Effect of Microstructure On Edge Wear Mechanisms in WC-Co
Effect of Microstructure On Edge Wear Mechanisms in WC-Co
Effect of Microstructure On Edge Wear Mechanisms in WC-Co
www.elsevier.com/locate/ijrmhm
Abstract
Edge wear of nano to coarse WC–Co grades was investigated under two-body abrasion conditions using SiC abrasive at 2 and 15 N
loads. With the Wne abrasive utilized, a transition from homogeneous plowing in submicron grades to inhomogeneous grains pullout in
coarse hardmetals was observed at the beginning stages. At Wnal stages, plowing was detected in all the grades. When using the coarse
abrasive, for all grades, Xaking at the beginning and plowing at the Wnal stages occurred. The performed numerical Wnite element simula-
tion of the microstructural mechanical response showed good agreement to experimental results and was discussed in relation to the oper-
ative wear mechanisms observed. The found results demonstrated a remarkable inXuence of microstructure and abrasive size on the
heterogeneous to homogeneous wear mechanisms transition at initial stages.
© 2006 Elsevier Ltd. All rights reserved.
0263-4368/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijrmhm.2006.04.004
172 P.V. Krakhmalev et al. / International Journal of Refractory Metals & Hard Materials 25 (2007) 171–178
and fragmentation of the surface, while softer ZrO2, SiC or excludes possible eVects of adhesion or chemical interac-
SiO2 abrasives caused preferable binder removal followed tions, which might complicate identiWcation of the abrasive
by detachment of WC grains [5,7]. It has also been demon- wear mechanisms. Wear mechanisms and volumetric wear,
strated that coarser materials exhibited predominantly brit- were investigated and discussed in dependence on the slid-
tle response whereas plastic deformation has been observed ing distance and normal force. The main attention was
in submicron materials [5,7]. It is important that in all those devoted to identiWcation of wear micromechanisms and the
research, the size of abrasive exceeded microstructural fea- correlation to microstructural response.
tures of WC–Co composite signiWcantly.
In a contradiction, if 1–10 m diamond, SiC and Al2O3 2. Materials and methods
abrasives have been used, a decrease in wear rates with
increase of the WC size has been observed recently, Totally, six WC–Co hardmetals of diVerent WC grain
although, the same wear micromechanisms have been size were investigated. The selected grades found industrial
detected [8,9]. With the hard diamond, binder and WC applications as cutting, drilling or milling tools of wide
grains are worn with the same rate, while binder phase materials range. The HM1-C and HM2-M grades are
removal followed by pullout of undermined WC particles is applied for woodcutting of soft and hard or frozen wood,
the main wear mechanisms if SiC or Al2O3 are utilized. the HM3-F grade is for sawing of laminated particle
Such a behavior has been explained by an increase in resis- boards, the HM4-EF grade is typical for medium or high
tance to carbide pullout as the size of carbide increases, due density Wberboard and, the HM5-UF and HM6-N grades
to increase in a depth of the binder needed to be removed. applied for drilling, milling and routing of printed circuit
For materials selection and cutting tool design, wear boards and glass reinforced plastics. Selected properties of
resistance is the vital property, determining tool perfor- the investigated materials are collected in Table 1.
mance and costs for manufacturing and exploitation. The specimens were ground and polished from two sides
Although there are the ASTM G65 and ASTM B611-85 with 3 m diamond paste to produce mirror-Wnished 60°
standards for the abrasive wear evaluation, these test meth- sharp edge. As an example, Wnal edge sharpness of the
ods usually employ coarse abrasive and do not represent HM2-M specimen is present in Fig. 1a. Polished specimens
the geometry and parameters of the real cutting or drilling were mounted onto the vertically movable holder, built on
operations. This disagreement resulted in recent interest in the drill machine. A normal force of 2 and 15 N was applied
tribological phenomena at the edge of a tool that is impor- by a static weight. The abrasive paper was Wxed on a hori-
tant practically, since sharpness loss inXuences dramatically zontal movable X–Y-table, which was drawn along the
the tool performance and Wnal quality of the product. loaded specimen manually, using worm screwfeed, at slid-
To characterize the edge fracture resistance and evaluate ing speed of about 0.5–1.0 cm/s. The speed was measured in
fracture toughness of hardmetals, a tribotester for investiga- advance and not currently controlled during the sliding. A
tion of the edge wear on prismatic specimens has been sketch of the specimen arrangement and sliding geometry is
designed by Scieszka [10,11]. It has been found that initial present in Fig. 1b. Long sliding distances were collected by
stages are controlled by brittle fracture whereas microabra- a number of strokes of about 15 cm each until required slid-
sive wear mechanisms dominated the steady-stage wear [10]. ing distance was reached. Every new stroke drew along the
Furthermore, the edge Xaking tendency has been examined fresh abrasive paper surface. The 4000 (»5 m grit size) and
using an indentation method. Good correlation between 80 (»200 m grit size) abrasive papers were used. The tests
fracture toughness and so-called edge toughness has been were performed without lubricant.
