Content uploaded by Junyu Tian
Author content
All content in this area was uploaded by Junyu Tian on Nov 16, 2021
Content may be subject to copyright.
Comparison of the Impact Wear Performances of
Quenching and Partitioning and Quenching and
Tempering Steels
Zhirui Wei, Xiaolong Gan,* Man Liu, Junyu Tian, Zhenye Chen, and Guang Xu*
1. Introduction
Cracks and material degradation often occur in high-strength
steels due to impact wear.
[1–3]
A hardening layer is generally
formed during impact wear, increasing the strength and hard-
ness of steel and also reducing the plasticity and fracture tough-
ness. The wear resistance of steel is related to both hardness and
toughness, and the decrease in toughness leads to the formation
of defects.
[4]
Conventional wear-resistant steels are normally
processed by a direct quenching and low-temperature tempering
(QT) treatment; thus, a good combination of strength and tough-
ness is obtained because of the formation of a tempered martens-
ite structure.
[5–7]
The tempering process improves the toughness
with a small reduction in hardness; how-
ever, in most cases, the improvement of
elongation and fracture toughness is
limited.
[8,9]
To further improve the comprehensive
property of martensite wear-resistant
steels, the quenching and partitioning
(QP) process has been widely used.
[10–12]
A completely austenitized steel is directly
quenched to a temperature between the
martensite starting temperature (M
s
) and
the martensite finishing temperature
(M
f
), followed by partitioning for some
time at a temperature higher than M
s
.
The QP treatment improves the toughness
by increasing the amount of retained aus-
tenite (RA). Since Speer et al.
[13]
have pro-
posed the QP process and the constrained
carbon paraequilibrium (CCE) model,
many studies have been conducted on
QP steels. Liu et al.
[14]
reported that the best comprehensive
mechanical property depended on the maximum volume frac-
tion, mechanical stability, and morphology of RA. Oskari
et al.
[15]
compared the wear resistances of QT and QP steels
and indicated that the presence of RA seemed to have no signifi-
cant influence on wear resistance. Ding et al.
[16]
claimed that as
compared with QT steels, the impact toughness of QP steels was
improved by about 43.79%, the strength only decreased by about
6.38%, the hardness hardly decreased, and the wear resistance
was improved because of the improvement of the comprehensive
performance. Different opinions are available on whether the
wear resistance of QP steels is improved by RA. Therefore, it
is necessary to clarify the effect of the QP process on the wear
performance of high-strength steels.
In this study, a series of QP and QT processes were carried
out to investigate the wear behavior of a low-carbon and high-
strength steel. The wear performances of the samples at two dif-
ferent impact energies were studied. The correlation of wear
loss, hardness, and fracture toughness (K
IC
) was proposed to
predict the wear performances of the test steel under different
heat treatment processes. Moreover, the wear experiment of
steels is very complicated and time consuming, so the fitting
model is useful for simplifying the comparison of wear resis-
tance of the tested steel after different heat treatments. For other
steels, the new parameter of min the model can be calculated
with several groups of wear tests, and the new fitting curves also
can predict the wear loss of samples treated by other heat
treatments.
Z. Wei, X. Gan, M. Liu, J. Tian, G. Xu
The State Key Laboratory of Refractories and Metallurgy
Key Laboratory for Ferrous Metallurgy and Resources Utilization of
Ministry of Education
Wuhan University of Science and Technology
Wuhan 430081, China
E-mail: xuguang@wust.edu.cn; ganxiaolong@wust.edu.cn
Z. Chen
Plate Department
Technology Research Institute of HBIS
Shijiazhuang 050000, China
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/srin.202100325.
DOI: 10.1002/srin.202100325
A low-carbon, high-strength steel is treated by different quenching and parti-
tioning (QP) and quenching and tempering (QT) routes in a salt bath furnace, and
its wear performance is evaluated by impact abrasive wear tests. It is observed as
compared with the traditional QT steel; the QP steel manifests better wear
performance at the quenching temperature of 220 C. Stable film-like retained
austenite (RA) and fine martensite laths improve the wear resistance of the QP
steel at the quenching temperature of 220 C, whereas unstable blocky RA formed
in the QP steel at the quenching temperature of 190 C decrease the wear
resistance. In addition, the lower critical impact stress for crack initiation at the
higher impact energy decreases the wear resistance; however, the relative wear
resistance is improved greatly at the higher impact energy due to the better
fracture toughness of the QP steel. Moreover, the correlation of wear loss,
hardness, and
K
IC
is modeled to compare the wear resistances of the test steel
after different heat treatments.
RESEARCH ARTICLE
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (1 of 12) © 2021 Wiley-VCH GmbH
2. Experimental Section
The chemical composition of the test steel is shown in Table 1.
