ORIGINAL RESEARCH PAPER

Sliding wear behaviour of laser surface

modified pearlitic rail steel
S. M. Shariff*1, T. K. Pal2, G. Padmanabham1 and S. V. Joshi1
Improving wear resistance of rails has a direct impact on the performance of rail–wheel system in
railroad technology. Enhancement of sliding wear resistance at curved track, where factors such
as adhesion, high slip ratios and contact fatigue act at contact patch of rail–wheel system, is
particularly desirable. In the present investigation, influence of laser surface modification on
sliding wear performance of a pearlitic rail steel (used in Indian railways) under two different
conditions, namely, laser hardening (without any melting involved) and laser melting (with thin
surface layer melting), has been studied under laboratory conditions. Before sliding wear testing,
the effect of laser scanning speed on the treated layer depth has been optimised, utilising a 9 kW
CO2 laser system. Sliding wear tests were carried out using a pin-on-disc device, with laser
treated and untreated pearlitic rail steel discs and sliding pins made of wheel steel material,
tungsten carbide (WC) and high speed steel (HSS). The tests were performed under normal
prototypic loads and unlubricated conditions. Microhardness in the laser melted layer was in the
range of 830–900 HV as against 890–1070 HV in the hardened layer, and was found to depend
on the laser scanning speed. Sliding wear resistance of both hardened and melted layers was
found to be significantly improved compared to untreated rail steel. The coefficient of friction was
also marginally reduced in the laser surface melted layers.
Keywords: Laser hardening, Laser melting, Rail steel, Sliding wear, Adhesion, Austenite

Introduction leading to reduced sliding wear can considerably

improve the life of rails, without adversely influencing
Steady increase in average car loading, annual tonnage the performance of wheels.
and concentration of traffic on main lines has placed Another problem of concern in a rail–wheel system is
more stringent demands on rails worldwide. Factors subsurface cracking of rail at gauge corners over a
contributing to the performance and durability of period of its usage, which could be attributed to severe
railroad components include wear resistance, internal cyclic loading that results in strain accumulation and
fatigue damage resistance, toughness, plastic deforma- deformation.9–11 The above problem gets exacerbated at
tion resistance, residual stresses and weldability.1–8 curves, crossings, swings, etc., where the combined
Engineers involved in maintenance of rails and wheels loading from normal forces and steering forces is higher.
have always had to cope with the progression of wear on Such cracking can have serious consequences and may
rails and wheels. This progression was observed to sometimes even lead to derailment of trains. Reduction
change the shape and condition of rail and wheel, in friction in the above regions by adopting remedies
leading to various defects over a period of their usage. such as mechanical lubrication, boride coating, etc. can
These defects could be of different types, ranging from considerably extend rail life.12 Although such remedies
mix of wear to rolling contact fatigue (RCF) damage to were found to yield considerable life extension in rails
plastic flow and corrugations. For the rails and wheels to and wheels, they pose certain inherent associated
be reusable again, these defects are periodically removed disadvantages such as labour intensiveness, complexity
by reprofiling to ensure improved rail–wheel contact. in application and environmental pollution. Recent
Hence, improving wear resistance of rail and wheel can reports suggest possible improvements by utilising laser
directly impact extension of their lifetimes. In particular,
surface treatment technologies such as hardening,
appropriate surface treatment of curved portion of rail
melting/glazing and cladding of rail/wheel.13–17 While
wear resistance can be enhanced through laser hardening
1
Center for Laser Processing of Materials, International Advanced and laser cladding, mitigation of rail–wheel friction in
Research Center (ARCI) for Powder Metallurgy and New Materials, undesirable areas of rails is possible through laser
Hyderabad, India
2
Department of Metallurgical and Materials Engg, Jadavpur University,
melting/glazing.
Kolkata, India Various conventional methods of heat treatment of
*Corresponding author, email [email protected]
rails that are in vogue include direct hardening through

ß 2010 Institute of Materials, Minerals and Mining

Published by Maney on behalf of the Institute
Received 30 November 2008; accepted 11 May 2009
DOI 10.1179/174329409X455458 Surface Engineering 2010 VOL 26 NO 3 199
Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

