to be dominant for silica and haematite. This process occurs
sive size: an increase in abrasive mean size decreases the in the micro-scale abrasion test when a significant proportion
number of particles present at the interface, thus promoting a of the particles slide without rolling at the specimen/ball in-
substantial increase on the force per abrasive particle. Fig. 2 terface, thus producing a series of fine parallel grooves in the
shows that alumina grits have the largest mean diameter, fol- specimen surface, Fig. 7b and c. Trezona et al. [25] suggested
lowed by silica and haematite, respectively. The greater the that the abrasive particles are embedded in the surface of the
ball bearing and act as fixed indenters.
In addition, the analysis of the wear mechanisms showed
that those produced by silica grits remarkably reproduced the
Fig. 10. Typical aspects of abraded surfaces. Test time 360 s. SEM. (a) A31 abraded by SiO2 ; (b) A34 abraded by SiO2 ; (c) A61 abraded by SiO2 ; (d) A64
abraded by SiO2 ; (e) A31 abraded by Fe2 O3 ; (f) A34 abraded by Fe2 O3 ; (g) A61 abraded by Fe2 O3 ; (h) A64 abraded by Fe2 O3 .
wear mechanisms acting on the real conditions experienced For the exogenous abrasives (Al2 O3 and SiO2 ), increas-
by steam-treated sintered iron components, Fig. 8. This sug- ing the hardness produces a decrease in the wear coefficient,
gests that the use of micro-abrasion tests using the same tri- which reaches a minimum, Fig. 9a and b. It can be noticed
bological conditions may constitute a powerful tool for the that the samples associated with hardness values lower than
optimisation of these components. the value corresponding to the minimum in the wear coef-
The correlation between hardness and abrasive wear is ficient were produced by using the larger powder grades (3
quite complex too, Fig. 9. and 4), while those associated with values above the min-
174 W.M. da Silva et al. / Wear 258 (2005) 166–177
Fig. 12. Typical aspects of the interactions between abrasive particles and pores. SEM. (a) A34 abraded by Al2 O3 , test time 10 s. (b) A34 abraded by Al2 O3 ,
test time 240 s. (c) A64 abraded by Al2 O3 , test time 10 s. (d) A64 abraded by Al2 O3 , test time 240 s. (e) A34 abraded by Fe2 O3 , test time 30 s. (f) A34 abraded
by Fe2 O3 , test time 240 s. (g) A61 abraded by Fe2 O3 , test time 30 s. (h) A61 abraded by Fe2 O3 , test time 240 s.
176 W.M. da Silva et al. / Wear 258 (2005) 166–177
distributions in the surface itself as well as their interactions nism of pore closure, the micro-abrasive wear behaviour is
with the pores. After coating the samples with gold, they were governed by a compromise between hardness and pore size.
observed by scanning electron microscopy. Wear coefficient decreases with hardness up to a critical value
(roughly 110 kgf mm−2 ), Fig. 9. From this point, which corre-
Fe(s) → Fe(aq) 2+ + 2e− (E0 = 0.44 V) (3) sponds to a critical value in pore size, the hardness continues
to increase, the pore size also continues to increase, as shown
Cu(aq) 2+ + 2e− → Cu(s) (E0 = 0.34 V) (4)
by Figs. 9 and 10, and negatively affects the wear resistance.
Fe(s) + Cu(aq) 2+ → Fe(aq) 2+ + Cu(s) (E0 = 0.78 V) (5)
References [12] T.S. Eyre, R.K. Walker, Wear of sintered materials, Powder Metall.
19 (1976) 22–30.
[1] A. Molinari, G. Straffelini, Dry sliding behaviour of steam treated [13] K. Gopinath, The influence of speed on the wear of sintered iron
sintered iron alloys, Wear 159 (1992) 127–134. based materials, Wear 71 (1981) 161–178.
[2] R. Binder, Influência dos Processos de Ferroxidação nas Pro- [14] S.C. Lim, J.H. Brunton, The unlubricate wear of sintered iron, Wear
priedades do Ferro Sinterizado, M.Sc. Dissertation, Universidade 113 (1986) 371–382.