found for brittle materials including hardmetals [12–14]. In The tested specimens were examined in a GEMINI LEO
both approaches described above, the size of abrasive parti- 1530 scanning electron microscope (SEM) to explore the
cles or size of the indenter exceeded considerably features of operative wear mechanisms. The volumetric wear for speci-
the microstructure and therefore, micro-scale edge behavior mens slid along the Wne abrasive paper at 2 N load was cal-
still has not been completely discovered. culated from the mean of 10–15 measurements of wear
In the present research work, abrasive wear of the wedge width (the 2 parameter in Fig. 1b), measured in the SEM.
specimens of WC–Co hardmetals from nano to coarse The coarse abrasive produced very rough and uneven worn
grades, was investigated under two-body abrasion condi- surface so that calculations of volumetric wear were
tions. The specimens sharpened and polished to 60° edge assumed inaccurate.
sharpness slid along the SiC abrasive paper in a direction To analyze the eVect of microstructure on the strain dis-
perpendicular to the edge line. Such geometry was selected tribution in the wedge specimen at the Wne abrasive and 2 N
to model a situation when the edge meets abrasive particles normal load test conditions, Wnite element simulations were
embedded into the substrate (for instance inclusions in cut- conducted with Abaqus® software. The contact situation
ting timber, drilling of hard particle reinforced composite was interpreted as a simultaneous multiasperity contact
material etc.). The disadvantage of this test arrangement is between specimen side and abrasive and was simpliWed as a
that the specimen does not penetrate the substrate as in the distributed pressure applied from the side normally to the
real situation. Nevertheless, such a scheme describes con- side surface of the specimen. The applied pressure value of
tact between the edge and particles appropriately and 500 MPa was calculated so that a component normal to the
P.V. Krakhmalev et al. / International Journal of Refractory Metals & Hard Materials 25 (2007) 171–178 173
Table 1
Selected properties of the WC–Co hardmetals investigated
HM1-C HM2-M HM3-F HM4-EF HM5-UF HM6-N
Coarse Medium Fine Extra Wne Ultra Wne Nano
Grain size, m 3.5–4.9 1.4–2.0 1.0–1.3 0.5–0.9 0.3–0.5 0.1–0.3
Co, wt.% 11 11 6 3 8 9
Hardness, HV30 1150 1280 1620 1900 1900 1950
Fracture toughness, MPa/m3/2 19.0 16.0 11.0 7.0 8.5 8.0
Transverse rupture strength, N/mm2 3200 3200 2800 2100 4300 4700
Width (1), mm 4.1 4.3 5.5 4.6 3.3 3.3
Fig. 1. Edge sharpness of the polished HM2-M specimen in front view (a), and a sketch of the specimen arrangement and sliding geometry (b).
abrasive surface would satisfy the 2 N normal load test con- 3. Results
ditions. The contact area was approximated as 1 m in
height. The plane strain conditions were assumed to prevail 3.1. Wear diagram
in the middle of the specimen, since the thickness of the
specimen exceeds other dimensions signiWcantly. As the A diagram summarizing results of worn volume in
boundary conditions, the nodes located at the top surface dependence on sliding distances using 4000 abrasive grit
of the specimen, Fig. 7, were restricted from movement dur- and 2 N normal load test conditions is present in Fig. 2.