The steel was refined in a 50 kg vacuum induction furnace and
hot rolled to a 17 mm plate. Cylindrical samples of
Φ10 120 mm
2
and cuboid samples of 70 15 50 mm
3
size
were machined from the hot-rolled plate. The M
s
was measured
by a Gleeble-3500 thermomechanical simulator using the cylin-
drical samples. The cylindrical samples were heated to 950 Cat
10 Cs
1
, held for 5 min, and quenched to ambient tempera-
ture at 20 Cs
1
.TheM
s
was measured as 315 Cbasedon
the dilatation curve during the quenching process. Heat
treatments were conducted on the cuboid samples in a heating
furnace and five salt bath furnaces, and the proposed heat-
treatment procedures are shown in Figure 1. The samples were
first austenitized at 950 C in a heating furnace. For the QT
treatment, the samples were directly water quenched to ambi-
ent temperature, followed by tempering at 200 C for 60 min in
a salt bath furnace, and the corresponding samples were labeled
as QT200. For the QP treatment, the samples were put into
three salt bath furnaces at different quenching temperatures
of 190, 220, and 250 C, followed by partitioning in another salt
bath furnace at 350 C. The partitioning temperature of 350 C
was 30 ChigherthanM
s
to make the partitioning of carbon
more homogeneous and complete. In addition, the bainite
and film-like RA could be formed at higher partitioning temper-
ature, which increased the toughness of test steel without
decreasing much hardness. Longer partitioning time of
15 min was corresponding to the tempering time and to get
more bainite and film-like RA (fRA) for improving the tough-
ness.
[17]
High content of Si inhibited most precipitation of car-
bides during partitioning.
The tempering temperature was selected as 200 Cbasedon
the tempering technology parameters of the tested steel in
industrial production. The maximum amount of residual aus-
tenite temperature (T
RAMAX
), the fine martensite starting tem-
perature (T
FMS
), and one temperature between them were
selected as quenching temperatures. T
RAMAX
was calculated
as 193 C by the CCE model proposed by Speer et al.
[13]
The
quenching temperature versus RA volume curve for the test
steel is shown in Figure 2. The volume of RA (V
RA
)first
increased as the quenching temperature decreased because of
the high stability of RA and then decreased. T
FMS
was deter-
mined as 253 C based on the method proposed by Liu
et al.
[14]
The temperature–dilatation curve and the first-order
dilatation curve for M
s
are shown in Figure 3.Figure3ashows
the formation of fine martensite. The transformation of coarse
martensite first accelerated, and then the transformation kinet-
ics was significantly retarded. Transformed coarse martensite
grains applied hydrostatic pressure on untransformed austenite
and refined austenite grains. Both fine austenite and high-
dislocation density refined the martensite structure after the
burst transformation. According to Liu et al.,
[14]
the linear part
of the first-order dilatation curve could be considered as the
burst transformation period. The temperature deviating from
the linear fitting was definedasT
FMS
(Figure 3b). Quenching
temperatures were finally determined as 190, 220, and
250 C, and the corresponding samples were labeled as
QP190, QP220, and QP250, respectively.
The effects of different heat-treatment processes on wear
resistance were evaluated by various characterization techni-
ques. Each test was repeated at least three times to obtain more
accurate results. The samples were polished and etched in 4%
Nital to observe the microstructure. Microstructures after heat
treatments and morphologies after wear tests were observed by
a ZEISS optical microscope and a Nova 400 Nano field-emission
scanning electron microscope (FE-SEM). Then the samples
were electrolytic polished with 1:4 perchloric acid and glacial
acetic acid to eliminate the stress layer, and the electrolytic
Table 1. Chemical composition of the tested steel (wt%).
CSiMnMoCrVNPS
0.208 1.93 1.94 0.252 1.07 0.036 0.004 0.008 0.003
Figure 1. Heat-treatment processes.
Figure 2. Quenching temperature versus RA volume fraction curve of the
test steel predicted by the CCE model.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (2 of 12) © 2021 Wiley-VCH GmbH
voltage was 20 V, electrolytic current was 0.2 mA. RA volumes
were measured by X-ray diffraction (XRD) under Co Kαradia-
tion at 40 kV and 40 mA. Microhardness measurements were
carried out on a Zwick hardness tester with 1 kg load.
