Multiscale and Multidisciplinary Modeling, Experiments and Design

https://doi.org/10.1007/s41939-024-00570-9

ORIGINAL PAPER

The evaluation of wear processes of the rolling-sliding contact

by means of flake wear debris in dry and wet contact
Antoni John1 · Henryk Bakowski
˛ 2 · Bogusław Śleziak3

Received: 23 February 2024 / Accepted: 9 August 2024

© The Author(s) 2024

Abstract
One of the tasks of this paper is to explain the cause-and-effect relationship of the influence of operating parameters on the
wear of the wheel-rail system where fatigue processes are initiated. In this way, learning about and defining the nature of the
phenomenon, makes it possible to determine the reliability of component operation. The aim of this study was to determine the
influence of operational parameters on tribological properties of the rolling-rail contact representing the wheel-rail association
in real operating conditions. Determining the form of wear on the basis of laboratory tests carried out on an Amsler-type
wear testing machine in the wheel-roller system makes it possible to determine the nature of changes occurring as a result of
cyclic interaction of normal and tangential forces on the surface of the tested material, without the need for long-term and
costly observation of the tested object in normal operation. Quantitative metallographic analysis of the morphology of wear
debris was used to determine the intensity of surface wear. Determination of the influence of operational parameters on the
stereological characteristics of wear debris makes it possible to predict the regenerative grinding of rails in service in the track
and allows practical application by the sections of operation and diagnosis of the railway system.

Keywords Rolling contact fatigue · Wear debris · Wet and dry contact

1 Introduction the wear of the rolling-slip association and the RCF (rolling
contact fatigue) condition, which do not allow for a compre-
Wear in the wheel-rail system is an extremely complex and hensive analysis. The influence of the type of rolling stock,
difficult process to analyse. The constant change and over- the different value of sliding taking into account the track
lapping of different factors in a very short time make wear profile (most common) or the different speed depending on
difficult to interpret. Thanks to the laboratory tests carried out the restrictions introduced or the intended use were not taken
on the Amsler test stand in the wheel-rail system in a wheel- into account. The paper (Gallardo-Hernandez and Lewis
rail contact, it was possible to determine the mechanism and 2008) describes studies to determine the values of normal
intensity of wear depending on the rolling material, the road and tangential stresses and the results of tribological tests on
profile and the constraints existing on the railway track, i.e. a railway wheel. The contact stress values (Table 4) are con-
load, sliding and speed, as well as atmospheric conditions. sistent with the results reported in the paper (645–882 MPa)
Current studies (Zakharov and Zharov 2002; Donzella (Clayton 1995).
et al. 2011a) have been carried out in a partial way consid- In papers (Zhong et al. 2011; Hiroyuki et al. 2011), labo-
ering the influence of slips, selected pressures or speeds on ratory tests in dry contact with a slip of 2.5–25% or 0.06% at
an assumed speed value of 500 min–1 were described. Tests
B Antoni John in the presence of a lubricating medium were described in
antoni.john@polsl.pl papers, in which the slip was correctly mapped to operating
conditions (0.5–5%). In contrast, the authors did not carry
1 Faculty of Mechanical Engineering, Silesian University of
out tribological tests taking into account the effect of speed,
Technology, Konarskiego 18A, 44-100 Gliwice, Poland
as they proposed a constant value of 400 min–1.
2 Faculty of Transport and Aviation Engineering, Silesian
More extreme conditions were proposed by the authors
University of Technology, Krasińskiego 8, 40-019 Katowice,
Poland of the paper (Clayton 1995; Dayot and A. Saulot n, C.
3 Godeau, Y. Berthier 2012) in which the speed was between
Court Expert, Katowice, Poland