water quenching from rolling mill, hardening by a high of used T-12 rail sections. Laser treatment on the surface
speed stream of water, head hardening with differential of the T-12 rail steel discs was carried out using a 9 kW
hardness gradient, etc. Some of these methods incorpo- CW CO2 laser (Model: ML-108, MLI Lasers Ltd,
rate a cycle that controls the cooling rate and aims at Israel). Before laser treatment, all discs were surface
reducing pearlitic lamellar spacing, thereby resulting in ground and coated with a 10–15 mm thick graphite layer
increased hardness leading to improvement in rail life.18–23 to enhance the coupling efficiency of the laser beam with
However, these methods have certain shortcomings, the surface. A rectangular laser beam spot of size
namely, requirement of post-treatment operations such 1961 mm, tailored by using facetted and elliptical
as straightening and machining due to massive distor- copper mirrors, was used under an argon gas shield
tion that results during processing. In contrast, laser with a constant laser power of 3500 W. It is pertinent to
surface treatment has the ability to harden locally and mention that using a larger laser beam spot is
precisely, resulting in negligible distortion. Additionally, advantageous for facilitating wide area coverage in a
with the advent of new and compact laser sources, single pass, instead of multiple passes with overlapping,
hardening by mounting the laser on a mobile carrier is which can be detrimental either because of development
also now realistically possible. In laser surface hardening of soft zones in the overlapped region due to tempering
of steel components, a laser beam with a desired spot effect or incorporation of cracks in the treated layer due
shape/size is scanned on the surface of the material to to uneven stress concentrations. The disc, fixed to a
raise the temperature above austenetisation range and rotary axis, was rotated at speeds ranging from 3?33 to
subsequent self-quenching results in hard martensitic 13?33 rev min21 to vary the laser beam scan speed on
structure due to high cooling rates. It has the added the disc surface so that the effect of laser interaction time
advantages of high case hardness with better uniformity, on the resulting treated layer could be studied across
absence of quenching media and negligible distortion conditions leading to both transformation hardening as
due to controlled localised treatment. Rail surfaces can well as melting of a thin surface layer. Preliminary laser
also be melted/glazed by increasing the laser intensity to treatment trials were first conducted to identify a
create a shallow surface melt which, upon solidification, parametric window in terms of laser power and rotary
can result in glassy/nanocrystalline phases due to high axis speed over which a cross-over from laser hardening
cooling rates that prevail in the process. Such laser to laser surface melting, with good treated layer
melted/glazed surfaces, having low friction coefficient, characteristics, was achievable. At a laser power of
can potentially serve as a solid state lubricant, and 3500 W, the layers treated at disc speeds below
mitigate subsurface cracking while also being compliant 3?33 rev min21 showed excessive melting with undula-
with contacting wheel profiles.13,14 Laser hardening and tions and cracking. The layer treated at speeds above
laser melting/glazing technologies have also been proven 13?33 rev min21 exhibited low case depth with non-
in various applications in automotive and other uniformity in hardness distribution.
industrial sectors.24–26 Furthermore, with the recent Both untreated and laser treated rail steel discs were
commercial availability of high power energy efficient extensively characterised for microstructure, hardness
lasers, cost effective surface treatment of components and phase constitution. Optical microscopy with image
like rails is also possible. However, implementation of analysis was used for case depth evaluation and
these laser based surface modification technologies microstructure observation. Identification of various
requires extensive research for optimising the process phases in the untreated and laser treated layers of rail
to achieve defect free layers. steel was carried out by X-ray diffraction (XRD)
The present study is aimed at investigating and techniques. In case of laser processed samples having a
understanding the efficacy of adopting laser surface melted layer on the surface, XRD analysis of the
treatment, by hardening and/or melting, for improving underlying hardened zone was also carried out by
the sliding wear performance of rail steel under normal grinding away the melted layer carefully. In addition,
loads. Laser hardening and laser melting treatment of a determination of the amount of retained austenite and
typical rail steel was carried out with a 9 kW CW CO2 estimation of grain size from XRD patterns according to
laser. Effect of processing variables on case depth, Scherrer formula were made using MATLAB software
hardness and microstructure of the treated layers has integrated with the system.
been discussed. Results of sliding wear tests conducted Sliding wear tests were carried out using a pin-on-disc
with a pin-on-disc device on laser treated and untreated device under normal loads in unlubricated dry condi-
rail steel are compared and assessed for improvement in tions to assess the tribological behaviour of untreated
sliding wear performance. discs as well as laser treated discs with both hardened/
melted surface layers. Pins of 20 mm length were
fabricated from the unaffected portions of used R-19
Experimental procedure steel wheel of Indian Railways Standard (composition
T-12 rail steel of Indian Railways Standard (composi- provided in Table 1). Similar pins, fabricated from
tion provided in Table 1) was used as the substrate standard tungsten carbide (WC) and high speed steel
material. Eighty millimetres diameter discs of 8 mm (HSS) materials, were also used in pin-on-disc tests
thickness were fabricated from the unaffected portions carried out to as part of this study. The sliding contact

