Wear 328-329 (2015) 378–390

Contents lists available at ScienceDirect

Wear
journal homepage: www.elsevier.com/locate/wear

Effect of WC grain size and content on low stress abrasive wear

of manual arc welded hardfacings with low-carbon or stainless
steel matrix
Vytenis Jankauskas a, Maksim Antonov b,n, Valentinas Varnauskas c, Remigijus Skirkus a,
Dmitri Goljandin b
a
Institute of Power and Transport Machinery Engineering, Aleksandras Stulginskis University, Studentų 11, Akademija, LT-53361 Kaunas District, Lithuania
b
Department of Materials Engineering, Tallinn University of Technology, Ehitajate tee 5, 19086 Tallinn, Estonia
c
Department of Materials Science and Welding, Faculty of Mechanics, Vilnius Gediminas Technical University,J. Basanavičiaus g. 28,
LT-03224 Vilnius, Lithuania

art ic l e i nf o a b s t r a c t

Article history: Abrasive wear resistance of tillage and harvesting tools is highly important for the agricultural sector
Received 17 November 2014 because abrasive wear by hard soil particles is the main factor limiting their lifetime. Manual arc welding
Received in revised form is among the easiest, most convenient, and economically feasible methods not only for coating metallic
26 February 2015
tools in small and medium scale farms, but also for mining operations. The aim of the current work was
Accepted 27 February 2015
Available online 11 March 2015
to develop electrodes for manual arc welding that enable a significant reduction of wear under three-
body abrasive conditions. Reinforcement by tungsten carbide powder was used due to possibility of
Keywords: production these powders through recycling of hardmetal scrap. The effects of three variables, namely
Hardfacing (1) binder material (low-carbon ferritic–pearlitic or austenitic stainless steel), (2) WC content, and
Metal-matrix composite
(3) WC grain size, on three-body abrasive wear resistance of hardfacings, were evaluated using the ASTM
Three-body abrasion
G65, dry sand/rubber wheel, test method (Procedure B). A reduction in wear rate by as much as a factor
Scratch testing
Soil-engaging tools of 9 was achieved by introducing of 42–43 wt% of WC. A discussion of the wear mechanisms for the
ASTM G65 hardfacings is provided, based on data and observations using scanning electron microscopy, energy
dispersive spectroscopy, X-ray diffraction, and optical emission spectroscopy.
& 2015 Elsevier B.V. All rights reserved.

1. Introduction are carbides of tungsten, chromium, titanium, tantalum or nio-

Abrasive wear resistance of tillage and harvesting tools is
highly important for the agricultural sector because abrasive wear
by hard soil particles is the main factor limiting their lifetime [1].
Several technologies are used nowadays for restoration of worn or
protection of as-received metallic tools: plasma transfer arc weld-
ing (PTAW), submerged arc welding (SAW), laser beam welding
(LBW), gas metal arc welding (GMAW), oxy-acetylene flame
brazing, friction stir processing, tungsten inert gas welding (TIG),
high-velocity oxy-fuel coating (HVOF) and arc welding by electro-
des [2,3]. The applied layer has usually higher hardness than that
of the base substrate material that is required to enhance protec-
tion against abrasive wear. In order to meet required hardness and
to provide enough resistance against impact the layer has to be
comprised of hard reinforcement and tough matrix with strong
adhesion between constituents. Typical reinforcement particles
n
Corresponding author.
E-mail address: Maksim.Antonov@ttu.ee (M. Antonov).
bium while borides and nitrides can also be used [3–6]. The most
preferable one and the most often used is tungsten carbide due to
its high hardness, high thermal conductivity, significant plasticity
(if compared to other ceramics), and good wettability by cobalt
[4–6]. Approximately 90% of ceramic–metal composites based on
WC have cobalt as the binder due to sufficient wetting and
resulting strong adhesion between phases. Binder content of bulk
composites is usually ranging from 3 to 25 wt% while coatings
have more narrow range of Co content (from 6 to 15 wt%) with the
content of 10 wt% being the most often used [4,6,7]. There is
growing interest to replace cobalt and to make Co-free cermets
mostly due to cobalt's high cost and toxicity [7]. Alternative
binders include combinations of Fe, Ni, Co, Cr, Mo, Mn, Al, B, C
and Si (Ni, FeNi, FeCo, CoCr, FeCoNi, NiCr, NiCrAl, FeNiCrMo, FeMn
(C), FeCrAl, NiCrBSi(C), NiCrFeBSi(C), etc) are studied and it is
predicted that more widespread acceptance of iron-bonded mate-
rials will take place [3–9]. WC-Co has low resistance against
oxidation at temperatures above 600 1C and materials of this
group have high wear rate at these temperatures [10]. Stainless
steel binder might improve the resistance against oxidation and