123
 Multiscale and Multidisciplinary Modeling, Experiments and Design

Table 1 Overview of precipitation values in Europe (GUS 2022) has an extremely significant impact on the phenomena and
durability of the wheel-rail contact.
Europe
One of the main tribological properties is the coefficient
Areas Coasts Lowlands of friction. In a wheel-rail contact, these values range from
mountainous 0.3 to 0.6, depending on the values of the operational factors
areas
(Garnham and Davis 2011).
Precipitation 1000–2000 3000–4000 400–600
The final factor affecting the wear of the wheel-rail system
value is the phenomenon of fatigue crack propagation as a result
precipitation, of the spreading action of the fluid (Fig. 1). This process is
mm closely related to the so-called fluid closure effect in the gap,
which occurs when the wheel is rolled through the contact
area (Halamaa 2011; Kvarda et al. 2021). Fluid closure results
in a short-circuiting effect of the crack edges, forced by the
160–300 km/h with a load of 16,000–25,000 N and a wheel contact load, with the crack walls inside the crack opening up
diameter of 0.75–1.25 m. A lower velocity results in a greater (pitting). During the development of fatigue cracks, curing by
depth of plastic deformation than is the case at higher veloc- pitting and embrittlement following dispersion curing after
ities. Analysing the effect of the loads, it was found that pitting occurring in the material immediately above the crack
fatigue cracks appeared at the lower value (16000 N/wheel), face follow in succession (Al-Juboori et al. 2019). This leads
while for the higher velocity, a tendency to develop cracks to the formation of cracks with critical lengths for the loads
of greater length was revealed, which were particularly dan- acting in the tribological system. The rate of development and
gerous because they tended to propagate cracks in depth. the critical crack length are a function of the material prop-
The presence of a lubricating medium in the wheel-rail erties and primarily depend on operational factors [Halamaa
contact during curve travel is undoubtedly an important oper- 2011, Zhou et al. 2023].
ational factor that alters the mechanism and intensity of wear. The effects of operational factors on the tribological node
A study on a real-world object was described, in which the at the rolling-sliding contact lead to wear and the formation
presence of a lubricant (oil spray on the wheels) led to twice of wear debris. This is all the more important as, in an era
as much wear (Mazzù et al. 2024). However, the presence of of increasing interest in methods of early damage detection
water on the rolling surface of the rail, which would signifi- and determination, in addition to strength tests and chemical
cantly alter the wear mechanism, was not taken into account. composition, quantitative metallographic evaluation of flake
According to the Central Statistical Office (GUS Poland wear debris makes it possible to find cause-and-effect rela-
statement), the presence of water in the rolling-rail contact in tionships between the stereological features of the particles
Poland is 130 days, i.e. from April to October there is rainfall, under study and operational parameters. Assessment of the
i.e. the annual precipitation is between 500–700 mm, which technical condition of an object (i.e. the wheel-rail system)
can be calculated as approximately 500–700 L per m2 (GUS on the basis of features makes it possible, consequently, to
2022). In most of Europe (Table 1), the annual rainfall is predict the service life of the friction node under considera-
300–800 mm (300–800 L per m2 ). tion. Skilful use of methods for studying the products of wear
The presence of water on the rolling surface of the rail- and tear during service life can have an impact on increasing
way rails during the winter season, which is formed from safety (Zhang et al. 2023).
melted snow, should also be noted. Thus, the number of days According to the dislocation theory of defects, stress con-
when water is deposited on the rail surface increases, which centrators are clusters of dislocations formed when obstacles

Fig. 1 Cracks on the rolling

surface caused by high-speed
train traffic: a surface cracks,
b grid of cracks, c pit after 19 mm
removal of flake
b)

23 mm 19 mm

a) c)

123
Multiscale and Multidisciplinary Modeling, Experiments and Design

crack

mikropory
microcracks subsurface
szczelina crack
podpowierzchniowa
clusters of dislocaons
tangenal stress
Fig. 2 Wear model according to Suh’s flake wear theory (Fleming and Suh 1997)

wheel
koło
loading
obciążenie

szyna
tarcie
rail
fricon
wheel
koło loading
obciążenie

rail
szyna

increase thewwzrost ciśnienia płynu

pressure
szczelinie
in gap
a) b)

Fig. 3 View of the worn surface a and scheme presented effect of the fluid closing in the gap b

are encountered. These clusters give rise to micropores, rail, any geometrical and structural notches (including dis-
which grow and merge leading to the formation of wear prod- continuities) that cause stress concentration (Freisinger et al.
ucts. The separation of the particles itself occurs under the 2024; Zhou et al. 2023; Wang et al. 2023).
action of tangential forces. The present study is aimed at determining the basic tribo-
According to the flake theory of wear, in the sliding fric- logical properties, i.e. the dependence of wear and friction
tion process, the maximum accumulation of dislocations coefficient and stereological features of wear products on
occurs at a certain distance from the surface. Large accumu- selected operating factors of the wheel-rail system (load, slid-
lations of dislocations result in the formation of micro-slots, ing, velocity, water (Freisinger et al. 2024)), which are the
which merge by growth. This leads to the formation of gaps main causes of fatigue and contact wear in the rolling-rail
located parallel to the friction surface. When the gaps reach contact. The development of diagnostic levels on the basis
a critical length, the material separates in the form of a flake of quantitative metallographic analysis will make it possi-
(Roylance et al. 2000; Hu et al. 2020). ble to determine the type of wear mechanism depending on
It is assumed that when the edge of the fracture closes the operating conditions. Constant monitoring of wear and
(Fig. 2), there is fluid between the fracture walls below the damage is essential in the normal operation of railway rails.
edge, which has entered the fracture from the external contact The aim of this study was to determine the influence
surface. When the crack is closed, there is a high fluid pres- of operational parameters on the tribological properties of
sure inside the crack, which acts on the crack walls causing a rolling-sliding contact representing a wheel-rail associa-
the crack to enlarge (Olofsson et al. 2019). The formation tion in real operating conditions. Determining the form of
and development of damage on the surface and just below wear on the basis of laboratory tests conducted on an Amsler
it of the components working in the rolling-sliding contact stand wear testing machine makes it possible to determine the
depends, on the one hand, on the load and stress level in nature of changes occurring as a result of cyclic interaction
the most stressed area of the rail and, on the other hand, on of normal and tangential forces on the surface of the tested
the properties of the material of which the rail is made, and material, without the need for long-term and costly observa-
above all on its strength in this area (Fig. 3). Stress levels are tion of the tested object in normal operation. Laboratory tests
influenced not only by external factors, but also by the inher- reflected the most significant operating parameters, occurring
ent stress state, the topography of the rolling surface of the in the real object, affecting the durability of the wheel-rail