Tensile tests were conducted on a UTM-5305 electronic univer-
sal tensile tester at ambient temperature at 0.36 mm/min ten-
sile rate. An impact test machine was used to carry out impact
tests on V-notched samples at ambient temperature. In addi-
tion, impact abrasive wear tests were carried out on an MLD-
10 impact abrasive wear machine. The impact abrasive wear
machine consisted of a 10 kg hammer, abrasives, a test sample,
and a downsample (GCr15; austenitized for 30 min and water
quenched) (Figure 4). The impact energy range of the impact
abrasive wear machine was 1–5 J, and two different impact
energies (2 and 4 J) were used in this experiment to simulate
the impact process under different working conditions. The
impact frequency was 150 times per minute. The downsample
was rotated at 200 rpm. The attrition rate of wear particles was
1.5kgmin
1
. The test samples were cleaned by an ultrasonic
oscillation, and weighing 6 times with an electronic balance
during 1 h impact abrasive wear test. Angular brown fused alu-
mina of 3 mm size was used as abrasive particles (Figure 5). The
grain structure of brown fused alumina could be easily con-
trolled during solidification; thus, their size distribution
remained uniform. When brown fused alumina particles are
refined to a very small size, they hardly experience any wear;
hence, 3 mm brown fused alumina particles were used in this
experiment.
Figure 3. a) Temperature dilatation curve and b) first-order dilatation curve for M
s
.
Figure 4. Schematic diagram of the impact abrasive wear machine.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (3 of 12) © 2021 Wiley-VCH GmbH
3. Results
3.1. Microstructure
Figure 6 shows the microstructures of the samples after the QT
treatment at 200 C and different QP processes. Typical martensite
lath bundles (LM) and wedge-type tempering martensite (TM) were
observed in the QT200 sample (Figure 6a). The areas with white lath
RA paralleling to lath bainite ferrite were considered as bainite,
whereas irregularly blocky and wedge-shaped areas were defined
as martensite.
[18,19]
In comparison with the QP samples, the
QT200 sample had thicker martensite laths and bundles, whereas
wedge-type martensite grains were smaller due to the lower temper-
ing temperature. The microstructures of the QP samples consisted
of martensite laths, wedge-type martensite, and a small amount of
bainite formed during isothermal holding at 350 C(Figure6b–d).
The morphologies of bainite and RA at higher magnification
are shown in Figure 7. More bainite laths are observed in QP250.
Residual austenite after quenching transforms to bainite during
partitioning process. More residual austenite grains were
retained after quenching at 250 C; thus, the QP250 samples con-
tained more bainite than QP220 and QP190. In addition, mar-
tensite laths in the QP250 sample were relatively thicker
because the quenching temperature was close to T
FMS
No obvious RA was detected in the QT200 sample, and the
morphology of the QP samples consisted of fRA or blocky RA
(bRA). The amount of bRA in QP190 was higher than those
in QP250 and QP220 (Figure 6d), and this result is consistent
with the RA changing trend calculated by the CCE model
(Figure 2). Some stable undercooled austenite grains in
QP190 were not transformed into bainite during the partitioning
process or into fresh martensite during second cooling; thus,
they were retained as bRA.
3.2. Mechanical Properties
The XRD diffraction profiles of different samples are shown in
Figure 8. No obvious RA peaks were observed in the QT200
sample, and the amount of RA was calculated by Equation (1).
Figure 5. Abrasive particles.
Figure 6. SEM microstructures of different samples at 5000 magnification: a) QT200, b) QP250, c) QP220, and d) QP190.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (4 of 12) © 2021 Wiley-VCH GmbH
Vi¼1
1þGðIα=IγÞ(1)
where V
i
is the amount of RA, Gis the ratio of the integrated
intensity of each ferrite peak to that of each austenite peak, I
α
and I
γ
are the integrated intensities of ferrite and austenite peaks,
respectively. The Gvalue was taken from the literature.
[20]
The
RA volume fraction (V
RA
) in the QT200 sample was calculated
as 0.1%. The V
RA
in the QP samples matched with the trend cal-
culated by the CCE model, and the measured V
RA
values of
QP190, QP220, and QP250 were 19.25%, 14.5%, and 10.1%,
respectively.
The mechanical properties are shown in Table 2. The fracture
toughness (K
IC
) was calculated by Equation (2).
[21]
KIC ¼δyffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
θαk
δy
!
ω
v
u
u
t(2)
where the δyis the yield strength, αkis the impact toughness, θis
the sample size coefficient. ωis the Shear lip width coefficient.
The QT200 sample had the highest yield strength and hardness;
however, its impact toughness was the worst, as shown in
Table 2. The hardness of the QP samples decreased with the
decreasing quenching temperature.
Figure 7. Samples at 10 000 maganification: a) QT200, b) QP250, c) QP220, and d) QP190.
Figure 8. XRD diffraction profiles.
Table 2. Mechanical properties of different samples.