Table 1 Chemical composition of rail and wheel steels utilised in this study

C Mn Si Cr Ni Mo Cu V Al S P

T-12 rail steel 0.67 1.13 0.27 Nil Nil Nil Nil Nil ,0.01 ,0.02 ,0.02
R-19 wheel steel 0.51 0.74 0.36 0.23 0.19 0.05 0.17 ,0.02 Nil ,0.02 ,0.02

200 Surface Engineering 2010 VOL 26 NO 3

 Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

2 Effect of rotary axis speed in revolutions per minute

during laser scanning on case depth of treated layer

evaluated by a perthometer. Mass loss in discs was

measured by using a weighing scale with 0?1 mg
resolution after every 0?5 km of sliding distance, while
the pin weight loss was determined only at the end of the
wear test corresponding to a sliding distance of 6 km. The
sliding coefficient of friction was also determined using a
friction monitor attachment available with the pin-on-
disc device. Morphologies of worn surfaces of untreated
and laser treated surfaces of rail steel disc were examined
a T-12 rail steel; b R-19 wheel steel under a scanning electron microscope (SEM) in an
1 High magnification SEM images of rail steel and wheel attempt to understand the possible wear mechanisms.
steel materials used in study
Results and discussion
pins had an apparent contact area of 2?8361025 m2,
Figure 1 shows the typical microstructures of T-12 rail
and the hardness values of the pins were 295–340 HV
steel and R-19 wheel steel used in the study. It is evident
for R-19 wheel steel, 810–850 HV for HSS and 1400–
that both the rail steel and wheel steel are almost fully
1550 HV for WC. Three different normal prototypic
pearlitic with variation in ferrite (dark etched regions)
loads (25, 50 and 75 N) were applied with a constant
content, determined by image analysis to be about 2% in
sliding speed of 1 m s21 for all the tests. Conditions for
rail steel and 15–18% in wheel steel. As a consequence of
sliding wear testing are listed in Table 2 and these are
the above variation in ferrite content, the hardness
similar to those adopted in the study conducted by
measured was in the range of 340–375 HV in rail steel
Viafara and co-workers.7 Before performing sliding and 295–340 HV in wheel steel, with the relatively high
wear tests, both untreated and laser treated rail discs ferrite content being responsible for lower hardness in
and pins were surface ground to eliminate role of surface wheel steel. Microstructures of these steels also indicate
asperities. The surface roughness of the pin and disc the pearlitic lamellar spacing to be in the range of 200–
after grinding was y0?5 mm Ra. The wear depths were 500 nm.
The effect of laser beam scanning speed, resulting
Table 2 Pin-on-disc sliding wear test conditions from varying disk rotation speed of the rotary axis
holding the discs during laser treatment, on case depth is
Sliding wear test conditions
illustrated in Fig. 2. The cross-sectional micrographs of
Disc material 80 mm diameter disc the treated layers processed at different disk rotation
of T-12 rail steel (ground) speeds, from which the case depths were determined, are
Pin material R19 wheel steel, HSS shown in Fig. 3. Low magnification microstructures of
and WC the treated layers shown in Fig. 3a and b clearly indicate
Sliding velocity, m s21 1 defect free hardened and melted layers respectively, with
Normal load, N 25, 50 and 75
Sliding distance of 0.5 sound interfaces between hardened layer/substrate and
each test, km melted/hardened layers. The case depth varied signifi-
Total tests performed 12 cantly with scan speed. The total case depth, comprising
Disk rotation speed, 541 melted and hardened regions, decreased from 940 to
rev min21 320 mm with increase in disk rotation speed from 3?33 to
Ambient atmosphere 25–32uC with 60–70%RH 13?33. The layer processed at 3?33 rev min21 showed a
Condition Dry
melted layer depth of about 55 mm as against no melting

Surface Engineering 2010 VOL 26 NO 3 201

Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

a optical micrograph of layer processed at 13?33 rev min21; b optical micrograph of layer processed at 3?33 rev min21; c
SEM image of melted region of layer processed at 3?33 rev min21; d SEM image of hardened region of layer processed
at 3?33 rev min21
3 Cross-sectional microstructures of laser treated rail steel layers processed under different conditions