123
 Multiscale and Multidisciplinary Modeling, Experiments and Design

Fig. 4 Laboratory stand—a,

kinematic pair b

a) b)

contact (Zhang et al. 2022). For this purpose, a dimensional operational conditions. The operating factors were chosen
analysis reproducing the actual conditions prevailing at this to reflect typical operating conditions existing on railway
critical point was used. A special feature of laboratory testing tracks. Three types of rail steel were used in the tests, i.e.
is that it is more possible to control the direct influence of a perlitic steel without and after heat treatment and bainitic
single selected factor than is the case in real conditions. steel in both dry and wet contact. Quantitative and qualita-
tive metallographic studies of wear products detaching from
the rolling surface were carried out. Profiographometric stud-
2 Test procedure ies of wear products and selected surfaces after cooperation
were also performed. Following the proposed studies, an in-
The laboratory tests reflected the most relevant operating depth analysis was carried out and final conclusions were
parameters, occurring in the real object, affecting the durabil- formulated. The final conclusions were divided according
ity of the wheel-rail contact. For this purpose, a dimensional to their nature, i.e. cognitive conclusions, utilitarian (useful)
analysis was used to represent the actual conditions prevail- conclusions and developmental (prospective) conclusions.
ing at this critical point. A special feature of laboratory testing Based on a literature review and our own research, it was
is that it is more possible to control the direct influence of determined that operational parameters significantly affect-
a single selected factor than is the case in real conditions. ing the durability of railway tracks are load, slip, and speed.
The core of tribological research is laboratory testing on an Table 2 compiles typical operational conditions found on a
Amsler-type machine in a roll-to-roll contact. During tribo- selected track section and their representation in the labora-
logical testing, mass loss was measured in a cyclic way (at tory setup.
specific time intervals). In this way, mass loss was recorded The testing conditions in the laboratory setup are similar to
without leading to a significant increase in temperature at the those occurring in real objects, relying on data drawn from
point of contact of the cooperating elements. a selected section of the railway track. Comparative tests
of rail steels were conducted on an Amsler laboratory stand
in accordance with the PN-H-04332:1982 recommendations.
2.1 Description of the research object Tribological tests were performed in a rolling-sliding contact
under dry friction conditions with three repetitions, and the
The bench tests were carried out on an Amsler-type wear
result (weight loss and friction coefficient) represented the
testing machine containing a tribological node operating in
mean value. Rollers with a diameter of 38 mm were made
the rolling-slipping contact, in a roll-to-roll arrangement. The
from the tested steels, and they interacted with the counter-
bench tests were carried out in the rolling-slip contact under
sample, also in the form of a roller. Despite the fact that in
dry friction conditions. Amsler’s bench (Fig. 4) is a roll-
operational conditions, railway tracks interact, for example,
to-roll friction junction device, where both the sample and
with B6 steel (with a pearlitic microstructure and ferrite pre-
the counter-sample are rollers. As a result of the rotational
cipitations along grain boundaries) used for railway wheel
movement of the rollers with a common contact area, the
rims, in the laboratory test, 100Cr6 steel was used as the
phenomenon of surface fatigue or abrasive/adhesive wear of
counter-sample. It has a microstructure composed of carbides
the material (depending on the slip value) can be observed.
in a martensitic matrix with high hardness – 62 HRC. This
The rollers were mounted on two rotating rollers.
steel was chosen due to its common use as a counter-sample
material in tribological studies, allowing for a comparison
2.2 Discussion of working conditions of obtained results with literature data and minimizing the
influence of counter-sample wear on the test sample.
On the above-mentioned stand, tests were carried out on cur-
rently manufactured and used three rail steels under various