Federal de Santa Catarina, Florianópolis, 1996, 121 pp. (in Por- [15] E.R. Leheup, D. Zhang, J.R. Moon, Fretting wear of sintered iron
tuguese). under low normal pressure, Wear 221 (1998) 86–92.
[3] P. Franklin, B.L. Davies, The effects of steam oxidation on porosity [16] C. Amsallem, A. Gaucher, G. Guillot, The unlubricate frictional
in sintered iron, Powder Metall. 20 (1977) 11–16. behaviour of sintered iron, Wear 23 (1973) 97–112.
[4] K. Razavizadeh, B.L. Davies, Influence of powder type and density [17] K. Razavizadeh, B.L. Davies, The effects of steam treatments on the
on pore closure and surface hardness changes resulting from steam wear resistance of sintered iron and Fe–Cu alloys, Wear 69 (1981)
treatment of sintered iron, Powder Metall. 22 (1979) 187–192. 355–367.
[5] P. Beiss, Steam treatment of sintered parts, Powder Metall. 34 (1991) [18] K.L. Rutherford, I.M. Hutchings, Theory and application of a micro-
173–177. scale abrasive wear test, J. Testing Eval. 25 (1997) 250–260.
[6] J.D.B. de Mello, R. Binder, A.N. Klein, I.M. Hutchings, Effect [19] R.I. Trezona, I.M. Hutchings, Three-body abrasive wear testing of
of compaction pressure and powder grade on microstructure and soft materials, Wear 233–235 (1999) 209–221.
hardness of steam oxidized sintered iron, Powder Metall. 44 (2001) [20] W.M. Da Silva, Efeito da Pressão de Compactação e Tamanho do
53–61. Pó de Ferro na Microabrasão do Ferro Sinterizado Oxidado a Vapor,
[7] K. Volenı́k, H. Volrábová, J. Neid, M. Seberı́ni, Structure of oxida- M.Sc. Dissertation, Universidade Federal de Uberlândia, Uberlândia,
tion products of sintered steel in superheated steam, Powder Metall. Brazil, 2003, 98 pp. (in Portuguese).
21 (1978) 149–154. [21] W.M. Da Silva, R. Binder, J.D.B. de Mello, Efeito do Tipo de Abra-
[8] A. Molinari, G. Straffelini, Surface durability of steam treated sin- sivo sobre o Desgaste Abrasivo do Ferro Sinterizado e Oxidado a
tered iron alloys, Wear 181–183 (1995) 334–341. Vapor, in: Proceedings of the 58th Annual Congress of ABM (Brazil-
[9] J.D.B. de Mello, I.M. Hutchings, Effect of processing parameters ian Association for Metallurgy and Materials), Rio de Janeiro, Brazil,
on the surface durability of steam-oxidized sintered iron, Wear 250 July, 2003, pp. 740–749 (in Portuguese).
(2001) 435–448. [22] I.M. Hutchings, Tribology: Friction and Wear of Engineering Mate-
[10] J.D.B. de Mello, I.M. Hutchings, Surface durability of steam oxi- rials, CRC Press, Boca Raton, 1992, p. 273.
dised sintered iron: effect of surface topography, in: Key Papers of [23] K.L. Zum Gahr, Microstructure and Wear of Materials Tribology
the Second World Tribology Congress, Vienna, Austria, 2001, pp. Series 10, Elsevier, 1987, p. 560.
105–112. [24] G.W. Stachowiak, A.W. Batchelor, Engineering Tribology, 2nd ed.,
[11] A.C. Bozzi, R. Binder, J.D.B. de Mello, Effect of industrial process- Butterworths–Heinemann, 2001, p. 744.
ing parameters on the surface durability of steam oxidised sintered [25] R.I. Trezona, D.N. Allsopp, I.M. Hutchings, Transition between two-
iron, in: Proceedings of the Sixth International Tribology Conference body and three-body abrasive wear: influence of test conditions in
(AUSTRIB’02), Perth, Australia, December 2–5, 2002, pp. 383–390. the micro scale abrasive wear test, Wear 225–229 (1999) 205–214.