http://dx.doi.org/10.1016/j.wear.2015.02.063
0043-1648/& 2015 Elsevier B.V. All rights reserved.
 V. Jankauskas et al. / Wear 328-329 (2015) 378–390
other types of corrosion [11]. Hardfacings having high resistance
against wear and corrosion are required for the agricultural sector
and other application areas for protecting of such parts as
conveyor screws and spreaders of fertilizers and other chemicals.
The size of reinforcing particles of ceramic–metal composites is
critical since it influence the mechanical and thermal properties.
Usually composites with fine WC grains are harder, have higher
compressive and fatigue strength but lower fracture toughness
[4,12,13] as compared to the coarse-grained composites. However,
it was found that in case of HVOF coatings the hardness of coatings
with fine and medium WC grain size is lower than that of coatings
with coarse grain size [14].
Manual arc welding is among the easiest, most convenient, and
economically feasible methods for maintenance and manufactur-
ing of new components in the agricultural sector and for mining
operations. Such coatings may be applied by personnel with basic
welding skills, even under field conditions (using portable power
generators) [5,15,16]. However, in contrast to fully automated
processes these coatings may have irregular microstructure and
suffer from pores, inclusions, and other defects.
The aim of the current work was to develop electrodes for
manual arc welding enabling significant reduction of wear in
three-body abrasive conditions. The electrodes were prepared
using existing industrial equipment. The reinforcement particles
are placed into flux covering. This approach differs from traditional
method when the tungsten carbide rod or carbide powder is
encapsulated in steel tube [17]. Usually, electrodes for manual arc
hardfacing with central core wire are nickel or cobalt based
[15,16]. The studied electrodes are iron based to reduce their cost
and to rise the availability for the end-user.
Wear testing of materials was conducted in conditions being
very similar to those experienced by soil-engaging tools according
to ASTM G65 standard [18]. Abrasive particles are introduced
between the test sample and the rotating soft rubber wheel
providing support for particle and excluding their breakage [19].
The wear mechanisms of hardfacings with low and high WC
grain concentration in low stress abrasive conditions is similar to
those observed for steels and hard metals (cermets) respectively
[20–29]. The main stages include: (1) removal of the binder by
multiple scratching, grooving and indenting; (2) fragmentation of
WC grains; (3) breakaway of insufficiently supported WC grains.
The wear resistance of the binder is to some extent enhanced by
(4) re-embedding of WC and sand fine fragments into a binder and
the formation of mechanically mixed layer that is less evident for
low stress wear test. Formation of (5) white-etching layer experi-
enced by steels abraded under same wear conditions takes place
while it is not always enabling improved wear resistance [20].
Fig. 1. Schematic overview of the (A) covered electrode for manual arc welding and (B) the structure of the coating.
379
Reinforcement by tungsten carbide powder allows in future
their possible replacement by composite powders obtained through
recycling of hardmetal scrap. Cermet composite particles allow to
produce coatings with the lower dissolution of WC (lower content
of brittle phases) and more homogeneous distribution of WC along
the coating [30,31]. In order to give the basis for selection and
production of the optimum electrodes the effect of (1) binder
material (low-carbon ferritic–pearlitic or stainless steel), (2) WC
hard phase content and (3) grain size on wear rate was studied.
2. Materials and methods
2.1. Materials preparation
Preparation of covered electrodes and application of the
manually welded arc layers was performed in cooperation with
JSC “Anykščių Varis”, the producer of electrodes on an industrial
scale [32].
The welding electrode (Fig. 1A) for application of the composite
coating was composed of the central core wire covered by the flux
covering with inclusions of WC. The diameter of the core wire was
3.2 mm and the length was 350 mm. The chemical compositions of the
low-carbon ferritic–pearlitic steel (SWRY-21 according to JIS G3503 or
Св08А according to GOST2246-70) and stainless steel core wires
(ER308L according to AWS A5.9/A5.9 M:2006) are given in Table 1.
In order to ensure homogeneous distribution of constituents in the
flux covering the mixing of components was performed in two steps in
the 5 dm3 (5 litres) mixing device. During the first step, the basic flux
covering similar to UONI 13/55 (GOST 9466-75; E 7015/AWS A5.1; E42
4 B22H5/EN ISO 2560-A) was prepared out of minerals (marble,
limestone, fluorspar, quartz sand and muscovite), ferroalloys (ferro-
manganese, ferrosilicon and ferrotitanium) and sodium silicate liquid
glass. This mixture was dried, crushed and mixed during the second
step with tungsten carbide, graphite, cellulose and sodium silicate liquid
glass. The content of components added during second step of mixing
was varied depending on WC size and content. Graphite content was
adjusted to avoid formation of graphite or brittle “eta” phase [6,17,33].
During both steps, powders were dry mixed and sodium silicate liquid
glass was introduced shortly before the end of mixing.
Electrodes were formed by an industrial Oerlikon CEP 141
electrode extrusion press line. The diameter of the draw plate for
forming of the electrodes was varied between 5.0 mm and 7.8 mm
for electrodes with low and high concentration of the WC particles.
After forming, electrodes were dried for 24 h at room tempera-
ture and then for 1 h at 300 1C. The duration of the heating and
cooling cycle from room to 300 1C was 2 h.
380 V. Jankauskas et al. / Wear 328-329 (2015) 378–390

Table 1
Chemical compositions of the core wires (wt%).

Core wire C Si Mn Cu Cr Ni P S Fe

Low-carbon ferritic–pearlitic steel 0.10 0.02 0.45 0.04 0.04 0.02 0.019 0.018 Balance
Stainless steel 0.03 0.03 1.90 0.13 19.54 9.9 0.022 0.008 Balance

Table 2 (higher level of porosity, impurities, spattering, arc craters, etc.).

