WEAR

ELSEVIER Wear 188 (1995) 61-65

Study of wear of as-cast and heat-treated spheroidal graphite cast iron

under dry sliding conditions
M.A. Islam, A.S.M.A. Haseeb, A.S.W. Kurny
Department of Metallurgical Engineering. Bangladesh University of Engineering and Technology, Dhaka-1000, Bangladesh

Received 9 June 1994; accepted 26 January 1995

Abstract

Wear behaviour of as-cast and heat-treated spheroidal graphite (SG) cast iron has been studied under dry sliding conditions using a pin-
on-disc type apparatus. Wear tests were carried out at a linear sliding speed of 0.88 m s-‘, under a constant load of 1.5 kg. All tests were
performed in ambient air at room temperature. Extent of wear damage and wear mechanisms were investigated by means of weight loss
measurement, optical microscopy, microhardness measurement and X-ray diffractometry on wear debris. The wear rate measured after 9500 m
of sliding is found to be about three times higher in the as-cast sample than in the heat-treated material. In the case of the heat-treated material,
abrasive wear is the main wear mechanism. A combination of adhesive wear, delamination and surface fatigue is believed to operate in the
as-cast material.

Keywords: Wear; Dry sliding; Spheroidal graphite cast iron; Wear mechanism; As-cast and heat-treated iron; Adhesive and abrasive weat

1. Introduction available, information on the wear resistance of this material

is relatively scarce. This study was, therefore, undertaken to
investigate the wear resistance of SG iron both in the as-cast
Spheroidal graphite cast iron (SG iron) is a relatively new and heat-treated conditions. Improved wear resistance of SG
addition to the cast iron family. The essential feature of SG
iron which is expected through heat-treatment will be of great
iron is that the graphitic carbon exists in the form of nodules,
value in gears, pinions, crankshafts and similar parts where
whereas in conventional grey cast iron, the graphitic carbon
wear resistance is of prime importance.
is in the form of flakes. The flaky morphology of graphite
provides easy sites for stress concentration at the sharp edges
of the flakes and this makes grey cast iron extremely brittle. 2. Experimental details
On the other hand, the nodular morphology of the graphitic
carbon in SG iron reduces the stress concentration effect and Spheroidal graphite cast iron, both in the as-cast and heat-
can even act as a crack arrester. SG iron thus possesses much treated conditions, was used in this study. Details of the com-
improved ductility compared with grey cast iron, or even position, production method etc. of the SG iron used can be
steel. SG iron has, therefore, increasingly been used in recent found elsewhere [ 31. The heat-treatment consisted of hard-
years to replace both cast iron and fabricated steel with a ening from 900 “C, followed by tempering at 550 “C for 1 h.
substantial cost advantage in manufacturing [ 11. The as-cast SG iron has a pearlitic matrix (hardness:
The mechanical properties like hardness, strength, elon- 382 KHN) with a ferrite ring around the graphite nodules,
gation, modulus, impact resistance etc. of SG iron are well while the heat-treated SG iron has a matrix of tempered mar-
documented [ 21. Strength and elongation of SG iron can be tensite (hardness: 580 KHN) . Wear tests were carried out in
varied within a wide range by changing its matrix. Ferritic a pin-on-disc type apparatus (Fig. 1) under dry sliding con-
SG iron, which is softer, will give maximum elongation while ditions in the ambient air at room temperature. SG iron in the
pearlitic SG iron is stronger and harder with moderate elon- form of a cylindrical pin of 8 mm diameter and 6.5 mm length
gation. The strength and hardness of the SG iron can be was used. Grey cast iron discs of 80 mm diameter and about
improved further through proper heat-treatment as in marten- 10 mm thickness were used as the counter body. During the
sitic SG irons. Although adequate information about the prop- tests, the SG iron pin was pressed against the disc under a
erties of SG iron relevant to structural applications is constant load of 1.5 kg. The disc speed was 300 rev min- ‘,

0043-1648/95/$09.50 0 1995 Elsevier Science S.A. All rights reserved

SsO10043-1648(95)06605-5
 M.A. Islam et al. /Wear 188 (1995) 6145

Load 5

APin (Fixed)