Sample Tensile strength
[MPa]
Hardness
[HV]
Impact toughness
[J cm
2
]
K
IC
[MPa m
1/2
]
RA [%]
QP190 1340 7 426 6922 95.31 19.3 0.2
QP220 1401 11 447 3964 100.82 14.5 0.3
QP250 1357 9 435 5933 97.59 10.1 0.2
QT200 1429 15 482 8882 89.91 0.1 0.05
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (5 of 12) © 2021 Wiley-VCH GmbH
3.3. Wear Loss
The wear loss versus time curves at two different impact energies
are shown in Figure 9. The scatter bars of three groups of repeat-
ing tests were added to eliminate the error. The wear losses at 2 J
were lower than those at 4 J. It happened because the higher
impact energy increased the number of defects and promoted
the propagation of cracks. The wear loss in the first 10 min at
2 J was greatly reduced due to the transformation-induced plas-
ticity (TRIP) effect. However, the wear loss in the first 10 min
changed slowly at 4 J, because more bRA was transformed into
deformed induced martensite (DIM) in first 10 min at higher
impact energy. The wear loss after 30 min tended to be stable
due to the completion of the TRIP effect. The wear loss trends
of QP220, QP250, and QP190 were similar at 2 and 4 J.
The relative wear resistance (W
O
) was used to compare the
improvement of wear resistance at two impact energy. W
O
was
defined as W
S
/W
T
,whereW
S
is the total wear loss of the original
sample without any heat treatment, and W
T
is the total wear loss of
the test samples. The total wear losses of the original sample after
60 min at 2 and 4 J were 212 and 335 mg, respectively. The larger
the W
O
value, the better the improvement of wear resistance.
Figure 10 shows the relative wear resistances of the samples at
the two impact energies. The W
O
values of the QP steelsat 4 J were
greater than those at 2 J. However, the W
O
of the QT200 sample at
4 J was smaller than that at 2 J. Therefore, the QP samples with
larger K
IC
exhibitedmore improvement in wear resistance as com-
pared with QT200 with smaller K
IC
at the higher impact energy.
3.4. Wear Surface
The wear morphology characteristics at 2 and 4 J were almost
identical. The wear surface morphologies of different samples
at 4 J consisted of wear debris, pits, cracks, flakes, and cuts
(Figure 11).
[22]
Typical flakes were observed in the QT200 sample
due to its high hardness (Figure 11a). Cuts, cracks, and flakes
appeared in QP250 because of its relatively smaller hardness
and K
IC
(Figure 11b). Fragments of abrasive particles and larger
deformations were observed in QP190 due to its lower hardness
(Figure 11d). The number of cuts and the deformation degree
were reduced in QP220.
Figure 12 shows the subsurface wear morphologies of different
samples at 4J. Deformations, cracks, and pits were observed. The
deformation degree gradually decreased with the increase of hard-
ness, and pits formed by abrasive particles were more obvious in
QP250 and QP190. Flakes peeled off from the matrix (Figure 12a).
Furthermore, pits in QP220 were relatively smaller than those in
QP250 and QP190, and cracks were rarely observed in QP220.
4. Discussion
4.1. The Role of Microstructure on Wear Performance
The amount and stability of austenite have great influence on
wear performance. Higher stability of austenite and strength of
bainitic ferrite led to better wear resistance, and the stability of
Figure 9. Wear losses per minute at a) 2 and b) 4 J.
Figure 10. Relative wear resistances at two different impact energies.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (6 of 12) © 2021 Wiley-VCH GmbH
Figure 11. Wear surface morphologies of different samples at 2500 magnification: a) QT200, b) QP250, c) QP220, and d) QP190.
Figure 12. Subsurface wear morphologies at 2500 magnification: a) QT200, b) QP250, c) QP220, and d) QP190.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (7 of 12) © 2021 Wiley-VCH GmbH
austenite outweighed the influence of amount of RA on wear resis-
tance.
[23]
bRA was unstable as compared with film-like RA; thus, it
was easily transformed into deformation-induced martensite
(DIM) during the wearing process.
[24]
It was proven that there
was a partial RA transformation during deformation in the
carbon-depleted area in the bRA region.
[25]
The strain-induced
transformation of blocky austenite into martensite was detrimen-
tal to the wear performance of the test steel because martensite
could not accommodate the inhomogeneous transformation
stress,
[26]
which could promote strain concentration and provided
a pathway for crack propagation at the boundaries of strain-
induced martensite and tempering martensite.
[27]
In comparison
with the other QP samples, QP220 manifested better yield
strength, impact toughness, and K
IC
. More RA in QP190 and
more bainite in QP250 decreased their strength and hardness,
whereas bRA in QP190 and relatively thicker martensite laths
in QP250 decreased their impact toughness and K
IC
.Therefore,
the best comprehensiveproperty of QP220 could be obtained from
the proper proportion of bainite, fine martensite, and RA.