List of hardfacings with indication of main properties. Coatings were deposited in four layers (Fig. 1B). The substrate was
placed horizontally during welding. The fourth layer was partially
Grade Binder WC powder mean WC Hardness of Density,
designation grain size, mm content, coating, HV10 g cm
3 removed during grinding and polishing of the samples.
wt%
2.3. Characterisation
LC540/0 LC 540 0.1 306 7 17 8.0
LC540/7 LC 540 7.1 492 7 20 8.3
LC540/13 LC 540 13.3 604 7 40 8.6 Chemical composition of the welded samples was estimated by
LC540/20 LC 540 20.0 5517 31 8.9 Belec Compact-Lab-N optical emission spectrometer. The analysed spot
LC540/27 LC 540 26.6 634 7 24 9.2 size diameter was 8 mm. Measurements were performed 5–7 times.
LC540/32 LC 540 31.7 6477 51 9.5
The layer of material being 50–100 mm thick was removed by grinding
LC540/43 LC 540 42.7 690 7 70 10.1
SS540/0 SS 540 0.3 244 7 15 8.0 between the measurements to exclude the effect of inhomogeneous
SS540/9 SS 540 8.6 254 7 18 8.4 carbide distribution. The accuracy of measurement for highest content
SS540/13 SS 540 12.7 392 7 20 8.5 of carbon and tungsten was 0.04 and 0.10 wt%. The X-ray diffraction
SS540/18 SS 540 17.4 4177 17 8.7 (XRD) analysis was done by Bruker D8 with Cu K–alpha radiation X-ray
SS540/23 SS 540 22.8 4157 30 9.0
tube. Microstructural analysis was performed with the help of Hitachi
SS540/26 SS 540 26.2 493 7 25 9.2
SS540/35 SS 540 34.9 5117 21 9.7 S-3400N-II scanning electron microscope (SEM) equipped with energy
SS540/42 SS 540 41.9 592 7 22 10.1 dispersive X-ray spectroscopy (EDS) XFlash 5040 QUAD detector from
LC23/34 LC 23 33.5 8017 33 9.6 Bruker, Zeiss ULTRA 55 with the compositional contrast detector EbS,
LC68/37 LC 68 37.3 756 7 35 9.8
and by Zeiss EVO MA15 SEM supplied with EDS – INCA analyser.
LC135/37 LC 135 37.0 699 7 45 9.8
LC268/33 LC 268 32.8 706 7 43 9.5
Hardness of the coatings was determined with the help of
LC428/32 LC 428 31.9 623 7 51 9.5 Indentec 5030SKV hardness tester according to ISO 6507 (ASTM
LC605/36 LC 605 36.1 680 7 80 9.7 E384) with dwell time of 10 s. Measurements of the surface
hardness were repeated at least five times and the average values
LC – Low-carbon ferritic–pearlitic steel core wire; SS – austenitic stainless steel
along with deviations were stated throughout the paper. Hardness
core wire
of WC grains was evaluated by Buehler Micromet 2001 microhard-
ness tester with Vickers indenter and load of 0.49 N. Hardness of
Two sets of electrodes for low-carbon and stainless steel core wire 10 grains was determined and the range of values is provided.
were prepared (Table 2). The first set of electrodes was prepared with After welding and grinding the top and cross-sectional samples
varied WC content (0.1–42.7 wt% of WC) incorporating particles of the intended for microstructural investigation were polished by apply-
same size (540 mm).The second set of electrodes was prepared with ing SiC sand paper from P240 to P4000 using Phoenix 4000
varied WC grain size (23–605 mm)while keeping the same fixed (Buehler) sample preparation system.
reinforcement phase content (close to 35 wt% of WC). The final content Scratch testing of coatings was performed by sharp diamond
of WC and density of hardfacings is stated in Table 2. The values were Vickers indenter on CETR/Bruker UMT-2 tribometer with speed of
calculated according to the results of chemical analysis performed by 1 mm min
1 and constant load of 0.5 or 2.0 N. Scratch testing of
optical emission spectroscopy (content of tungsten, Table 4). carbide, matrix and the diffusion zone between the carbide and the
WC grains with average size of 540 mm were received from matrix was required to track the variation in hardness and the mode
Torez factory (Ukraine) and crushed by disintegrator DSL-350 and of fracture (ductile or brittle) of hardfacing. The sufficiently low load
DSL-175 (rough and fine crushing) produced at Tallinn University was selected to provide the narrow width of scratches comparable to
of Technology [34]. Powders were sieved by Fritsch Analysette the scale of the microstructure (size of carbides and dendrites).
3 into six fractions: 710–500, 500–355, 355–180, 180–90, 90–45 Wear testing of coatings was performed according to ASTM G65
and 45–0 mm. The corresponding average sizes (605, 428, 268, 135, test method, Procedure B by applying the MMTS device further
68 and 23 mm) are stated throughout the paper. Sample SEM described in [18,35]. Test conditions are given in Table 3. To
images of WC powder are given in Fig. 2. quantify the weight loss during the experiments, the specimens
were ultrasonically cleaned with acetone and weighed before and
2.2. Weld process after the tests to the nearest of 0.1 mg using GR-202, A&D
Instruments balance. Wear tests were repeated at least twice and
Substrate material for coating welding was S355 steel according the average values were stated throughout the paper. The value of
to EN 10025. The substrate was polished by SiC P60 coarse grit standard deviation of results was in most of the cases less
abrasive paper prior to coating application in order to clean the than75% (max. 712%) and is not shown on the graphs.
surface and to improve the adhesion between the hardfacing and the
substrate. Manual arc welding was done with the help of Fronius
inverter MagicWave 2600 with 115–125 A and 20–27 V direct 3. Results
current. Energy input was between 0.9 and 1.4 kJ mm
1. Higher
value of electric current was used for welding of thick electrodes 3.1. Microstructure and chemical composition of the hardfacings
with high WC content and for electrodes with stainless steel core
wire. Coatings made with stainless core wire and those with high WC Due to the droplet mode transfer of (1) steel from core wire and
content were hard to weld and the quality of the welds was lower (2) WC particles from flux covering during welding the carbides
 V. Jankauskas et al. / Wear 328-329 (2015) 378–390 381

Fig. 2. SEM images of as-received tungsten carbide grains (A), and those after crushing and sieving (B) with mean size of 605 mm and 68 mm respectively.

Table 3
Abrasive wear test conditions.

Parameter Description

Scheme Block-on-Wheel
Description of wheel Rubber lined steel wheel, diameter 228.6 mm, width 12.7 mm and hardness Shore A-60
Abrasive AFS 50/70 Ottawa silica sand (SiO2) with size 0.2–0.3 mm, HV0.05 ¼ 1100 and feed rate 370 g min
1
Circumferential velocity 2.4 m s
1
Linear abrasion 1436 m
Force again specimen 130 N
Atmosphere Air, 237 2 1C, relative humidity 45 75%

Table 4
Chemical composition of the coatings obtained by optical emission spectrometer (wt%).