123
Multiscale and Multidisciplinary Modeling, Experiments and Design

Table 2 Selected operational and geometric parameters of the chosen railway surface in experimental modes

Type of research Operational parameters Geometric parameters

Load Sliding, % Velocity Contact surface area, mm2 Width of contact, mm

Stationary P1 = 500 N γ1 = 0,3 n = 100 min.−1 1,5 0,15

P2 = 1000 N γ2 = 2,6 n = 200 min.−1 2,0 0,2
P3 = 2000 N γ3 = 5,0 n = 300 min.−1
3,0 0,3
In real-world conditions Q1 = 5·104 N γ1 = 0,3 v = 40 km/h 80,0 8,0
Q2 = 6,2·104 N γ2 = 2,6 v = 70 km/h 100,0 10,0
Q3 = 10 ·104 N γ3 = 5,0 v = 100 km/h
150,0 18,0

for P i Q: 1 – Rail load during the passage of an empty freight train, 2 – Rail load during the passage of a passenger train, 3 – Rail load during the
passage of a fully loaded freight train;
forγ: 1 – indicates a straight track, 2 – curve, 3 – curve + incline/gradient;
The provided speeds refer to the restrictions present on Polish State Railways (PKP) in Poland

Table 3 The research plan

conforms to a D-optimal partial Test number Controlled factors on a normalized Controlled factors on a real scale
selection plan scale

Q̂ γ̂ n̂ Q, N γ, % n, min.−1

1 −1 −1 +1 500 0,3 300

2 +1 −1 −1 2000 0,3 100
3 −1 +1 −1 500 5,0 100
4 +1 +1 +1 2000 5,0 300
5 −1 0 0 500 2,6 200
6 +1 0 0 2000 2,6 200
7 0 −1 0 1000 0,3 200
8 0 +1 0 1000 5,0 200
9 0 0 −1 1000 2,6 100
10 0 0 +1 1000 2,6 300
11 0 0 0 1000 2,6 200

−1 lowest level; 0 central level; 1 highest level

Qmin = 500 N Qc = 1000 N Qmax = 2000 N
γmin = 0,3% γc = 2,6% γmax = 5%
nmin = 100 obr/min nc = 200 obr/min nmax = 300 obr/min

The analysis of measurement results based on the Hart- Table 4 Comparison of stress values depending on the type of system
ley’s plan (Table 3) involves calculating coefficients of a in experimental modes
second-degree polynomial, verifying their significance, and Loading in Stress in Stress in
checking the adequacy of the polynomial. The polynomial Wheel-rail roller-roller
coefficients are calculated according to the general princi- Real Laboratory contact* system*
conditions stan
ples of the least squares method, applying matrix calculus
(normal equations) with electronic computation techniques 100 kN/wheel 2000 N 836 MPa 875 MPa
taken into account (Table 4).
62,5 1000 N 661 MPa 618 MPa
Tribological tests were carried out on the basis of an estab- kN/wheel
lished experimental plan. The experimental plan required the 27 kN/wheel 500 N 434 MPa 437 MPa
determination of the limiting values of the input quantities
* –The stress values calculated from Hertzian formulas
for which measurements of the output quantities are carried
out. On the basis of the experimental plan, which assumed
three levels of the controlled factor, i.e. minimum indicated of the values of the output quantities were carried out. The
by “−1”, central “0” and maximum “ + 1”, measurements obtained measurement results were repeated three times and

123
 Multiscale and Multidisciplinary Modeling, Experiments and Design

then subjected to statistical analysis as a basis for substantive 3 Results

analysis. The results of a correct experimental plan are output
quantities called result factors (wear or friction coefficient). Laboratory studies constitute a crucial element in under-
In order to determine the forces acting under laboratory standing and assessing the wear process occurring in rolling-
conditions in relation to real-world conditions in the wheel- sliding contact, depending on operational factors such as load
rail system, the theory of similarity was employed, utilizing (in N), sliding (as a percentage), and speed (rotation speed
the following formula and assuming geometric similarity of in 1/min). A distinctive feature of laboratory research is the
the considered friction nodes: greater ability to directly control the impact of a selected
    factor compared to real-world conditions.
P P N N
 