-.
x -..$s-cast SG iron
7-E 3 .. .. A
Counter body ‘.A_______------

I’
“E
?ii+ & 2.. ,lJ’
2 I
Fig. 1.Schematic of the wear test apparatus
I’ Heat treated SG iran
0
which gave a linear speed of 0.88 m s-’ at wear track. For
each experiment a new pin and a new disc were used. Test
durations ranged from l-3 h. Before the tests, both the pin
and the disc were degreased, cleaned thoroughly in water and 0 2,000 4.000 6,000 8.000 10,000
dried immediately in acetone. All tests were carried out in
ambient air at room temperature. Sliding distance, m
After testing, the worn surface of SG iron pins was exam-
Fig. 2. Wear rate of SG iron pins as a function of sliding distance.
ined by optical microscopy. These were then cleaned thor-
oughly in running water, dried in acetone and again
examined. Pins were weighed before and after the tests to -1
determine the weight loss due to wear. At least three tests
were carried out for each set of conditions and the average
weight loss is reported in this paper. To reveal any change in .*
A
the subsurface region of the SG iron pin, microhardness meas- P _-As-cast
--__ -_
SG iron __,H~*
r*
urements were carried out on cross-sections through the wear I/ - ‘_&’
/I
track, using a Knoop indenter. Identification of the wear
//
debris collected during wear testing was made by X-ray dif-
//
fraction, using Cu K-a radiation.
i ;pG iron, _

3. Results and discussion

The wear rates of both heat-treated and as-cast SG iron are 0 2,000 4,000 6,000 8,000 ro,t
shown as a function of sliding distance in Fig. 2. The wear Sliding distance, m
rate of the as-cast sample initially reaches a high value, then
Fig. 3. Normalized wear rate with respect to contact pressure vs. sliding
decreases and falls to a steady value. In the case of the heat- distance.
treated sample, the wear rate does not show a maximum.
After the initial increase, the wear rate becomes steady. In elongated in the direction of sliding. This indicates consid-
this test geometry, as the test proceeds, pin wear changes the erable deformation around graphite nodules on the as-cast
initial line contact to an area contact. Hence the contact pres- sample. It may be mentioned that the graphite nodules in the
sure decreases continuously during the test. Such a change in as-cast sample are surrounded by soft ferrite rings which are
contact pressure during wear testing is also encountered in likely to get deformed. Sliding marks are observed on both
other test geometries such as in the cross-cylinder test [4]. samples. The as-cast sample sustains more damage due to
The wear data normalized with respect to instantaneous con- wear. The scar on the as-cast sample also shows areas with a
tact pressure is redrawn in Fig. 3. Both Figs. 2 and 3 are compact transfer layer, especially for longer test runs. Fig. 5
qualitatively similar although the difference in the wear rate shows a typical example of a transfer layer. It is not contin-
between the as-cast and heat-treated samples is more pro- uous and occurs randomly on the scar. The wear scar on the
nounced when the nominal contact pressure change is taken heat-treated samples did not contain such a transfer layer.
into account. Fig. 6 shows wear scars after thorough cleaning in running
Fig. 4 shows the wear scars on the as-cast and heat-treated water. The scar on the as-cast sample did not change much
samples after sliding to 6333 m. Both micrographs were taken due to cleaning. Patches of debris, greyish in colour, are
immediately after the test without any cleaning. Greyish attached to the surface. Deformed areas are seen on the scar.
debris covers much of the surfaces. Dark areas are graphite The wear scar on the heat-treated sample is clean and smooth.
nodules. The graphite retains its nodular shape on the heat- Very fine ploughing/sliding marks are the main feature of
treated samples; while on the as-cast sample, the graphite is this scar. The wear action has caused mainly polishing on this
 M.A. Islam et a.! /Wear 188 (1995) 61-65 63

greater than that of the Fe304 line, possibly indicating a

greater amount of Fe0 in the debris. It is clear that during the
wear process, oxidation of iron had taken place. It is well
known that oxidation also takes place during the wear of steels
[ 5,6]. Although both Fe304 and Fe0 were identified, Fe203
could not be detected in the debris. This may be explained in

Fig. 5. Transfer layer formed on as-cast sample after 9500 m of sliding.