Grade designation C W Si Mn Cr Ni

LC540/0 0.45 0.1 0.82 0.89 0.07 0.02

LC540/7 0.40 6.7 0.79 0.76 0.07 0.03
LC540/13 0.80 12.5 2.32 1.48 0.29 0.19
LC540/20 0.77 18.8 1.60 1.02 0.10 0.04
LC540/27 1.14 25.0 1.32 0.92 0.10 0.04
LC540/32 1.13 30.6 1.41 0.84 0.11 0.04
LC540/43 1.39 40.1 1.26 0.55 0.18 0.11
SS540/0 0.63 0.3 1.17 2.06 18.40 10.20
SS540/9 0.48 8.1 1.13 1.75 17.30 8.10
SS540/13 1.18 11.9 1.83 2.06 14.00 7.10
SS540/18 0.90 16.3 1.27 1.51 14.50 8.90
SS540/23 0.84 21.4 1.69 1.36 14.00 6.60
SS540/26 0.94 24.6 1.62 1.56 13.00 8.60
SS540/35 0.97 32.8 2.14 1.89 11.60 5.80
SS540/42 1.67 39.3 1.18 1.03 10.70 7.90
LC23/34 1.87 31.4 1.07 0.67 0.11 0.07
LC68/37 1.68 35.0 1.14 0.61 0.10 0.06
LC135/37 1.77 34.7 1.18 0.66 0.10 0.06
LC268/33 1.75 30.8 1.26 0.75 0.11 0.06
LC428/32 1.22 29.9 1.24 0.69 0.10 0.05
LC605/36 1.99 33.9 1.06 0.71 0.11 0.07

Fig. 3. SEM image of the cross-section of the sample SS540/35 (A); and top view of SS540/42 (B).
382 V. Jankauskas et al. / Wear 328-329 (2015) 378–390

Fig. 4. SEM images with indication of locations where EDS analysis was performed (see also Table 5). Samples shown in A, B, E and F are polished while samples shown in C,
D, G and H are after wear testing (indicated).
 V. Jankauskas et al. / Wear 328-329 (2015) 378–390 383

Table 5
Results of EDS analysis of polished and worn surfaces of coatings. Locations are depicted in Fig. 4.

Position Description Elements, wt%

C W Si Mn Cr Ni Fe

a1 Binder zone located close to carbide grain 2.5 22.9 0.4 1.1 – – 73.1
a2 Binder zone having a brighter appearance (more dendrites) 2.2 11.8 1.0 1.3 – – 83.7
a3 Binder zone having a darker appearance (less dendrites) 2.5 7.6 0.9 1.2 – – 87.8
a4 Bright and dark binder zones 2.3 10.0 0.9 1.2 – – 85.6
b1 Dark carbide grain zone (WC þC) 14.0 86.0 – – – – –
b2 Bright carbide grain zone (WC) 7.3 92.7 – – – – –
b3 Thin diffusion zone located close to carbide grain (Fe2W2Cþ C) 20.1 60.3 – 0.4 – – 19.2
b4 Thin diffusion zone located aside from carbide grain (Fe6W6C þC) 13.1 63.6 – 0.3 – – 23.0
b5 Dendrite 3.0 67.1 – 0.7 – – 29.2
b6 Matrix zone located close to carbide grain 3.3 15.2 0.8 0.9 – – 79.8
b7 Matrix zone located aside from carbide grain 3.2 15.1 0.6 1.2 – – 79.9
b8 Matrix zone located far aside from carbide grain 3.0 17.0 0.6 1.1 – – 78.3
b9 Zone including dendrite and matrix 2.8 34.4 0.1 0.9 – – 61.8
c1 Binder zone 3.8 10.7 1.3 1.4 – – 82.8
c2 Carbide grain (WC) 7.5 90.7 0.0 0.1 – – 1.7
d1 Dendrite with cracks 3.8 42.6 – 1.3 – – 52.3
d2 Dendrite 3.0 42.5 – 1.4 – – 53.1
d3 Matrix with a dark dot 3.0 10.4 – 1.0 – – 85.6
d4 Matrix 3.1 9.0 – 1.4 – – 86.5
e1 Binder zone located close to carbide grain 3.5 31.3 0.3 1.5 10.9 5.1 47.4
e2 Binder zone located far aside from carbide grain 3.3 32.0 0.0 1.3 8.4 4.2 50.8
e3 Large binder zone 3.0 30.7 0.2 1.3 8.7 4.2 51.9
f1 Bright carbide grain zone (WC þ C) 18.9 80.8 – – – – 0.3
f2 Dark carbide grain zone (WC þC) 16.4 83.3 – – – – 0.3
f3 Thin diffusion zone (Fe2W2C or Fe6W6C þC) 6.7 67.5 – 0.3 4.5 1.5 19.5
f4 Dendrite located close to carbide grain 4.2 66.4 – 0.4 5.9 1.6 21.5
f5 Dendrite located aside from carbide grain 13.8 58.1 – 0.4 5.2 1.5 21.0
f6 Thick dendrite grain 4.8 68.0 – 0.3 4.8 1.5 20.6
f7 Matrix zone located close to carbide grain 4.3 10.7 0.7 1.8 9.2 5.9 67.4
f8 Matrix zone located aside from carbide grain 3.7 8.7 0.9 2.0 8.9 6.1 69.7
f9 Zone with dendrite and matrix 3.5 33.4 – 1.1 7.4 4.5 50.1
g1 Binder zone 3.9 54.2 0.6 0.9 8.9 3.5 28.0
g2 Carbide grain (WC) 6.8 92.2 – – – 0.3 0.7
h1 Matrix 10.7 9.2 1.8 2.0 12.4 8.0 55.9
h2 Matrix with dark inclusions 6.8 13.8 1.5 2.7 22.1 4.8 48.3
h3 Thick dendrite grain 3.1 69.5 – 0.5 5.6 2.4 18.9
h4 Thick dendrite grain 3.7 69.0 – 0.6 5.7 2.5 18.5