= = (1)
R ·L R·L m 2 m2
3.1 Tribological Tests
where:
P’ i P – load under laboratory conditions and in a real Comprehensive tribological studies were conducted on three
object N, types of rail steels: untreated, heat-treated, and bainitic steel,
R’ i R – equivalent radius under laboratory conditions and in dry and wet contact. The experiments were based on a
in a real object m, polyselection D-optimal experimental design according to
L’ i L – contact length under laboratory conditions and in Hartley with triple repetition. The research results under-
a real object m. went thorough correction to confirm the main trends in the
According to the operating conditions on the PKP tracks, formation of the analyzed functions (Figs. 5–7).
the load for a passenger car is 62,500 N/wheel, while for Figure 5 shows an increase in wear as the load increases
an empty and loaded freight car, it is 20,000 N/wheel and in the dry contact (798–1476 mg), while in the wet contact
100,000 N/wheel, respectively. This determines the magni- the increase dominated in the initial phase( 297–342 mg),
tude of forces and, consequently, the stresses occurring in the to decrease slightly at the maximum load value (335 mg).
rolling-sliding contact in the friction node. By comparing the The load may already be so high that water deposited in the
compressive stresses generated in the contact zone during the cracks of the irregularities acts on the surface by increasing
rolling-sliding friction of two interacting elements, both in pressure in the form of an elastohydrodynamic wedge. The
a real-world and laboratory setting, it is possible to closely coefficient of friction in both cases increases with increasing
approximate the conditions prevailing in both friction zone load.
(Table 4). In dry contact, an increase in the load of the railway rolling
The PN-82/H-04332 standard specifies the conduct of stock results in an increase in wear intensity (798–1476 mg)
tests in both unlubricated and lubricated contact, aiming to and friction coefficient values (0.53–0.66). The most critical
reflect real-world conditions on the track. Dry friction is com- event is the passage of a loaded (freight) train at high speed
monly employed in the tests due to the ease of controlling on a straight section, as it increases the wear intensity. The
the controlled factors. Untreated pearlitic steel is the stan- passage of a fully loaded freight train on a curve at a low
dard material used for railway tracks. A series of studies speed also leads to increased wear (Fig. 6).
have been conducted to determine the primary directions of The most concerning signals during tribological testing
wear formation and the friction coefficient. reflecting the passage of a passenger or freight train on a
On the basis of the literature analysis and our own research straight section at maximum speed. The wear value is close to
(Tables 2 and 4), it was concluded that the operating parame- the maximum wear value, which may lead to the occurrence
ters with a significant impact on the durability of the railway of fatigue damage on the friction surface (Fig. 7).
rail are load, slippage and speed. The choice of minimum
and maximum values was determined by the conditions on 3.2 Metallographic examination
the selected track section. The minimum load level reflects
the passage of an empty goods train, while the maximum Samples after the Amsler tribological test were subjected to
level reflects the passage of a fully loaded goods train. The microscopic examinations to elucidate wear mechanisms and
minimum value of slip occurs when the train is running on observe differences in dry contact. Differences in the wear
a straight section of track and the maximum on a curve process were observed on the friction surfaces of the samples
combined with a gradient and/or an incline. The speeds are after the tests, indicating a change in the wear mechanism
defined by the restrictions on the track section. under the same operational parameters, i.e., load, sliding,
and speed (Figs. 8–12).
As a result of the tribological tests, wear debris detach-
ing during friction from the rolling-sliding association were

123
Multiscale and Multidisciplinary Modeling, Experiments and Design

samples from untreated steel samples from heat-treated steel samples from bainitic steel

a)

b)

Fig. 5 The relationship between mass loss and friction coefficient in rolling-sliding contact after tribological testing under load during passenger
train: a dry contact, b wet contact; under various sliding conditions and rotation speeds

samples from untreated steel samples from heat-treated steel samples from bainitic steel

a)

b)

Fig. 6 Dependence of weight loss and coefficient of friction in a rolling-sliding contact after a tribological test with load during the passage of a
goods train loaded to the maximum: a in dry contact, b in wet contact under different sliding conditions and rolling speed

123
 Multiscale and Multidisciplinary Modeling, Experiments and Design

samples from untreated steel samples from heat-treated steel samples from bainitic steel

a)

b)

c)

d)

Fig. 7 The relationship between mass loss and friction coefficient in rolling-sliding contact after a tribological test corresponding to a train passage
at rotation speed: a, c minimum in dry contact, b, d maximum in wet contact; under different load and sliding conditions

123
Multiscale and Multidisciplinary Modeling, Experiments and Design

rail surface from untreated steel rail surface from heat-treated steel rail surface samples from bainitic steel

a)

b)

Fig. 8 The surface of samples from the steel used for rails after the tribological test in: a dry contact, b wet contact

a) b) c)