The wear loss of QP190 is the largest among all samples
(Figure 9). Figure 6 and 7 shows that QP190 contains more
amount of bRA than other samples. The blocky austenite is det-
rimental to the wear performance of the test steel as mentioned
earlier. Furthermore, the hardness of QP190 is the smallest
among all samples (Table 2); therefore, more bRA and low hard-
ness resulted in the largest wear loss of QP190.
In addition, the total wear resistance of the QP220 sample is
the best among all samples at both 4 and 2 J (Figure 9). Figure 6
and 7 shows that QP220 consisted of film-like RA and a small
amount of bRA. bRA was transformed into DIM during the wear
process, and crack propagation was retarded by stable film-like
RA; hence, it can be inferred that film-like RA improved the wear
resistance. The QP220 sample has the largest hardness and K
IC
among all QP samples (Table 2); hence, it has the best wear
resistance.
The wear resistance of QP250 is worse than QP220 and better
than QP190 (Figure 9). The QP250 sample contains the largest
amount of bainite because more undercooled austenite is
retained for bainitic transformation during partitioning.
Figure 13 shows the hardness versus surface depth curves after
the wear process. The presence of more bainite in QP250 leads to
a limited increase in hardness after the wear process. As QP250
has a lower amount of RA (10.1%) than QP220 (14.5%), lower
hardness, and smaller K
IC
, it has lower wear resistance. In con-
trast, due to lower hardness and smaller K
IC
, QP190 has worse
wear resistance than QP250.
Moreover, the wear loss of QT200 is the lowest at the begin-
ning of the wear process because of its highest hardness.
However, the wear performance of QP220 is better than that
of QT200 after 20 min of wearing; it happens because the hard-
ness of QP220 obviously increases after initial wearing. Due to its
increased hardness, higher K
IC
, and more RA, QP220 samples
manifest better wear performance than QT200 at the later stage
of the wear process. However, the other QP-treated samples
exhibit worse wear resistance than QT200 due to the comprehen-
sive effect of hardness and K
IC
.
4.2. The Role of Impact Energy
The increase in impact energy increased the impact stress, which
directly affects the wear resistance of steel. It is known that K
IC
has little influence on the wear resistance of steel if the impact
stress is too low for crack propagation.
[28]
Therefore, the contact
stress between abrasives and samples (P
C
) and the critical impact
stress (P
L
) required for crack propagation were calculated to
explain the influences of fracture toughness on the wear resis-
tance of the test samples. The critical impact stress required
for crack propagation and the contact stress between abrasives
and samples can be calculated as
[29–32]
PL¼2:8KIC
H
HS
Hstanα
FI
1=4
(3)
FI¼0.136 J3R1E2
1þR1
R2
!
1=5
(4)
Figure 13. Hardness versus surface depth a) 2 and b) 4 J.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (8 of 12) © 2021 Wiley-VCH GmbH
PC¼0.388 FIE2
R1
1=3
(5)
where His the original hardness, H
s
is the hardness after the
wear process, αis the angle of cracks (45), R
1
is the size of
abrasive particles, R
2
is the size of steel sample, which is
assumed to be infinite, so R
2
¼∞.Eis elastic modulus
(53 GPa), Jis impact energy (2 and 4 J), and. F
I
is impact force.
The calculated results are shown in Table 3, and it is noticeable
that P
C
was far larger than P
L
. Hence, the wear loss was affected
by both hardness and K
IC
at 2 and 4 J. The increase in impact
energy or the decrease in hardness could reduce P
L
. The wear
loss per minute at 4 J is higher than that at 2 J (Figure 9). It hap-
pens because the lower P
L
at the higher impact energy results in
an easy propagation of cracks and more wear loss. Moreover, the
improvement of the relative wear resistance at 4 J is greater than
that at 2 J, except for the QT200 sample (Figure 10). For the QP
samples, higher K
IC
restrain the propagation of cracks and pro-
mote relative wear resistance. The improvement of the relative
wear resistance increases with the increase in the impact energy.
It happens because cracks easily propagated at the higher impact
energy; thus, the wear loss of the original sample without any
heat treatment is larger at the higher impact energy.
Therefore, the restraining effect to crack propagation and the
decrease in the wear loss in the higher K
IC
QP samples are more
obvious at the higher impact energy; hence, the relative wear
resistance is improved greatly at the higher impact energy.
The smaller K
IC
of QT200 could not restrain crack propagation
at the higher impact energy; thus, the improvement of wear resis-
tance is reduced. Hence, the wear performance of the QP sam-
ples with higher K
IC
is better at the higher impact energy.
4.3. Correlation of the Hardness,
K
IC
, and Wear Loss
The total wear loss versus hardness curves and the total wear loss
versus K
IC
curves at two different impact energies are shown in
Figure 14. It is observable that both hardness and K
IC
were non-
linear to the total wear loss. The total wear loss decreases with the
increase in hardness and then begins to increase when hardness
reaches 480 HV. The increase in K
IC
cannot improve the wear
resistance at the beginning; however, the wear loss decreased
when K
IC
is high enough to restraint the growth of cracks.