are distributed unevenly and zones with varying WC concentra-
tions are present. Besides, the tungsten carbide particles have
higher density than binder and they tend to sink inside the molten
metal weld puddle. The formed structure of the coating is
consisting of alternating zones with high or low concentration of
WC particles inside the metal binder (Figs. 1B and 3).
The chemical composition of the coatings obtained by optical
emission spectrometer is given in Table 4.
The content (wt%) of phosphorus, sulphur and titanium was in
the range from 0.01 to 0.04, from 0.02 to 0.08 and from 0.02 to
0.06 respectively. The rest was iron.
Hardfacings LC540/0 and SS540/0 were prepared without the
addition of WC grains. Traces of tungsten found are caused by using
the same mixing device for preparation of all batches. This may also
explain the presence of chromium and nickel in the hardfacings made
with low-carbon steel core. The carbon content is higher for hard-
facings with higher tungsten. Deviations from the trend are explained
by the instability of manual welding and by the design of the
electrode. Increased (comparing to the core wires one, Table 1) content
of Si and Mn is due to minerals and ferroalloys of flux covering.
At least 10% of WC grains and sufficient amount of carbon were lost
during welding. This was calculated by comparing the initial concen-
tration during preparation of the electrodes and the actual content of
elements in hardfacing measured by optical emission spectroscopy.
X-ray diffraction results indicated that coatings with high
tungsten content have a high content of WC, W2C, double carbides
(Fe2W2C and Fe6W6C) and low content of Fe0.95(WC)0.05 solid
solution.
Macrohardness values (HV10) of coatings are higher for coat-
ings with high reinforcement content. Coatings welded by apply-
ing low-carbon steel core wire have higher hardness than those
welded with stainless steel core wire (valid for concentrations of
WC below 40 wt%). Microhardness (HV0.05) of the WC grains
measured was in the range from 1960 to 2440, which is similar to
results published by other researchers [2,8].
Results of EDS investigation of different characteristic zones are
given in Fig. 4 and Table 5. Grades LC540/13 and SS540/42 were
selected for investigation due to their low wear rates.
The compositions given in the description of Table 5 are selected
from those found during XRD measurements according to the ratios of
atomic weights of corresponding elements. The content of tungsten
decrease when the grain (b1, b2), diffusion zone (b3, b4) and matrix
(b6–b8) are compared. Binder located closer to carbide grain has
higher tungsten content (positions a1–a3, Table 5). Carbide grain has
zones with WC and those with excessive carbon WCþC (positions b1,
b2, c2, f1, f2 and g2). Diffusion zones are composed of double carbides
Fe2W2C or Fe6W6C. It is possible to suppose that Fe2W2C is located
close to carbide grain while Fe6W6C is located aside due to the
corresponding change in carbon content (positions b3, b4 and f3,
Table 5). Eutectic or eutectoid dendrites (having tree-like shape) of
double carbides have varying carbon content depending on position
(b5, f4–f6). Worn dendrites of low-carbon hardfacings have increased
content of iron (compare b5, d1 and d2, Table 5). Composition is
similar to zone b9 (dendrite and matrix, Table 5). Matrix of hardfacings
made with stainless steel core wire has Cr and Ni while it was usually
lower than in core wire (Table 1; f7 and f8, Table 5). The content of
384 V. Jankauskas et al. / Wear 328-329 (2015) 378–390

Fig. 5. SEM images of scratches performed on coatings with different loads (indicated).