Fig. 9 View of flake wear debris: a wear debris example obtained from real object – wheel-rail system, b, wear debris, c metallographic specimen
of wear debris obtained from laboratory stand

obtained. Once the products were collected, they were quan- it possible to improve safety in train operations (Fig. 10, 11).
titatively and statistically analysed and computer acquisition The determination of diagnostic levels made it possible to
was performed (Fig. 9). The analysis allows the wear process classify the wear mechanism according to the dimension of
to be quickly identified and the wear rate of the tribologi- the wear debris.
cal system to be assessed. During the study, weight loss and At the dry contact, an increase in rolling stock loading also
observation of the surface of the samples were systematically increases the wear intensity (798–1476 mg) and the friction
measured and wear products were collected. coefficient value (0.53–0.66). Comparing the dry and wet
Carrying out a quantitative metallographic evaluation of contact, there was a high agreement with the results obtained
the flaky wear products made it possible to determine the empirically on the real object, the wear also decreased more
distribution of the studied particles and to determine their than twice (798–297 mg) for an unloaded goods train and
density function. Based on the histograms of the distribution more than four times for a maximally loaded goods train
of geometric features, the average diameter of the flake wear (1476–335 mg). With a light load, the effect of water squeez-
debris was determined. It is extremely difficult to carry out ing into the gap is less and thus the pressure in the gap is lower.
a quantitative metallographic analysis of wear debris on the At high loading, the pressure is high enough to cause damage
basis of the area or parameter, based on the above-mentioned to the surface layer of the rail. In addition, the lower veloc-
stereological features. On this basis, it will be possible to ity results in a longer contact time between the two mating
infer the nature and type of wear and tear, which will make surfaces, which contributes to the perpetuation of the fatigue

123
 Multiscale and Multidisciplinary Modeling, Experiments and Design

Fig. 10 Histogram of the wear debris geometrical parameters and their density functions

gap dislodgement effect and the enlargement of RCF dam- is not very high and may lead to the occurrence of surface
age. The presence of a lubricating medium in the form of fatigue of the rolling surface.
water on both straight section and and in the curve, reduces The most worrying signals come from a tribological test
the wear value by more than 2–4 times. This does not change reflecting the passage of a passenger or goods train on a
the fact that the passage of a train in a curve in a lubricated straight section at maximum speed. The wear value is close
contact at low speed can lead to the appearance of RCF-type to the maximum wear value. This can lead to the appearance
damage. of fatigue damage on the friction surface. The mass loss of
In a dry contact, the most dangerous event is the passage the rolls after heat treatment and without changes paraboli-
of a loaded (freight) train at high speed on a straight section, cally with increasing slip reaching a maximum at γ = 2.6%,
as it increases the intensity of wear. The passage of a loaded which is almost seven times greater than at 0.3%. The rea-
goods train in a curve at low speed also leads to an increase in son for such an association behaviour is, among other things,
wear. In the wet, straight section, the passage of an unloaded the increase in temperature at the interface between the two
freight or passenger train can lead to an intensification of cooperating elements (from 22 °C to 47 °C, measured by the
the wear process. The situation is similar when a train passes contact method). The increase in temperature changes the
through a curve. As the train load increases, the value of wear wear mechanism and the appearance of the friction surface,
decreases slightly, but the value of the friction coefficient where colour changes and numerous collapses appear. An
increases and no RCF damage appears on the running surface increase in slip causes an increase in frictional force, break-
of the rails.As the speed increases, the wear value decreases ing of frictional bonds and a shift in the point of onset of
in both the dry and wet contact, this is due to the formation plastic deformation towards the surface.
of an elastohydrodynamic film of negligible thickness. The The samples after the tribological test on the Amsler test
effect of speed on the wear mechanism is significant, as it stand were subjected to microscopic examination in order
changes the value of the wear intensity. A high wear value to clarify the wear mechanisms and observe the differences
was also observed in the wet contact reflecting the passage in the dry and wet contact. On the friction surface of the
of an unladen passenger or goods train at a low value of the samples after the tests, differences in the wear process were
friction coefficient. Under such conditions, the wear intensity observed, which show a change in the wear mechanism at the
same operating parameters, i.e. load, sliding and speed. The

123
Multiscale and Multidisciplinary Modeling, Experiments and Design

samples from untreated steel samples from heat-treated steel samples from bainitic steel

a)

b)

c)

d)