Fig. 4. Optical micrographs of wear scars after 6333 m of sliding. Micro-

graphs were taken immediately after the tests without cleaning of (a) the
as-cast sample and (b) the heat-treated sample. The sliding direction is
represented by the horizontal.

sample. Fig. 7 shows the cross-section through wear scars on

both samples, the top being the scar surface. The edge of the
wear scar on the heat-treated sample is sharp and smooth
whereas that on the as-cast sample is rough and irregular.
Spalling of layers at the surface of the as-cast sample is seen
(as indicated by the arrow).
From the examination of the surface and cross-section of
wear scars, the dominant wear mechanism can be suggested.
For the heat-treated samples, abrasive wear is the main mech-
anism. Fine debris particles generated in the wear process are
believed to cause the microcutting/ploughing on the surface.
Abrasive action of this fine debris also smoothens the scar.
On the other hand, several mechanisms are believed to oper-
ate in the case of the as-cast sample. The presence of the
transfer layer suggests the occurrence of adhesive wear in the
as-cast sample. Delamination and/or surface fatigue also
occurs, which causes spalling of material as seen in Fig. 7(b) .
Abrasive wear seems to play an insignificant role in this case.
X-ray diffraction data obtained from the wear debris from
the heat-treated sample during 9500 m of sliding is summa-
Fig. 6. Wear scar after thorough cleaning in running water of (a) the as-cast
rized in Table 1. The debris consists mainly of iron powder, sample and (b) the heat-treated sample. Both samples were tested to a sliding
Fe0 and Fe304 powder. The intensity of the Fe0 line is distance of 3166 m.
 M.A. Islam et al. /Wear 188 (1995) 61-65

takes place with the generation of lamellar wear particles. At

low loads and speeds, mild wear is encountered and the cor-
responding wear particles are very fine and oxidized. This
has also been confirmed by others [9]. In the present study
of cast irons, where a load of 1.5 kg and speed of 0.88 m ’ s-
have been used, wear particles are very fine and oxidized and
the wear is of the mild type.
Sexton and Fischer [ 91 and Czichos [ lo] have found that
the wear mechanism in steel depends upon its hardness. In
martensitic steel with higher hardness, abrasive wear tends to
take place. But for low hardness values, adhesive wear is
encountered. The results of the present study also conforms
to the above. For heat-treated SG iron with a martensitic
structure, abrasive wear has been found to be the main mech-
anism while on the softer as-cast samples, adhesive wear has
taken place.
During wear tests, the properties in the surface region of
the specimen may change and, thus, influence the wear proc-
ess. To investigate whether such a change has taken place,

“0°0
I

50 firn
0: _I
Fig. 7. Optical micrographs showing cross-section through wear scar on (a) 0 40 80 120 160
the as-cast sample and (b) the heat-treated samples. The sliding direction is
perpendicular to the plane of the paper. (a) Distance below wear scar, grn
1,000
Table 1
Summary of the X-ray data of the wear debris collected during 9500 m
sliding of the heat-treated sample 0
z 800
0
Possible phases (M) Relative intensity I 0
Peak position (“) 0
Y 00 0
0
0 0
35.6 Fe304 (311) weak vi 600
- B u u 08 0
41.91 Fe0 (200) strong 2 ,8Boo:o -u
S bo - u
44.7 Fe (110) strongest
5
82.5 Fe (211) weak I 400
S!
.I!
ZE
terms of the free energy of formation of oxides. The standard 200
free energy of formation of Fez03 is much less negative in
the temperature range of O-2000 “C compared with those of
OL
Fe0 and Fe304 [ 71. The formation of Fe0 and Fe304 is, 0 50 100 150
therefore, more likely than the formation of Fe,O,.
Welsh [ 81 studied the wear behaviour of steels in the load ( b) Distance below wear scar, ,um
range of 0.05-40 kg and speed range of 0.017-2.66 m s-l, Fig. 8. Variation of Knoop hardness below the wear scar as a function of
observing that wear can be severe or mild depending upon depth, (a) the as-cast and b) the heat-treated specimens. Both specimens
the load and speed. At higher loads and speeds, severe wear were tested up to 9500 m of sliding.
 M.A. Islam et al. /Wear 188 (1995) 6165 65