The increase in hardness is accompanied by the decrease in
fracture toughness. The wear loss is determined by both hardness
and K
IC
unless the impact stress is too low for cracks to propagate.
The wear loss can be separated into two parts—the wear loss of
wear debris and flakes depending on the hardness and the wear
loss of cracks depending on K
IC
. According to Lewis,
[33]
the wear
loss depending on hardness can be expressed as
WkPNx
HþKNJn(6)
where kis the wear coefficient of impact wear, Kis the wear coef-
ficient of slide wear,Pis impact stress, Nis impact frequency, xis
sliding distance, Jis impact energy, and nis the number of impact
cycles. The aforementioned equation indicates that the wear loss
could be decreased by increasing the hardness; however, it does
not completely agree with the test results. It is reported that wear
resistance is not completely linear to hardness,
[34–37]
and the effect
of fracture toughness is vital to wear loss. According to Shao and
Zhang,
[38]
the relationship between wear loss and K
IC
can be
described as
Table 3. Impact stresses of different test steels.
Sample P
C
(2 J) [MPa] P
L
(2J) [MPa] P
C
(4J) [MPa] P
L
(4J) [MPa]
QP190 6427.9 744.9 7357.2 672.8
QP220 6427.9 816.4 7357.2 734.8
QP250 6427.9 745.5 7357.2 670.0
QT200 6427.9 770.4 7357.2 693.4
Figure 14. Total wear loss at two impact energy a) 2 and b) 4 J.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (9 of 12) © 2021 Wiley-VCH GmbH
W∝R1=2
1P4=5
K3=4
IC H1=2
M
(7)
To correlate the relationship among wear loss, hardness, and
K
IC
, the aforementioned formula can be modified as
W∝kPNx
HþKNJnþcR1=2
1P4=5
K3=4
IC H1=2
M
(8)
where R
1
is the size of abrasive particles, cis the wear coefficient,
and H
M
is the hardness of abrasive particles. Moreover, N,x,e,n,
d,H
M
, and Pall are constants. Therefore, the aforementioned
formula can be simplified as
W∝kA
HþcC
K3=4
IC
þKB ∝
mK3=4
IC þH
HK3=4
IC
(9)
where the A,B,C, and mare constants. Qualitatively, the wear
loss is proportional to mK3=4
IC þH
HK3=4
IC
. Hence, the model can be used to
compare the wear resistances of the test steel after different heat
treatments without carrying out wear tests.
Quantitatively, the wear loss was modeled with Hand K
IC
, and
the coefficient mwas calculated according to wear loss results of
the QP samples. Both hardness and K
IC
were affected by the
TRIP effect. For more accurate fitting results, the wear loss of
each sample was separated into two parts before and after the
TRIP effect. According to Figure 9, the TRIP effect was consid-
ered to be completed after 30 min of the wear process. The total
wear loss before the TRIP effect (W
1
) depended on the original
properties of steel, whereas the total wear loss after the TRIP
effect (W
2
) depended on H
S
and K
IC
0. The obtained results
for each sample are shown in Table 4. The dynamic fracture
toughness (K
IC
0) was calculated by Equation (10), where His
the original hardness of steel.
[32]
KIC
0¼H
HS
KIC (10)
Six groups of wear loss results for QP190, QP220, and QP250
before and after the TRIP effect, H, and K
IC
were considered, and
multiple unary first-order equations were established to calculate
the value of m. The suitable value of mfor the test steel was found
as 15.
Figure 15 shows the relationships among hardness, K
IC
, and
wear loss at two different impact energies, and the corresponding
curves are fitted by the results in Table 4. The fitting formulas are
shown as follows.
Wð2JÞ¼ð6:60:88Þ15K3=4
IC þH
HK3=4
IC
ð0.37 0.01Þ(13)
Wð4JÞ¼ð7.7 0:71Þ15K3=4
IC þH
HK3=4
IC
ð0.41 0:04Þ(14)
The correlations were linear at both impact energies, indicating
that the mvalue of 15 and the correlations were suitable for the test
samples at both impact energies. The result of the QT200 sample
is shown in Figure 15, and the experimental wear loss (marked by
the blue dot) is almost similar to the predicted result (red line).
Hence, the fitting curves of wear loss can be used to calculate
the wear losses of the other samples at 30 min. Generally, wear
loss increases linearly with wearing time; hence, long-term wear
losses can also be calculated based on wear losses at 30 min.