in the layer processed at 13?33 rev min21. The above pearlite in the unaffected rail steel substrate. The
observations are a direct consequence of the longer laser additional peak noted at 51u Bragg angle is suggestive
beam interaction time with the rail disc surface at lower of substantial presence of austenite in the melted layer.
disk rotation speed. Hence, by controlling process The XRD pattern of the hardened layer exhibited a
parameters like scan speed, laser power, beam spot size, predominantly martensitic phase with few carbide
etc., the case depth of the treated layer can clearly be phases. A small peak corresponding to austenite, which
controlled.27–29 High magnification micrographs shown is absent in the unaffected steel substrate, can also be
in Fig. 3c and d indicate distinguishable microstructural observed in the XRD pattern of the hardened region.
changes in the hardened and melted regions respectively, The retained austenite content was evaluated by
of the layer processed with 3?33 rev min21. The layer deconvoluting peaks of XRD patterns using MATLAB
processed at 3?33 rev min21 indicated substantial melt- software integrated with the system. The amount of
ing. Figure 3c illustrates the presence of highly deformed retained austenite was found to increase from 2?1% in
martensitic needles with featureless austenite and the hardened layer to 9?8% in the melted layer.
cementite. The absence of a dendritic pattern in the Broadening of XRD peaks with reduction in peak
melted layer may be particularly noted, suggesting a intensity, as compared to the sharp peaks observed in
refined microstructure due to high solidification rates diffraction patterns of untreated substrate, was observed
typically experienced in the melt pool during laser at Bragg angles of 45, 65 and 82u in case of the melted
melting.13,14 Figure 3d shows the microstructure of the layer. This could plausibly be due to the presence of
hardened region below the melted region comprising of nanocrystalline phases and stress, arising from rapid
uniformly distributed refined martensite throughout its solidification of the melt due to high cooling rates
depth with few pockets of retained austenite. experienced in laser processing. In this context, it is
Figure 4 depicts XRD patterns from analysis of relevant to point out that the grain size of austenite in
various regions of the treated layer processed at the melted layer, evaluated according to Scherrer
3?33 rev min21 and unaffected substrate. The patterns formula from XRD patterns and calculated using
clearly indicate the presence of martensite and carbide MATLAB software integrated with the system, was
phases in the hardened and melted layers as compared to found to be in the range 28–44 nm. Prior reports by Cui

202 Surface Engineering 2010 VOL 26 NO 3

 Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

4 X-ray diffraction patterns of hardened and melted regions of laser treated layer in T-12 rail steel processed at
3?33 rev min21 and unaffected substrate

and co-authors involving laser processing of steels have rapid cooling of the melt, a glazed layer may be expected
also indicated formation of such nanocrystalline phases, to form, depending on the rate of cooling. The melted
depending on the process parameters adopted.24,25 layer depth increased with further decrease in disk
Scanning speeds below 3?33 rev min21 were found to rotation speed, from about 17 mm in case of layer
result in coarser austenite grains in a dendritic micro- processed at 9?99 rev min21 to 55 mm in case of layer
structure, with excessive melting and microcracking. In processed at 3?33 rev min21. The cross-sectional hard-
view of this, the lowest laser scanning was restricted to ness profiles in Fig. 5 show hardness in the range of 830–
3?33 rev min21 to retain the nanocrystalline nature 900 HV in the melted region and 890–1070 HV in the
without any surface undulations and cracks. In case of hardened region. The high hardness in hardened regions
the hardened layer, broadening of the diffraction peaks is attributed to martensite formation, as evident from
is less notable as compared to that in case of the melted microstructural evaluation and phase analysis explained
region, with the observed peak broadening being earlier. Relatively low hardness in the melted zone as
attributable to presence of refined martensite. compared to the hardened layer is evident from the
Hardness profiles across the depths of laser treated expanded hardness profile shown in the inset in Fig 5.
rail steel layers, processed with different laser scanning The reduced hardness in the melted region (830–
speeds, are shown in Fig. 5. The hardness variation with 900 HV) as compared to that in the hardened region is
depth in treated layer results due to thermal fields due to increased amount of retained austenite and
affecting microstructure as depicted by various reduced martensite as evident from the phase constitu-
authors.28,29 If the laser interaction time is such that tion revealed from XRD patterns. In spite of high
the material at a given location attains a temperature retained austenite in the melted layer, the hardness did
above the solidus, melting occurs as observed in case of not decrease to the extent generally observed in
the layer processed at 9?99 rev min21 or lower. Upon conventional processes, consistent with reports in

Surface Engineering 2010 VOL 26 NO 3 203

Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

5 Hardness profiles across depths of laser treated layers of T-12 rail steel processed at different rotary axis speeds
(inset shows enlarged view of hardness variation within top 100 mm case depth)