Fig. 11 Dependence of the mean diameter of the flaky wear debris in the rolling-sliding contact after the tribological test with the load during the
passage of: a, c passenger train, b, d maximally loaded goods train; under different sliding conditions and rolling speed

cracks that appear in the wet contact lead to jagged upward- Table 5 Suggested diagnostic levels
sloping flakes on the surface. This is in contrast to dry contact
Dimensions of the flake wear debris Object wear -
surfaces, where there is no wear in the form of pearl-like object condition/wear
flaking of wear products from the friction surface. The wet mechanism
contact friction surface has many irregularities, into which
water can additionally penetrate, contributing to further crack Below 200 μm Normal (abrasive)
propagation due to the fluid confining effect of the joint. The 200–400 μm Acceptable
water pressure is so great that it causes the crack to become (abrasive/adhesive)
fractured, leading to the appearance of increasingly large pits 400–600 μm Unacceptable (spalling
or pitting)
on the surface.
The condition of a machine usually cannot be determined Above 600 μm Requiring increased
control
from a single (only one) quantitative metallographic anal-
ysis of the wear process. Although an experienced user
of this technique can draw conclusions from the size of
the wear products—large chunks, for example, undoubtedly basis of extensive experience based on an analysis of change
demonstrate significant fatigue of the material. It would be trends, either on the same or a similar type of system.
extremely convenient if the ’absolute’ size of wear prod- On the basis of the studies carried out on the average diam-
ucts could give a definitive answer about the degree of wear eter of wear products, diagnostic levels can be proposed using
of a machine/system. However, this is only possible on the quantitative metallographic evaluation of the rolling/slipping
association depending on the operating conditions, defining
ranges of safe operation (Table 5).

123
 Multiscale and Multidisciplinary Modeling, Experiments and Design

In order to compare and explain the wear mechanisms of 4 Conclusion

the rails tested in the real objects and the rollers used for
the laboratory tests, metallographic tests were carried out on On the basis of the tribological and metallographic studies
samples taken from them in selected sections of the rails and carried out, it was concluded that:
rollers in the surface layer and in the core (Fig. 12).
From the metallographic examination of the surface layer
of the rail after operation, it was found that cracks formed 1. Destruction of the surface layer in a rolling-sliding asso-
at a certain depth and propagate parallel along the rolling ciation takes the form of a delamination wear mechanism
direction (Fig. 12). As the subsurface crack increases, the in the form of flaky wear products.
cracking can reach a critical value, after which the material 2. The detaching wear products have a shape and dimension
product can be detached. Analysing the surface layer of the that depends on material and operational factors affecting
roller after the co-operation carried out in laboratory tests, it their formation and propagation process.
was found that the wear mechanism follows the same pattern 1. The dimension and shape of wear products are a reflec-
as during operation in real track conditions. Furthermore, the tion of the condition of the surface layer, as a result of an
rolling surface of the rail is very similar to that obtained on efficient system for collecting and identifying wear prod-
the laboratory stand. ucts and the condition of the friction surface, the technical
condition of the object can be successfully monitored.
2. In a dry contact, there is a relationship between wear
and friction coefficient, leading to an increase in wear
intensity or a change in the wear mechanism.

a) b)

c) 100 μm
d)

Fig. 12 View of the rolling surface and metallographic specimens: a—rolling surface of the railway rail, b—rolling surface of the roller after
cooperation, c—un-etched cross-sectional view of the surface layer of the rail, d—etched cross-sectional view of the surface layer of the roller