ing the analysis. Generally, materials demonstrate similar behavior with the
Two model microstructures of WC–11%wt.Co compos- highest wear rate at the beginning, which further decreases
ites, one corresponding to the ultraWne and another to the with increase in sliding distance. Nevertheless, as seen in the
coarse grades, were used, Fig. 7. The microstructures used close view in Fig. 2b, if for the ultraWne HM5-UF and
in the analysis represented a 2D cross-section of the 3D nanoscale HM6-N composites the diagrams are described
specimen in this simulation. The Co and WC phases were with quite smooth curves, the diagrams for the HM1-C,
assumed to behave isotropic elastic with perfect phase HM2-M, HM3-F and HM4-EF materials look like a com-
interface, i.e. no separation along the Co/WC interface was bination of two regions with a transition after about 30–
allowed. 60 cm sliding. Such a behavior is well pronounced in the
Meshes and Abaqus® input Wles were created based on HM1-C, HM3-F and HM4-UF materials but not in the
micrographs of the materials through the “ppm2oof”and HM2-M grade. In the further discussion, the initial and
“oof2abaqus” software, respectively, developed at the the Wnal stage terms are used to describe the corresponding
National Institute of Standards Technology (NIST) [15]. portions of the wear diagram.
The diVerent phases were identiWed as separate element
groups, with signiWcantly reWned meshes in the Co phase, 3.2. Wear mechanisms
and phase boundaries. Totally, 3.0 £ 105 and 1.1 £ 105 plain
strain 6-node triangular elements were used to represent the Three regimes of the two-body abrasion were run: (i) 2 N
ultraWne and the coarse materials, respectively. The princi- load and Wne abrasive (4000 grit), (ii) 2 N load and coarse
pal strain distribution maps were calculated on the cross- abrasive (80 grit) and (iii) 15 N load and coarse abrasive (80
section parallel to the sliding direction and normal to the grit). Worn surfaces were examined in the SEM to identify
abrasive surface. The microstructures used for the calcula- wear mechanisms operative at each stage.
tions as well as strain distribution maps are present in The worn surfaces investigated after sliding under 2 N
Fig. 7. load and Wne abrasive test conditions in the coarse and
174 P.V. Krakhmalev et al. / International Journal of Refractory Metals & Hard Materials 25 (2007) 171–178
Fig. 2. Worn volume vs. sliding distance diagram under 2 N and Wne abrasive test conditions (a) and, close view of the beginning part of the diagram (b).
nano grades are illustrated in Fig. 3. It is seen that observed grooving indicated predominance of the plowing wear
wear mechanisms are not the same. In the coarse HM1-C mechanisms, although some indications of possible falling
material, at the initial stage, pullout and detachment of WC out of WC grain due to lost of the matrix support may be
grains were observed after 2 cm sliding, Fig. 3a. Further observed. Fig. 4b illustrates contact surface of the HM1-C
sliding led to the binder removal and some scratching of the grade after 30 cm sliding that corresponds to the deXection
WC grains, Fig. 3b. Surface plastic deformation and exten- on the wear diagram. A simultaneous operating of the WC
sive grooving wear were indicated in Fig. 3c and d illustrat- grain fracturing and detachment, binder deformation and
ing the worn surface observed after 2 and 200 cm sliding in removal and some grooving wear mechanisms observed
the nanoscale HM6-N grade. It is seen that plowing is the clearly.
main wear mechanism operative in the nanoscale grades at Fig. 5 illustrates worn surfaces after sliding at 2 N and
all sliding distances. coarse abrasive test conditions. It is seen that utilization of
The close view of the worn surface observed in the the coarse abrasive led to initiation of plowing and chip-
HM6-N grade after 2 cm is present in Fig. 4a. Extensive ping in both the coarse HM1-C and nano HM6-N materi-
Fig. 3. Worn surface observed at the initial and Wnal stages in the HM1-C (a) and (b), and in the HM6-N (c) and (d) grades under 2 N and Wne abrasive test
conditions, respectively.
P.V. Krakhmalev et al. / International Journal of Refractory Metals & Hard Materials 25 (2007) 171–178 175
Fig. 7. Microstructures of model WC–11 wt.%Co alloys used for simulation of the microstructure response in submicron (a) and micron (b) grades and,
maximum principal strain distributions calculated for the submicron (c) and micron (d) microstructures. Note strain localization in the binder phase in the
micron grade and bulk strain distribution in the submicron grade.