microhardness tests were carried out on the cross-sections I31 S.S.A. Mogni, M.Sc. Eng. Thesis, Bangladesh University of
below the wear scar on the as-cast and heat-treated samples. Engineering and Technology, Dhaka, 1992.
[41 Wear Testing with a Crossed-Cylinder Apparatus, G83, American
Fig. 8 shows the Knoop hardness as a function of distance
Society for Testing and Materials, Philadelphia, PA, 1984.
below the surface. A slight increase in hardness is detected [5] P.D. Goode, Nucl. Instrum. Meth. Phys. Res., 839 (1989) 521-530.
near the surface of the as-cast sample, whereas the heat- [6] B.M. Khusid, E.M. Khusid and B.B. Khina, Wear, 165 (1993) 109-
treated sample shows a decrease. Considerable increase in 112.
hardness near the surface during wear is reported by others [71 F.D. Richardson and J.H. Jeffes, J. Iron Steel ht., 160 (1948) 261.
[ 81 N.C. Welsh, Phil. Trans. Roy. Sot. (Lond.), Ser. A, 257 ( 1964) 31.
[ 111. The increase in hardness of the as-cast sample in the
[9] M.D. Sexton and T.E. Fischer, Wear, 96 (1984) 17.
present case is thought to be due to strain hardening of the [ 101 H. Czichos, in N.R. Loonis (ed.), New Direction of Lubrication,
pearlitic matrix at the surface region which predominates over Materials, Wear and Surface Interaction in Tribology in the 80’s,
any frictional heating effects. The decrease in hardness at the Noyes, NJ, 1989.
[ 111 J. Don, Ph.D. Thesis, Ohio State University, 1982.
contact surface of the heat-treated sample has been caused by
temperature rise due to friction. The martensitic matrix is
believed to be tempered by the frictional heat during sliding.
Biographies

M.A. Islam: obtained his B.Sc.Eng. (metallurgical) degree

4. Summary and conclusion
in 1992 and M.Sc.Eng. (metallurgical) degree in 1994 from
the Bangladesh University of Engineering and Technology,
In this study the wear behaviour of as-cast and heat-treated
Dhaka. After obtaining his B.Sc.Eng. degree, he joined the
spheroidal graphite cast iron has been investigated. The dif-
Department of Metallurgical Engineering, BUET, Dhaka as
ference between the as-cast and the heat-treated samples lies
a lecturer in 1992. He is now working as an assistantprofessor
in their matrix structure. The martensitic matrix of the heat-
in the same department. His present research interest is sur-
treated sample is harder than the pearlitic matrix of the as-
face engineering and heat-treatment of new materials.
cast sample. The high hardness of the heat-treated samples
results in lower wear rates as compared with the as-cast sam-
A.S.M.A. Haseeb: received his M.Sc. Eng. (metallurgical)
ples. Wear mechanism is also seen to be different. While
degree from the Bangladesh University of Engineering and
heat-treated samples wear mainly by an abrasive mechanism,
Technology (BUET), Dhaka in 1986 and Ph.D. in metallurgy
the as-cast samples undergo a combination of adhesive wear,
and materials engineering from the Catholic University of
delamination/surface fatigue. Under the present experimen-
Leuven, Belgium in 1992. He is presently working as an
tal conditions, wear is thought to be of the mild, oxidative
assistant professor at the Department of Metallurgical Engi-
type. The hardness near the surface of the heat-treated sam-
neering, BUET, Dhaka. His research area includes wear, sur-
ples is found to decrease during testing because of tempering
face engineering and thin films.
caused by frictional heating. But the as-cast sample shows a
slight increase in hardness due to strain hardening effect.
A.S.W. Kurny: was born in 1949. He obtained his B.Sc. Eng.
However, this hardness increase is not high enough to cause
(metallurgical) degree in 1970 and M.Sc. Eng. (metallur-
any subsequent decrease in wear rate in the as-cast sample.
gical) degree from the Bangladesh University of Engineering
and Technology, Dhaka in 1973. Then he joined the Depart-
ment of Metallurgical Engineering, BUET, Dhaka as a lec-
References
turer in 1974. After obtaining his Ph.D. in metallurgy from
the Indian Institute of Science, Bangalore in 1982, he has
[l] A.A. Zavaras and H.D. Broady, US Government Research
Announcement Index, PE-126647/X4& National Center for
been working as a professor in the Department of Metallur-
Manufacturing Science, 1990. gical Engineering, at BUET, Dhaka. His current areas of
[2] G.N.J. Gilbert, BCIRA J., 16 (4) (1968) 342. interest are surface engineering and minerals processing.