In general, the increase in hardness is accompanied by the
decrease in K
IC
. The fitting model is useful to predict wear resis-
tance of the tested steel after other heat treatments. For example,
at least three groups of pretests are considered to fit the value of
munder any wear conditions, and then the corresponding wear
loss versus mK3=4
IC þH
HK3=4
IC
curves are obtained. Hardness and K
IC
of
other heat treatment samples of the tested steel can be easily
measured, and the best wear resistance samples can be deter-
mined by taking the hardness and K
IC
into fitting curves. This
method avoids the repeating wearing tests of other heat treat-
ment samples by making only three groups of pretests.
Regarding to the limitation of this model, the practical value of
wear loss is closely related to the practical wear conditions, so the
fitting trends of present curve are applicable to the set of tests and
conditions in this study. In addition, for steels without the TRIP
effect, H
S
and K
IC
0hardly change after wear; hence, the wear loss
at any wearing time can be fitted with initial Hand K
IC
. However,
for TRIP steels, H
S
and K
IC
0change sharply after wearing;
thereby, it will be more accurate to fit the wear loss with hardness
and K
IC
0after the TRIP effect.
5. Conclusion
The wear performances of the test steel treated by different heat-
treatment routes were investigated at two different impact ener-
gies, and the corresponding microstructures and the effect of RA
on wear resistance were also analyzed. The correlation of hard-
ness, fracture toughness, and wear loss was proposed. The main
observations are presented as follows. 1) The correlation of wear
loss, hardness, and K
IC
was modeled as W∝mK3=4
IC þH
HK3=4
IC
.
Table 4. Wear properties of different samples.
H/HV H
S
/HV
(2 J)
H
S
/HV
(4 J)
K
IC
[MPa m
1/2
]
K
IC
0
[MPa m
1/2
](2J)
K
IC
0
[MPa m
1/2
](4J)
W
1
10
3
[mg] (2 J)
W
2
10
3
[mg] (2 J)
W
1
10
3
[mg] (4 J)
W
2
10
3
[mg] (4 J)
QP190 426 6 482 8 485 5 95.31 82.78 82.26 80 5713 107 3 100 4
QP220 447 3 483 9 482 4 100.82 92.54 92.73 59 4514874764
QP250 435 5 471 6 474 6 97.59 89.58 89.98 69 5603983934
QT200 482 8 490 9 488 5 89.91 89.44 89.81 55 4563844833
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (10 of 12) © 2021 Wiley-VCH GmbH
Qualitatively, the model could be used to compare the wear resis-
tances of the test steel under different heat treatments without
calculating the mvalues for wear tests. Quantitatively, the value
of mwas calculated as 15, and the fitting curves of the QP sam-
ples matched well with those of the QT samples. Hence, the fit-
ting model of several groups of wear results can predict the wear
loss of other heat treatment samples. 2) In comparison with the
QT samples, the QP220 sample manifested better wear perfor-
mance. The amount of RA and the morphology and width of
martensite laths affected the wear resistance of the test steel.
The optimum amount of RA corresponding to the minimum
wear loss was obtained at 220 C. Unstable bRA decreased the
wear resistance of the QP190 sample, and stable film-like RA
and fine martensite laths increased the wear resistance of the
QP220 sample. 3) The relative wear resistance of the samples
with larger K
IC
was improved at the higher impact energy.
Hence, the wear performance of the high fracture toughness
steel was better at the higher impact energy.
Acknowledgements
The authors gratefully acknowledge the financial support from the
National Nature Science Foundation of China (Nos. 51874216,
51704217), the key Project of Hebei Iron and Steel Group (HG2019313).
Conflict of Interest
The authors declare no conflict of interest.
Data Availability Statement
Research data are not shared.
Keywords
fracture toughnesses, impact wear, quenching and partitioning, retained
austenite
Received: June 1, 2021
Revised: July 31, 2021
Published online:
[1] J. Lankford, Eng. Fract. Mech. 1977,9, 617.
[2] M. Lindroos, K. Valtonen, A. Kemppainen, A. Laukkanen,
K. Holmberg, V.-T. Kuokkala, Wear 2014,322–323, 32.
[3] N. P. Suh, Wear 1977,25, 111.
[4] J. R. Fleming, N. P. Suh, Wear 1977,44, 39.
[5] C. Y. Wang, X.Li, Y. Chang,S. Han, H. Dong, Wear 2016,362–363, 121.
[6] H. Y. Song, S. H. Zhang, L. Y. Lan, C. Li, H. Liu, D. Zhao, G. Wang,
Acta. Metall. Sin. 2013,26, 390.
[7] X. M. Wang, P. Z. Li, Adv. Mater. Res. 2011,399–401, 259.
[8] J. Lu, H. Yu, P. Kang, X. Duan, C. Song, Wear 2018,414–415, 21.