chromium being sufficiently high in certain places points to the fact of c1; f8 and h1, Table 5) that may be explained by embedment of SiO2
possible formation of chromium carbides (zone h2, Table 5). The soft fragments. It is expected that matrix has Fe0.95(WC)0.05 solid solution
austenitic matrix has higher content of Si after wear (compare b8 and as it was determined by XRD investigation.
 V. Jankauskas et al. / Wear 328-329 (2015) 378–390
Fig. 6. Effect of WC reinforcement content (A) and grain size (B) on wear rate of composite coatings with low-carbon steel or stainless steel binder. The value of standard
deviation of results was in most of the cases less than7 5% (max. 712%) and is not shown on the graphs.
The diffusion zone between carbide grains and the matrix (Fig. 4B
and F) has varied width. It is evident that at least 20% of the WC
grains were completely dissolved during welding and fine dendrites
were formed (Fig. 4). The dissolution was more pronounced for
hardfacings with low WC content and in case of fine particles.
Tungsten carbide particles have good wetting by both binders;
there is a 2–10 mm wide diffusion zone between WC and binder
(Figs. 4B and F; 5D and E). Scratch test along the border between
the carbide and the binder has shown that the diffusion zone has
sufficient hardness and strength (Fig. 5D). There is no abrupt
change of scratch width during crossing the diffusion zone. The
material pushed to the edge of the scratch has limited rupture
indicating that there is considerable bonding between phases
(Fig. 5). The scratches done in other locations show also limited
rupture of the edges. The highest variations in shape of the scratch
were observed for hardfacings with stainless binder (Fig. 5G). The
abnormal fracturing of some carbides is illustrated in Fig. 5E.
3.2. Wear testing
The results of wear testing are given in Fig. 6. The corresponding
polynomial trendlines with equations and coefficients of determi-
nation are given. There is a correlation between the WC content and
the wear rate of coatings. These correlations are valid for both types
of coatings with mild steel and stainless steel binders.
Hardfacings with higher reinforcement content have higher
hardness (Table 2) and better wear resistance (Fig. 6A). There is an
optimum WC grain size (Fig. 6B) close to 0.14 mm (135 mm). The
low measured wear rate of hardfacings made from WC powders
with an average grain size of 0.6 mm (605 mm) is due to limitation
of test method applied. The sand with size of 0.2–0.3 mm used for
testing is not able to cut matrix sufficiently deep to remove the WC
grains and protruding grains deteriorate the rubber wheel. The
grooves formed on the surface of wheel reduce the pressure to be
supplied to the abrasive particles and the intensity of abrasive
cutting around the large WC grains decreases.
The characteristic features of the wear mechanism of studied
hardfacings is very similar to observations by other researchers found
in literature (Fig. 7). Material without the WC grains addition (LC540/
0, Fig. 7A) is worn due to scratching, grooving and indenting.
Harfacings with WC addition are worn preferentially (Fig. 7B, C and
D). Primary and precipitated carbides are worn much less than matrix
and provide protection until they become vulnerable to removal by
fracture or complete detachment (Fig. 7F and G). Silica and WC
fragments are embedded into the binder between the carbide grains
385
(Fig. 7E). These fragments, having same or higher hardness than
abrasive particles, improve the wear resistance. Conglomerates of WC
grains results in uneven distribution of hard faces and lowered wear
resistance (Fig. 7H). This is due to the improper carbide retention
leading to their fracturing and due to the absence of carbides in other
areas leading to increased removal of binder.
Tribolayers were found on most of the hardfacings between
primary carbide grains (Fig. 8). It is often identified as white-etching
layer formed due to melting of the test sample [20]. The thickness of
these layers reaches 40 mm (Fig. 8F) but in most of the cases it was in
the range from 1 to 15 mm. The thickness of the layer varied across the
width of the wear scar. It is possible to suppose that formation of
tribolayer is cyclic process, including formation, growth and fracturing
similar to that observed for other coating type [36]. An additional
thorough investigation is required to determine the mechanical
properties of tribolayers to provide conclusions [20]. Tribolayer may
serve as dumping element reducing stresses transferred from abrasive
particles to primary hardfacing. Microcracking of dendrites down to
the depth of 40 mm was observed below areas without tribolayer (Fig.
8G).
The defects, which are significantly reducing the integrity and
the wear resistance of hardfacings are given in Fig. 9. These
include incomplete fusion in overlay regions between weld beads
(Fig. 9A), large scale cross macrocracks through the whole hard-
facing (Fig. 9B), poor bonding between carbide grains and binder
leading to their complete or partial detachment (Fig. 9C and D),
pores (Fig. 9C), cracking of elongated carbides and dendrites
during loading (Figs. 8G and 9E), carbide grain microfracturing
during preparation (crushing) or coating stages (thermal stresses)
and subsequent loss of fine fragments (Fig. 9F), slag inclusions
(Fig. 9G), dissolution of carbide grains (Fig. 9H).
4. Discussion
The distribution of WC grains throughout the weld hardfacing is
uneven because of the application method used (Figs. 1B, 3). This is
due to transfer of weld constituents from different sources (steel core
and the covering) and due to instability caused by manual mode of
welding. This will result in uneven wear of coating in real applica-
tions. The wear of zones with high content of the WC will take longer
than that of zones with lower hard phase content. The temperature
and cooling rate throughout the weld bead differs that results in
local variations in type of steel structure formed [37]. However, as it
was shown in [20] the initial structure and hardness of steel
386 V. Jankauskas et al. / Wear 328-329 (2015) 378–390

Fig. 7. SEM images of typical wear features of hardfacings.

obtained by thermal treatment is not influencing much the wear rate
while the surface layer formed plays a major role. Similar observa-
tions were true for ceramic–metal composites [10,21]. Tribolayer was
found on top of investigated hardfacings with lowest and highest
wear resistance (LC540/0 and LC540/43, Fig. 8). It was found in both
groups of hardfacings made with low-carbon and stain-
less steel core wire. Almost no damage was found for the original
structure below tribolayer. The tribolayer has cracks that indicates
that it may experience spallation.
The dissolution of WC particles in arc welding results in lowering
the toughness and wear resistance of deposits down to 50% if
compared to gas welded overlay [17]. It is required to reduce the
 V. Jankauskas et al. / Wear 328-329 (2015) 378–390 387

Fig. 8. Cross-sectional SEM images of hardfacings after wear testing.

heat input and the length of arc to cut the loss of WC grains [38]. The
welder should be informed about this prior to the welding process.
The results presented in Fig. 4 and Table 5 indicate that grains are
composed of WC and WC with excessive (free) carbon (zones b2 and
b1). There are only some places corresponding to WC with excessive
carbon according to EDS analysis and the appropriate raw material
should be used to avoid the presence of such a phase. The zone of
WC with excessive carbon content is detrimental since it results in
lowering of mechanical and wear properties [39]. The behaviour of
carbide having such zones during single scratching is illustrated in
Fig. 5E. In contrast to W2C/WC, the carbide grain of WC with
excessive carbon can be easily broken apart (compare Fig. 5D and E).
Addition of WC up to 13 wt% into a stainless steel matrix is
unfavourable while addition of 42 wt% of WC results in reduction of
388 V. Jankauskas et al. / Wear 328-329 (2015) 378–390

Fig. 9. Typical defects influencing the wear resistance of hardfacings.