various other studies involving laser melting of considered favourable to melt only a thin layer (pre-
steels.27,30–32 The striking uniformity in the hardness ferably less than 100 mm) which can rapidly solidify.
distribution in the hardened region of the processed Further TEM analysis is in progress to investigate the
layers is due to prior uniform pearlitic distribution in the microstructural features within the melted layers
rail steel substrate. Similar hardness in the melted region obtained in the present study.
and uniform martensite formation in the hardened Initially, sliding wear tests were performed with
region has been reported in the study performed by untreated T-12 rail steel as disc and R-19 wheel steel
Demelfi and co-workers13 on laser treatment of pearlitic as pin under three different loads of 25, 50 and 75 N.
rail steel, as used in the present study. However, Subsequently, the untreated and laser hardened/melted
marginal changes in hardness values could be observed discs were also tested against WC and HSS pins. In each
across the depth of the treated layers, depending on the case, the discs and contact pins were ground to 0?5 mm
disk rotation speed, due to local morphological changes Ra surface roughness before sliding wear testing. The
in martensite and other phases. With further substantial mass loss of disc after every 0?5 km sliding distance was
reduction in the disk rotation speed, microstructural measured and the wear rate was evaluated in terms of
changes may lead to reduced martensite, accompanied by mass loss per unit sliding distance. The steady state wear
carbon loss and increased austenite, thereby resulting in regime was found to occur only after a certain initial
reduced hardness in the melted layer. Prior studies period of sliding, corresponding to the running-in or
reported by various groups have also shown considerable accommodation period, during which the noted mass
microstructural coarsening, with loss of nanocrystalline/ loss was high and the wear rate fluctuating. The high
amorphous phases in the melt pool, with increased laser wear rate during initial accommodation period could be
material interaction time.24,25,33 Hence, for obtaining a attributed to the changing surface conditions and wear
glazed layer of nanocrystalline/amorphous nature, it is process with deformation and removal of asperities.

204 Surface Engineering 2010 VOL 26 NO 3

 Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

6 Sliding wear performance of laser treated T-12 rail steel disc tested under 50 N load with different contact pins (inset
depicts sliding wear behaviour of untreated rail disc)

Surface profile analysis of worn surfaces performed intermittently noted during the steady state wear period.
intermittently during the accommodation period, using It is evident that, in case of HSS and WC sliding pins,
a surface profilometer, indicated smoothening of rail the mass loss in the T-12 rail steel discs during the initial
disc owing to removal of asperities accompanied by an accommodation period is higher compared to the case
increased contact area. involving the R-19 wheel steel pin as well as the
In case of untreated T-12 rail discs tested against R-19 corresponding steady state disc mass loss. For example,
wheel pins, the steady state wear rate in both the disc in case of WC pin sliding, the untreated T-12 disc mass
and the pin was found to increase with increasing load loss for initial 0?5 km sliding distance is as high as
(25 to 75 N). In case of rail disc, the steady state wear 130 mg as against 1?0 mg during the last 0?5 km lap at
rate increased from 3?661023 g m21 at 25 N load to the end of the 6?0 km sliding distance constituting the
8?461023 g m21 under 75 N load. In case of wheel pin, entire test. This could be attributed to the combined
the wear rate increased from 8?761023 g m21 at 25 N effect of initial accommodation period and surface
load to 15?461023 g m21 under 75 N load. An increase asperities, as well as the large difference in hardness
in load has a direct influence on the contact pressure, between the untreated disc and the mating pin. In a
apparent area of contact and stress and, thereby, previous study, conducted by Viafra et al., involving
influences the wear process significantly. Irrespective of sliding of pearlitic and bainitic steels, increase in
the load applied, the wear rate was always found to be hardness variation between disc and contact pin
relatively higher in the wheel steel pin compared to the materials was shown to promote roughness changes
rail steel disc. This can be attributed to the lower and reduced adhesion.7 In case of the untreated rail steel
hardness of wheel steel owing to the higher ferrite disc sliding against the R-19 wheel steel pin, the mass
content as observed in microstructures previously removal is relatively gradual as the difference between
discussed. Few pockets of iron oxides were observed the hardness values of the two materials is only about
on worn surfaces of discs after attaining steady state 40 HV, suggesting predominantly adhesive wear.
wear rate, i.e. after the initial accommodation period. The sliding wear behaviour of hardened and melted
The role of mild oxidation, resulting from frictional surface layers of laser treated rail steel discs against WC
heating, on the wear process has also been reported in and HSS pins is also illustrated in Fig. 6. It is evident
other studies involving dry sliding wear of steels.5–9 that the wear rate reduced greatly in laser processed
The sliding wear performance of the untreated rail discs as compared to the untreated counterpart, regard-
steel disc against contact pins made of wheel steel, HSS less of the laser processing conditions and irrespective of
and WC under identical testing conditions (normal load whether the laser treatment involved surface hardening
of 50 N) is shown as an inset in the top left corner of alone or surface melting with an underlying hardened
Fig. 6. The coefficient of friction values depicted on each layer. As compared to the heavy mass loss in initial laps
of the curves in the figure are the mean average of values during sliding with untreated discs, mass losses were