Utilization of the coarse 80 grit abrasive paper resulted the Xaking wear mechanism, activated at the beginning, is
in initiation of chipping (Xaking) wear mechanism at the changed to the plowing at the Wnal stages, Figs. 5 and 6.
beginning stages. This wear mechanism caused formation At the Wnal stages, more than few grains of WC are in a
of the unevenly worn surface and, therefore, diYculties in contact to the abrasive paper. The wear behavior at this
quantitative analysis of the volumetric wear. Generally, as stage generally depends on WC grain size, abrasive size and
seen in Figs. 5 and 6, chipping was observed at 2 and 15 N abrasive hardness. It has been reported that for coarse
loads at the beginning stages and more pronounced for abrasives, homogeneous plastic deformation and plowing is
higher load test conditions. Further sliding led to predomi- the most common wear mechanisms, while if the size of the
nance of plowing wear and, only minor chipping was asperity is comparable to the microstructure, selective
observed after 100 cm sliding at 2 and 15 N load. Such a removal of the binder phase is observed [8,9,16].
behavior was not sensitive to the WC grain size because the Results obtained in the present research for Wnal stage
interaction zone exceeds signiWcantly features of the micro- wear mechanisms are in a good agreement to these consid-
structure and, was observed in the nano and coarse grades. erations. It is seen in Fig. 3b, that under 2 N and Wne abra-
The Xaking phenomena observed in the present research sive test conditions, removal of binder and some scratching
can be understood using a concept of the edge chipping ini- of the WC grains is observed in the HM1-C hardmetal,
tiation in brittle materials developed in [12–14]. It has been while evidences of plastic deformation and plowing are
shown that there is a linear dependence between an indenta- observed in the same material slid under 15 N and coarse
tion load critical for Xaking initiation and a distance from abrasive test conditions, Fig. 6. For the submicron HM6-N
the edge. Additionally, initiation of Xaking has been found of HM5-UF grades, even the Wne abrasive utilized was
dependent on the edge geometry – the critical Xaking loads remarkably larger than the features of microstructure and
have been found higher for specimens with a chamfer com- therefore, microstructure response was similar at all the test
pared to sharp specimens, even the edge toughness value conditions. The plowing was found the main wear mecha-
remained the same [12,13]. In the present research, the nism operative at the Wnal stage for 2 and 15 N normal
rounded surface occurred due to wear apparently plays a loads in submicron grades.
role of the chamfer. As the edge is worn, the critical load Discussing the overall wear mechanisms, it is diYcult to
value required for Xaking initiation increased and, therefore, link the edge wear resistance directly to the mechanical
178 P.V. Krakhmalev et al. / International Journal of Refractory Metals & Hard Materials 25 (2007) 171–178
properties (collected in Table 1). Nevertheless, obviously wear initiated at the beginning stages led to quick sharp-
the microstructure response vitally inXuenced the wear ness loss and formation of the rough and uneven edge.
mechanisms and wear rates at the initial stages. If the fea- The further sliding resulted in a transformation of Xak-
tures of microstructure are comparable to the size of abra- ing to plowing wear in submicron and micron grades.
sive particles, inhomogeneous pullout of the WC grains was
observed. In nanoscale materials, microstructure is remark- Acknowledgements
ably Wner than the abrasive and therefore, plowing was the
dominant wear mechanism at all sliding distances. Numeri- Authors would like to thank Mr. Mathias Tillman and
cal simulation showed good correlation to the experimental Mr. Stefan Ederyd, Sandvik Hard Materials, for materials
results although experimentally tested grades had diVerent supply and fruitful discussion and analysis of the obtained
cobalt content. Nevertheless, systematic investigations of results. Also, authors are grateful to Mr. Nils Hallbäck,
the edge wear in dependence on cobalt content and WC Karlstad University, for productive discussions on numeri-
grain size are required for better understanding of the edge cal simulation and Mr. Prasad Angaru, for active participa-
wear phenomena. The found results demonstrated a tion in the experimental part of the research.
remarkable inXuence of microstructure and abrasive size on
the heterogeneous to homogeneous wear mechanisms tran- References
sition at the initial stages of wear that facilitate understand-
ing of the real tool edge wear behavior. [1] Spriggs GE. A history of Wne grained hardmetal. Int J Refract Met
Hard Mater 1995;13:241–55.