wear rates by as much a factor of 9 (Fig. 6A). This might be attributed
to lower hardness of stainless steel matrix that is not able to provide
required protection and support to the WC grains for hardfacings
with low binder content. The reduction of wear rates of hardfacings
with WC content higher than 13 wt% is due to lowering of distance
between carbide grains and providing protection against binder
abrasion. The change in size of abrasive particles will probably lead
to shifting of the WC threshold content. Higher (442 wt%) WC
concentration is difficult to achieve by proposed type of electrodes.
Hardfacings made with low-carbon matrix exhibit reduction in wear
 V. Jankauskas et al. / Wear 328-329 (2015) 378–390
rate by a factor of 2 even with minor addition of WC (13 wt%) while
further increase in reinforcement content is not so efficient since it is
failing to offer the sufficient rise of hardness of hardfacing (Table 2).
The hardfacings welded using electrodes with smaller WC grain
size have higher hardness as it is predicted in literature [4,13].
However the lowest wear rate (in the range of carbide grain sizes
from 23 to 428 mm) was exhibited by hardfacings with intermediate
grain size (135 mm). Intermediate size of carbides might be optimum
since carbides of this size are less suffering from excessive dissolu-
tion as finer fractions do and, besides, they are less prone to thermal
cracking experienced by coarser fractions. Removal of WC grains
takes sufficiently longer in low stress wear process by soft (hardness
of silica is lower than that of WC) rounded coarse abrasive. The
removal of coarse WC grains with size of 605 mm was almost
impossible to fulfil by sand particles with a size of 200–300 mm
that resulted in deterioration of the rubber wheel and low measured
wear rate of these hardfacings (Fig. 6B).
As it is demonstrated by Fig. 5 the different zones of hardfacing
behave sufficiently well during single scratching indicating good
bonding between phases and limited brittle fracturing required for
high wear resistance. The exception was discussed earlier (Fig. 5E).
The coatings with sufficient wear resistance (Fig. 6) and, presum-
ably, corrosion resistance (due to stainless steel core wire used)
may be used in conditions of low stress abrasive wear for
reinforcement of soil-engaging tools. The best industrial nonfer-
rous hardfacings tested by the same test method are having the
similar wear rate (10 mm3) [3]. The further optimisation of the
electrode, testing in high stress and field conditions are required.
The performance of hardfacings could be improved and the cost of
the electrodes could be reduced by application of the recycled
cermet composite particles instead of WC [30,31].
5. Conclusions
The effects of three variables, namely (1) binder material (low-
carbon ferritic–pearlitic or austenitic stainless steel), (2) WC con-
tent, and (3) WC grain size, on three-body abrasive wear resistance
of hardfacings, were evaluated using the ASTM G65, dry sand/
rubber wheel, test method (Procedure B).
1. A reduction in wear rate by as much as a factor of 9 was
achieved by compositional modification.
2. The hardness and wear resistance (valid for content of WC
below 40 wt%) of hardfacings made with stainless steel core
wire is lower if compared to low-carbon coatings of the same
WC content.
3. The presence of 13–20 wt% of WC in hardfacing made with
low-carbon steel core electrodes enables to halve the wear rate.
Further increase in reinforcement content is less cost-efficient.
4. The content of WC in hardfacing made with stainless steel core
wire lower than 42 wt% is not cost-efficient and below 13 wt%
it is even detrimental.
5. The hardfacings produced with mean WC grain size of 135 mm
have the lowest wear rate among the studied materials with a
range of sizes from 23 to 428 mm.
6. More efforts are required to provide homogeneous distribution
of WC, reduce its dissolution and formation of weak detri-
mental phases.
Acknowledgements
This work was supported by Institutional Research Funding IUT
(19–29, M.Antonov) of the Estonian Ministry of Education and Research
389
and by Estonian Science Foundation, Estonia (Grant 8850, M.Antonov).
The authors wish to thank M.Viljus and V.Mikli from Centre for
Materials Research of Tallinn University of Technology for help with
sample analysis.
References
[1] R.L. Kushwaha, J. Shi, Investigation of wear of agricultural tillage tools, Lubr.
Eng. 47/3 (1989) 219–222.
[2] P.F. Mendez, N. Barnes, K. Bell, S.D. Borle, S.S. Gajapathi, S.D. Guest, H. Izadi,
A.K. Gol, G. Wood, Welding processes for wear resistant overlays, J. Manuf.
Process. 16 (2014) 4–25.
[3] J.R. Davis, Hardfacing, weld cladding, and dissimilar metal joining, in:
D.L. Olson, T.A. Siewert, S. Liu, G.R. Edwards (Eds.), ASM handbook : Welding,
Brazing, and Soldering, vol. 6, ASM International, 1993, pp. 789–829.
[4] Hardmetals, in: V.K. Sarin (Ed.), Comprehensive Hard Materials, vol. 1, Elsevier,
2014.
[5] G.S. Upadhyaya, Cemented Tungsten Carbides, Production, Properties and
Testing, Noyes Publications, USA, 1998.
[6] C.M. Fernandes, A.M.R. Senos, Cemented carbide phase diagrams: a review,
Int. J. Refract. Met. Hard Mater. 29 (2011) 405–418.
[7] J.M. Tarragó, C. Ferrari, B. Reig, D. Coureaux, L. Schneider, L. Llanes, Mechanics
and mechanisms of fatigue in a WC–Ni hardmetal and a comparative study
with respect to WC–Co hardmetals, Int. J. Fatigue 70 (2015) 252–257.
[8] S. Gnyusov, S. Tarasov, Y.u. Ivanov, V. Rothstein, The effect of pulsed electron
beam melting on microstructure, friction and wear of WC–Hadfield steel hard
metal, Wear 257 (2004) 97–103.
́
[9] J. Rodrıguez, ́ R. Fernández, J.E. Fernández, An experimental study of
A. Martın,
the wear performance of NiCrBSi thermal spray coatings, Wear 255 (2003)
950–955.
[10] M. Antonov, I. Hussainova, J. Pirso, O. Volobueva, Assessment of mechanically
mixed layer developed during high temperature erosion of cermets, Wear 263
(2007) 878–886.
[11] A.P.I. Popoola, Hardness, microstructure and corrosion behaviour of WC-9Co-
4Cr þ TiC reinforced stainless steel, Int. J. Electrochem. Sci. 9 (2014) 1273–1285.
[12] Kennametal Specialty Carbide Products, A93-350(10)A5, available online:
〈www.kennametal.com〉 (12/11/2014).
[13] Cemented Carbide, Sandvik New Developments and Applications, Sandvik,
H-9116 Eng, 2005, available online: 〈www.hardmaterials.sandvik.com〉 (12/11/
2014).
[14] P. Chivavibul, M. Watanabe, S. Kuroda, K. Shinoda, Effects of carbide size and
Co content on the microstructure and mechanical properties of HVOF-sprayed
WC–Co coatings, Surf. Coat. Technol. 202 (2007) 509–521.
[15] Hardfacing Alloys, Kennametal Stellite, B-12-02851, available online: 〈www.
kennametal.com/stellite〉 (12/11/2014).
[16] EutecTrode, High Performance Manual Metal Arc Electrodes for Repair, Join-
ing, Rebuilding and Wearfacing Applications, Castolin Eutectic, available
online: 〈www.castolin.com〉 (12/11/2014).
[17] R. Chattopadhyay, Advanced Thermally Assisted Surface Engineering Pro-
cesses, Kluwer Academic Publishers, USA, 2004.
[18] ASTM G65-04 (2010), Standard Test Method for Measuring Abrasion Using the
Dry Sand/Rubber Wheel Apparatus, ASTM, 2010.
[19] K.G. Budinski, Guide to Friction, Wear, and Erosion Testing, ASTM Interna-
tional, USA, 2007.
[20] S. Das Bakshi, P.H. Shipway, H.K.D.H. Bhadeshia, Three-body abrasive wear of
fine pearlite, nanostructured bainite and martensite, Wear 308 (2013) 46–53.
[21] M. Antonov, I. Hussainova, Cermets surface transformation under erosive and
abrasive wear, Tribol. Int. 43 (2010) 1566–1575.
[22] M.G. Gee, A. Gant, B. Roebuck, Wear mechanisms in abrasion and erosion of
WC/Co and related hardmetals, Wear 263 (2007) 137–147.
[23] K.H. Zum Gahr, Wear by hard particles, Tribol. Int. 31 (1998) 587–596.
[24] R.J.K. Wood, Tribology of thermal sprayed WC–Co coatings, Int. J. Refract. Met.
Hard Mater. 28 (2010) 82–94.
[25] N.B. Dube, I.M. Hutchings, Influence of particle fracture in the high-stress and
low-stress abrasive wear of steel, Wear 233–235 (1999) 246–256.
[26] H. Chen, I.M. Hutchings, Abrasive wear resistance of plasma-sprayed tungsten
carbide–cobalt coatings, Surf. Coat. Technol. 107 (1998) 106–114.
[27] V.E. Buchanan, P.H. Shipway, D.G. McCartney, Microstructure and abrasive
wear behaviour of shielded metal arc welding hardfacings used in the
sugarcane industry, Wear 263 (2007) 99–110.
[28] S.M. Nahvi, P.H. Shipway, D.G. McCartney, Particle motion and modes of wear
in the dry sand–rubber wheel abrasion test, Wear 267 (2009) 2083–2091.
[29] C. Katsich, E. Badisch, Effect of carbide degradation in a Ni-based hardfacing
under abrasive and combined impact/abrasive conditions, Surf. Coat. Technol.
206 (2011) 1062–1068.
[30] A. Zikin Ph.D. thesis, Advanced Multiphase Tribo-Functional PTA Hardfacings,
TUT Press, Estonia, 2013 (12/11/2014).
[31] A. Zikin, M. Antonov, I. Hussainova, L. Katona, A. Gavrilović, High temperature
wear of cermet particle reinforced NiCrBSi hardfacings, Tribol. Int. 68 (2013)
45–55.
[32] Description of the company, available online: 〈http://www.varis.lt/index.php?
mln=en&id=8&ids=9〉 (12/11/2014).
390 V. Jankauskas et al. / Wear 328-329 (2015) 378–390