Surface Engineering 2010 VOL 26 NO 3 205

Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

7 Friction coefficient as a function of sliding distance for all rail disc surfaces involving HSS and WC pins

found to be relatively lower in case of treated discs. On lead to a marginal reduction in friction coefficient as one
the whole, too, the total mass loss after 6 km sliding would have expected, the reduction is not large enough
distance was about 193 and 106 mg in untreated rail disc to have any significant bearing on the weight loss data.
when tested against WC and HSS pins respectively, as A similar observation was made in case of tests carried
against 48 and 20 mg in laser surface melted discs, and out with HSS pins, with the coefficient of friction values
39 and 15 mg in laser hardened discs. The above results being of the same order and varying only over a small
clearly imply a reduction in wear loss by a factor of 4 to range. Thus, for the six different mating surface
5 and a vast improvement in the sliding wear resistance combinations depicted in Fig. 7, the average coefficient
due to laser treatment, which can be attributed to the of friction values vary only over the narrow band of
significantly increased hardness in the treated layers. 0?36–0?46. Thus, the above noted changes in coefficient
It is to be noted from Fig. 6 that, the wear rate is of friction are not expected to either dictate the overall
observed to gradually reduce and become nearly wear process or explain the above noted surprising trend
constant in all cases by about 2?5–3?0 km of sliding of the disc weight loss being higher when tested against
distance, as evidenced by the nearly constant slope of the HSS pins as compared to WC pins. For the sake of
curve depicted for each individual case. However, the completeness, it is also pertinent to point out that Fig. 6
steady state wear rate corresponding to the constant depicts a slightly higher weight loss in case of laser
slope region of the cumulative weight loss vs. sliding melted disc specimens despite the marginally lower
distance curves in Fig 6, is higher in laser hardened or friction coefficient, during sliding with HSS as well as
melted discs slided with HSS pin as compared to WC pin WC pins. This observation is by no means unique and
sliding. A similar trend is also observed in wear tests has also been reported previously. For example, DiMelfi
conducted with untreated rail discs. Although this et al. and Aldajah et al. have reported reduction in
observation is surprising because of the higher hardness friction in melted layers of rail steels and attributed it to
of WC pin as compared to HSS, even if one accounts for elastic compliance of the melted region.13,14
the possibility of work hardening in tool steels due to Scanning electron morphologies of worn surfaces on
high frictional heating as reported in other studies,30,34 untreated and laser processed rail steel discs, after
the above trend was consistently observed and demands completion of sliding wear testing with HSS and WC
further investigation. This was attempted through care- pins, are illustrated in Fig. 8 and provide further insight
ful monitoring of friction coefficient and a study of worn into the wear behaviour associated with different
surface microstructures as discussed subsequently. processing and testing conditions. The worn surfaces
Figure 7 illustrate variation of friction coefficient as a of untreated rail disc after sliding against WC and HSS
function of sliding distance for all mating surface pins can be seen in Fig. 8a and d respectively. If carefully
combinations involving untreated as well as laser treated observed, the only common feature in the above two
rail discs and HSS and WC pins. The friction coefficient micrographs is the evidence of scoring marks parallel to
values were found to be 0?36 and 0?42 respectively, for the direction of sliding and the presence of some
melted and hardened discs compared to 0?44 for particles, indicative of ineffective removal of debris.
untreated substrate for tests conducted with WC pin. However, the specimen tested against WC reveals severe
While these values do indicate that laser melted surfaces abrasion characterised by deep grooves and, over some

206 Surface Engineering 2010 VOL 26 NO 3

 Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

a untreated surface with WC-pin sliding; b hardened surface with WC-pin sliding; c melted surface with WC-pin sliding;
d untreated surface with HSS-pin sliding; e hardened surface with HSS-pin sliding; f melted surface with HSS-pin sliding
8 Worn surface morphologies of rail steel discs tested under different conditions