[9] Z. Cvijovi, M. Rakin, M. Vratnica, I. Cvijovi´cc, Eng. Fract. Mech. 2008,
75, 2115.
[10] Z. Wang, B. Huang, H. Chen, C. Wang, J. Ma, X. Zhao, J. Mater. Eng.
Perform. 2020,29,1.
[11] K. Dong, C. Lu, W. Zhou, D O. Northwood, C. Liu, Eng. Fail. Anal.
2021,123, 105290.
[12] D. A. Hills, D. Nowell, J. O’Connor, Wear 1988,125, 129.
[13] J. G. Speer, O. F. C. R. Assun, D. K. Matlock, D. V. Edmonds, Mater.
Res. 2005,8, 417.
[14] L. Liu, B. B. He, G. J. Cheng, H. W. Yen, M. X. Huang, Scr. Mater.
2018,150,1.
[15] H. Oskari, S. Mahesh, P. David, P. Kantanen, J. Kömi, N. Ojala,
V. Heino, Wear 2018,400, 21.
[16] L. P. Dong, Y. F. Sun, Study of Q&P Heat Treatment on Microstructure
and Properties of Low Alloy Wear Resistant Steel, Zhengzhou University,
Zhengzhou 2013, pp. 17–18 (In Chinese).
[17] X. F. Huang, W. L. Liu, Y. Y. Huang, H. Chen, W. Huang, J. Mater.
Process Tech. 2015,222, 181.
Figure 15. The relationships of hardness, K
IC
and wear loss under impact energy of a) 2 and b) 4 J.
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (11 of 12) © 2021 Wiley-VCH GmbH
[18] A. Navarro-Lopez, J. Hidalgo, J. Sietsma, M. J. Santofimia, Mater.
Charact. 2017,128, 248.
[19] C. Y. Wang, J. Shi, W. Q. Cao, H. Dong, Mater. Sci. Eng., A 2010,527,
3442.
[20] M. Liu, G. Xu, J. Y. Tian, Q. Yuan, M. Zhou, H. Hu, Steel Res. Int. 2019,
90, 1900159.
[21] H. F. Li, F. Z. Zhang, P. Zhang, Q. Q. Duan, G. Yao, Investigation on
Fracture Toughness and Crack Growth Mechanism of High-Strength
Steels, University of Science and Technology of China, Shenzhen
2019, pp. 46–47. (In Chinese)
[22] H. K. D. H. Bhadeshia, D. V. Edmonds, Met. Sci. 1983,17, 411.
[23] P. V. Moghaddam, J. Hardell, E. Vuorinen, B. Prakash, Wear 2019,
428–429, 193.
[24] G. Saha, K. Valtonen, A. Saastamoinen, P. Peura, V.-T. Kuokkala, Wear
2020,450–451, 203263.
[25] R. Rementeria, J. A. Jimenez, S. Y. Allain, G. Geandier,
J. D. Poplawsky, W. Guo, E. Urones-Garrote, C. Garcia-Mateo,
F. G. Caballero, Acta Mater. 2017,133, 333.
[26] J. Rendón, M. Olsson, Wear 2009,267, 2055.
[27] M. Meggiolaro, A. Miranda, J. Castro, L. Martha, J. T. P. Castro, Int. J.
Fatigue 2005,27, 1398.
[28] H. V. G. C. Sih, N. H. Int. Pub. 1973,7, 368.
[29] P.A. Engel, T.H. Lyons, J.L. Sirico, Wear 1973,23, 185.
[30] G. K. H. Zum, Microstructure and Wear of Materials, Tribology,
Elsevier, Amsterdam 1987, p. 132.
[31] Q. L. Yong, Second Phases in Structural Steels, Metallurgical Industry
Press, Beijing 2006, pp. 87–88. (In Chinese)
[32] B. Liu, W. Li, X. Lu, X. Jia, X. Jin, Wear 2019,428, 127.
[33] R. Lewis, Wear 2007,262, 1516.
[34] M. N. Mandikos, G. P. Mcgivney, E. Davis, P. J. Bush, J. Prosthet. Dent.
2001,85, 386.
[35] Y. H. Dou, H. L. Zhou, X. Z. Wang, J. W. Yang, Adv. Mater. Res. 2013,
634–638, 3595.
[36] E. Hornbogen, Wear 1975,33, 251.
[37] P. Valizadeh. Moghaddam, M. Rinaudo, J. Hardell, E. Vuorinen,
B. Prakash, Wear 2020,460–461, 203484.
[38] H. S. Shao, Q. Zhang, Abrasive Wear of Metals and Wear-Resistant
Material, China Machine Press, Beijing 1988. (In Chinese)
www.advancedsciencenews.com
l
www.steel-research.de
steel research int. 2021, 2100325 2100325 (12 of 12) © 2021 Wiley-VCH GmbH