123
Multiscale and Multidisciplinary Modeling, Experiments and Design

3. In wet contact, there is an inverse correlation of wear and Donzella G, Faccoli M, Mazz A, Petrogalli C, Roberti R (2011a) Pro-
coefficient of friction, leading to the appearance of RCF gressive damage assessment in the near-surface layer of railway
wheel–rail couple under cyclic contact. Wear 271:408–416
damage on the rolling surface. Donzella G, Faccoli M, Mazzů A, Petrogalli C, Roberti R (2011b) Pro-
4. In wet straight section contact, the wear mechanism gressive damage assessment in the near-surface layer of railway
changes for maximum values of speed and load, which wheel–rail couple under cyclic contact. Wear 271:408–416
can lead to the appearance of fatigue/contact damage. Fleming JR, Suh NP (1977) Mechanics of crack propagation in delam-
ination wear. Wear 44:39–56
5. At minimum speed, wear is much more severe (high Freisinger M, Fellner S, Gammer C, Riedl H, Hahn R (2024) Stratified
chance of fatigue wear) than at higher speeds, which is surface layers affecting crack propagation in wheel-rail contacts.
dictated by the formation of a lubricating medium layer Tribol Int 192:109319
between the mating surfaces (so-called elastohydrody- Gallardo-Hernandez EA, Lewis R (2008) Twin disc assessment of
wheel/rail adhesion. Wear 265:1309–1316
namic wedge).
Garnham JE, Davis CL (2011) Estimation of wheel/rail adhesion
6. It is advisable to make a mobile device to monitor the coefficient under wet condition with measured boundary friction
condition of the surface layer, taking into account the coefficient and real contact area. Wear 271:32–39
diagnostic criteria proposed in the paper. Performing GUS. Statistical analyses. Environment 2022. Available
30.12.2023. https://stat.gov.pl/files/gfx/portalinformacyjny/pl/
additional tests with the above-mentioned device is a fast,
defaultaktualnosci/5484/1/23/1/ochrona_srodowiska_2022.pdf
non-invasive and inexpensive solution to complement the Halamaa R (2011) Rostislav Fajkoˇs b, Petr Matuˇsekb, Petra Babkovac,
ultrasonic and magnetic tests in order to make the right Frantiˇsek Fojtik a, Leo Vaclavek: Contact defects initiation in
decision about the technical condition of the tested object. railroad wheels – experience, experiments and modelling. Wear
271:174–185
Hiroyuki M, Satohb Y, Kanematsub Y, Iwafuchib K (2011) On-site
investigation and analysis of flaking damage leading to rail break.
Author contribution All authors contributed to the study conception Wear 271:168–173
and design. Material preparation, data collection and analysis were per- Hu Y, Zhou L, Ding HH et al (2020) Investigation on wear and rolling
formed by Henryk B˛akowski. All authors read and approved the final contact fatigue of wheel-rail materials under various wheel/rail
manuscript. hardness ratio and creepage conditions. Tribol Int 143:106091
Kvarda D, Galas R, Omasta M, Shi LB, Ding HH, Wang WJ, Krupka
Funding Silesian University of Technology. I, Hartl M (2021) Asperity-based model for prediction of traction
in water-contaminated wheel-rail contact. Tribol Int 157:106900
Data availability No datasets were generated or analysed during the Mazzù A, Battini D, Zani N (2024) Computational assessment of
current study. ratcheting in rail-wheel contact with solid contaminant. Wear
546–547:205346
Declarations Olofsson U, Lewis R, Harmon M (2019) Tribology of the Wheel-
Rail Contact from. Handbook of Railway Vehicle Dynamics CRC
Press. NY
Conflict of interest The authors declare no competing interests. Roylance BJ, Williams JA, Dwyer-Joyce R (2000) View all authors
and affiliations wear debris and associated wear phenomena—-
Open Access This article is licensed under a Creative Commons fundamental research and practice proceedings of the institution
Attribution 4.0 International License, which permits use, sharing, adap- of mechanical engineers. Part J: J Eng Tribol 241(1):79
tation, distribution and reproduction in any medium or format, as Wang W, Li S, Ding H, Lin Q, Galas R, Omasta M, Meli E, Guo J, Liu
long as you give appropriate credit to the original author(s) and the Q (2023) Wheel/rail adhesion and damage under different contact
source, provide a link to the Creative Commons licence, and indi- conditions and application parameters of friction modifier. Wear
cate if changes were made. The images or other third party material 523:204870
in this article are included in the article’s Creative Commons licence, Zakharov S, Zharov I (2002) Simulation of mutual wheel/rail wear.
unless indicated otherwise in a credit line to the material. If material Wear 253:100–106
is not included in the article’s Creative Commons licence and your Zhang SY, Liu QY, Wang WJ, Spiryagin M, Lin Q, Ding HH, Wu Q,
intended use is not permitted by statutory regulation or exceeds the Zhou ZR (2022) Implications of water medium for the evolution of
permitted use, you will need to obtain permission directly from the copy- rolling contact fatigue under rail surface defect conditions. Tribol
right holder. To view a copy of this licence, visit http://creativecomm Int 175:107870
ons.org/licenses/by/4.0/. Zhang SY, Ding HH, Lin Q, Liu QY, Spiryagin M, Wu Q, Wang WJ,
Zhou ZR (2023) Experimental study on wheel-rail rolling contact
fatigue damage starting from surface defects under various opera-
tional conditions. Tribol Int 181:108324
References Zhong W, Hu JJ, Shen P, Wang CY, Lius QY (2011) Experimental
investigation between rolling contact fatigue and wear of high-
Al-Juboori A et al (2019) Evolution of rail surface degradation in the speed and heavy-haul railway and selection of rail material. Wear
tunnel: the role of water on squat growth under service conditions. 271:2485–2493
Eng Fract Mech 209:32–47 Zhou X, Li S, Wang J, Wang K, Jing L (2023) Fatigue crack growth in
Clayton P (1995) Predicting the wear of rails on curves from laboratory wheel-rail rolling-sliding contact: a perspective of elastic-plastic
data. Wear 181–183:11–19 fracture mechanics criterion. Wear 530–531:205069
Dayot C, A. Saulot n, C. Godeau, Y. Berthier. (2012) Tribological
behaviour of pearlitic and Bainitic steel grades under various slid-
ing conditions. Tribol Int 46(2012):128–136 Publisher’s Note Springer Nature remains neutral with regard to juris-
dictional claims in published maps and institutional affiliations.

123