regions of the wear track, morphology suggestive of evident from the pits in the worn surfaces of untreated
plastic deformation. The presence of deeper grooves in discs. Such contribution of mild oxidation to wear in
case of WC pin sliding as compared to HSS sliding steels during pin-on-disc sliding testing has been
indicate severe wear in untreated rail due to antagonistic previously reported by various authors studying sys-
hard WC material. This suggests that sliding of soft tems involving significant variation in hardness of
pearlitic rail disc against the considerably harder pin contacting surfaces.35–37
results in predominantly abrasive wear. The disc tested Examination of the worn surface morphologies of
against a HSS pin, on the other hand, does not exhibit laser treated tested disc specimens is equally informative
deep grooves but shows evidence of smearing on the and can also be explained based on the same considera-
surface. tions as discussed above. An immediate observation is
Scoring and deep grooves parallel to the direction of that two distinct types of worn surface morphologies
sliding could be due to abrasion by trapped debris result based on the type of pin used. As evident from
formed during early stages of sliding constituting the Fig. 8, the morphologies of laser hardened and melted
initial accommodation or running-in period. Material surfaces tested with a WC pin, depicted in Fig. 8b and c,
removal mechanisms in sliding include asperity defor- are largely similar. The morphologies of laser hardened
mation, fracture, delamination, adhesion, abrasion, etc. and melted surfaces tested with a HSS pin shown in
and observation of scoring and cutting grooves arising Fig. 8e and f are also similar, although quite distinct
from resulting debris entrapped between the sliding from those resulting from testing with a WC pin. This
surfaces has been reported in various studies involving can be easily explained by the fact that, since the laser
laser treated steels.34–37 Ineffective removal of debris hardened and laser melted layers exhibit comparable
during the wear test and its prolonged entrapment hardness values, the hardness difference between the
between the sliding surfaces can lead to smearing and/or mating surfaces is in the narrow range of 540–620 HV
plastic deformation. Prior studies investigating wear due for tests corresponding to Fig. 8b and c, while the
to sliding and involving hard metallic materials such as hardness difference is less than 100 HV for tests
WC, Al2O3, Ti3AlC2 and Ti–Co also reported entrap- corresponding to Fig. 8e and f. Smearing is predomi-
ment of debris or oxides and their subsequent smearing nantly noted in Fig. 8b and c, and as mentioned above,
on contact surfaces.38–40 It may be noted that the it is envisaged that considerable debris entrapment
difference between the hardness of WC and untreated between the sliding surfaces over the entire contact area
T-12 material is more than 1000 HV while that between is responsible for this observation. This also manifests in
HSS and the untreated T-12 material is around 450– the form of apparently lower disc weight loss in the laser
500 HV. This results in extensive generation of debris treated discs despite the much harder WC pin being
during initial stages of wear and it is plausible that a employed. In contrast, when the HSS pin is used against
good amount of the debris remains entrapped between the laser treated discs of comparable hardness, the
the sliding surfaces and gets smeared and reconsoli- material removal is far more gradual and accumulation
dated with continued sliding, more so in case of the WC and/or entrapment of debris are unlikely. Scoring marks
pin. Following the initial accommodation period, it is are indeed seen as a consequence of three body abrasion,
also conceivable that mild oxidation caused by the which is quite typical of sliding wear. Another observa-
inevitable temperature rise under dry sliding conditions tion that corroborates the earlier discussions is the
also plays a role in determining the wear behaviour, as presence of relatively smoother surfaces noted in Fig. 8c

Surface Engineering 2010 VOL 26 NO 3 207

Shariff et al. Sliding wear behaviour of laser surface modified pearlitic rail steel