[33] C.M. Fernandes, A.M.R. Senos, M.T. Vieira, Control of eta carbide formation in
tungsten carbide powders sputter-coated with (Fe/Ni/Cr), Int. J. Refract. Met.
Hard Mater. 25 (2007) 310–317.
[34] D. Goljandin Ph.D. thesis, Disintegrator Milling System Development and
Milling Technologies of Different Materials, TUT Press, Estonia, 2013 12/11/
2014.
[35] M. Antonov, I. Hussainova, R. Veinthal, J. Pirso, Effect of temperature and load
on three-body abrasion of cermets and steel, Tribol. Int. 46 (2012) 261–268.
[36] M. Antonov, I. Hussainova, E. Adoberg, Effect of loading system inertia on
tribological behaviour of ceramic–ceramic, ceramic–metal and metal–metal
dry sliding contacts, Trib. Int. 65 (2013) 207–214.
[37] J. Hornung, A. Zikin, K. Pichelbauer, M. Kalin, M. Kirchgaßner, Influence of
cooling speed on the microstructure and wear behaviour of hypereutectic
Fe–Cr–C hardfacings, Mater. Sci. Eng. A 576 (2013) 243–251.
[38] A. Hasui, O. Morigaki, Welding and Powdering, Mashinostroyenie, Moscow,
1985 (in Russian. Originally published in Japanese).
[39] A.S. Kurlov, A.I. Gusev, Tungsten carbides and W–C phase diagram, Inorg.
Mater. 42 (2006) 121–127.