and f as compared to Fig. 8b and e respectively, which 7. C. C. Viafara, M. I. Castro, J. M. Velez and A. Toro: Wear, 2005,
259, 405–411.
explain the marginally lower coefficient of friction values
8. J. Kalousek: Wear, 2005, 258, 1330–1335.
obtained in tests involving laser melted discs. 9. Y. Jiraskova, J. Svoboda, O. Schneeweiss, W. Daves and F. D.
Fischer: Appl. Surf. Sci., 2005, 239, 132–141.
Conclusions 10. G. Donzella, M. Faccoli, A. Ghidini, A. Mazzu and R. Roberti:
Eng. Fra. Mech., 2005, 72, 287–308.
The study demonstrated the feasibility of improving 11. P. Clayton: Wear, 1996, 191, 170–183.
sliding wear resistance in rail steel under prototypic 12. H. Muster, H. Schmedders, K. Wick and H. Pradier: Wear, 1996,
191, 54–64.
laboratory load conditions by laser hardening and laser
13. R. J. DiMelfi, P. G. Sanders, B. hunter, J. A. Eastman, K. J.
melting. The hardened/melted layers produced by rapid Sawley, K. H. Leong and J. M. Kramer: Surf. Coat. Technol., 1998,
cooling were harder than untreated pearlitic substrate 106, 30–43.
with sound interfacial bonding. The type of layer 14. S. Aldajah, O. Ajayi, G. R. Fenske and S. Kumar: J. Tribol., 2003,
(hardened or melted) and its depth was found to vary 125, 643–648.
15. J. W. Ringsberg, A. Skyttebol and B. L. Josefson: Int. J. Fatigue,
depending on the laser processing parameters. 2005, 27, 702–714.
Microhardness measurements revealed hardness values 16. M. Hiensch, P. O. Larsson, O. Nilsson, D. Levy, A. Kapoor, F.
of 830–900 HV in the melted layer, as against 890– Franklin, J. Nielsen, J. W. Ringsberg and B. L. Josefson: Wear,
1070 HV in the hardened layer. The wear behaviour was 2005, 258, 964–972.
found to vary depending on the laser treated layer 17. F. J. Franklin, G.-J. Weeda, A. Kapoor and E. J. M. Hiensch:
Wear, 2005, 258, 1048–1054.
characteristics and type of sliding contact pin. In terms 18. R. I. Carroll and J. H. Beynon: Wear, 2006, 260, 523–527.
of mass loss, the improvement in sliding wear resistance 19. K. M. Lee and A. A. Polycarpou: Wear, 2005, 259, 391–399.
was more than twofold in laser hardened/melted layer as 20. M. Ueda, K. Uchino and A. Kobayishi: Wear, 2002, 253, 107–113.
compared to untreated pearlitic substrate. The coeffi- 21. R. Heyder and G. Girsch: Wear, 2005, 258, 1014–1021.
22. M. G. M. F. Gomes, L. H. Almeida, L. C. F. C. Gomes and I. L.
cient of friction was found to have marginally reduced May: Mater. Char., 1997, 39, 1–14.
with laser melted layers as compared to hardened 23. P. Clayton and D. Danks: Wear, 1990, 135, 369–389.
counterparts. The results indicate possibility of applying 24. C. Cui, I. Hu, Y. Liu, K. Gao and Z. Guo: Appl. Surf. Sci., 2008,
defect free laser treatment on pearlitic rails, particularly 254, 6779–6782.
at curved portions, for improved sliding wear resistance. 25. C. Y. Cui, J. D. Hu, Y. H. Liu and Z. X. Guo: Appl. Surf. Sci.,
2008, 254, 3442–3448.
26. W. M. Steen: ‘Laser materials processing’; 1991, London, Spriger.
Acknowledgements 27. A. Basu, J. Chakraborty, S. M. Shariff, G. Padmanabham, S. V.
Joshi, G. Sundararajan, J. D. Majumdar and I. Manna: Scr.
The authors gratefully acknowledge Director, ARCI for Mater., 2007, 56, 887–890.
granting the permission to publish the work. They are 28. K.-A. Chiang and Y.-C. Chen: Mater. Lett., 2005, 59, 1919–1923.
also thankful to Dr G. V. N. Rao for phase analysis and 29. M. A. Ashby and E. Easterling: Acta Metall., 1984, 32, 1935–1962.
30. R. Colaco and R. Vilar: Wear, 2005, 258, 225–231.
Mr Shyam Rao, Technician for metallographic sample 31. J. H. Abboud, K. Y. Benyounis, A. G. Olabi and M. S. J. Hashmi:
preparation. J. Mater. Process. Technol., 2007, 182, 427–431.
32. M. R. James, D. S. Gnanamuthu and R. J. Moores: Scr. Metall.,
References 1984, 18, 357–361.
33. J. kausinski: Appl. Surf. Sci., 1995, 86, 317–322.
1. F. C. R. Hernandez, N. G. Demas, D. D. Davis, A. A. Polycarpou 34. Y. Wang, T. Lei and J. Liu: Wear, 1999, 231, 12–19.
and L. Maal: Wear, 2007, 263, 766–772. 35. M. S. Devangan and P. A. Molian: J. Mater. Process. Technol.,
2. A. Kapoor, F. J. Franklin, S. K. Wong and M. Ishida: Wear, 2002, 1990, 23, 41–54.
253, 257–264. 36. H.-J. Kim and Y.-G. Kweon: Wear, 1996, 193, 8–15.
3. L. Deters and M. Proksch: Wear, 2005, 258, 981–991. 37. D. I. pantelis, E. Bouyiouri, N. Kouloumbi, P. Vassiliou and A.
4. F. Alwahdi, F. J. Franklin and A. Kapoor: Wear, 2005, 258, 1031– Koutsomichalis: Surf. Coat. Technol., 2002, 298, 125–134.
1037. 38. S. Ndlovu, K. Durst and M. Goken: Wear, 2007, 263, 1602–1609.
5. W. R. Tyfour, J. H. Beynon and A. Kapoor: Wear, 1995, 180, 79– 39. L. Wu, J. Chen, M. Liu, Y. Bao and Y. Zhou: Wear, 2009, 266,
89. 158–166.
6. U. Olofsson and T. Telliskivi: Wear, 2003, 254, 80–93. 40. Y. Xue and H. M. Wang: Mater. Lett., 2008, 62, 3091–3094.

208 Surface Engineering 2010 VOL 26 NO 3

Copyright of Surface Engineering is the property of Maney Publishing and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.