[2] Brookes K. Some tribulation on the way to a nano future for hard-
5. Conclusions metals. Metal Powder Report 2005;60:24–30.
[3] Grearson A, James M, Tillman M, Norgren S, GustaVson P. The
Edge wear in sharpened wedge WC–Co specimens (0.1– future of Wner grain hard metals. In: Kneringer G, Rödhammer P,
0.3 to 3.5–4.9 m WC grain size) were investigated under Wildner H, editors. Proceedings of the 16th international plansee sem-
inar. Reutte, Austria: PLANSEE Holding AG; 2005. p. 314–38.
dry two-body abrasion conditions. The main obtained [4] Nordström J. On wear of wood-cutting tools. PhD thesis, Karlstad
results can be summarized as follow: University, Karlstad, Sweden, 2004.
[5] Fischer TE, Jia K. Abrasion resistance of nanostructured and conven-
• In case of the Wne abrasive (4000 grit) and low 2 N load tional cemented carbides. Wear 1996;200:206–14.
test conditions, the microstructure response inXuenced [6] O’Quigley DGF, Luyckx S, James MN. An empirical ranking of a
wide range of WC–Co grades in terms of their abrasion resistance
the wear mechanisms signiWcantly. The coarse grades measured by the ASTM standard B 611-85 test. Int J Refract Met
demonstrated inhomogeneous behavior with pullout of Hard Mater 1997;15:73–9.
WC grains from the edge, followed by binder removal [7] Allen C, Sheen M, Williams J, Pugsley VA. The wear of ultraWne WC–
and scratching of the WC at longer sliding distances. The Co hard metals. Wear 2001;250:604–10.
ultraWne and nanoscale grades showed homogeneous [8] Shipway PH, Hogg JJ. Dependence of microscale abrasion mecha-
nisms of WC–Co hardmetals on abrasive type. Wear 2005;259:44–51.
plowing behavior. DiVerences in wear mechanisms [9] Engqvist H, Axén N. Abrasion of cemented carbides by small grits.
observed in nano and coarse grades caused diVerent Tribol Int 1999;32:527–34.
wear rate and, therefore, lower wear resistance was [10] Scieszka SF. The wear transition as means for hardmetal fracture
observed in the nanoscale materials. toughness evaluation. Int J Refract Met Hard Mater 2001;19:101–8.
• Numerical simulation of the Wne abrasive and low load [11] Scieszka SF. Edge failure as a means of concurrently estimating the
abrasion and edge fracture resistance of hard-metals. Tribol Int 2005;
test conditions showed that micron and submicron 38:834–42.
microstructures respond the load diVerently. In the [12] McCormick NJ. Edge Xaking as a measure of material performance.
micron grade, plastic deformation is localized in the Met Mater 1992;8:154–6.
binder phase on the tensed side of a specimen. In the [13] Hangl M, Danzer R, Paar R. Edge toughness of brittle materials. In:
submicron grade, besides concentration on the tensed Petit J, editor. Mechanisms and mechanics of damage and failure,
Proceedings of the 11th biennial european conference on fracture –
side, strain is distributed into the bulk. The modeled ECF 11, vol. 1. Engineering Materials Advisory Services, Warley,
strain maps are well conWrmed to the wear mechanisms 1996. p. 707–13.
observed experimentally in micron and submicron [14] Morrell R, Gant AJ. Edge chipping of hard materials. Int J Refract
grades. Met Hard Mater 2001;19:293–301.
[15] Langer SA, Fuller Jr ER, Carter WC. OOF: an image-based Wnite-ele-
• Utilization of the coarse abrasive size (80 grit) at 2 and
ment analysis of material microstructures. Comput Sci Eng 2001;3:
15 N normal loads test conditions neglected inXuence of 15–23.
microstructure response and similar wear mechanisms [16] Gant AJ, Gee MG, Roebuck B. Rotating wheel abrasion of WC/Co
were observed for submicron and coarse grades. Flaking hardmetals. Wear 2005;258:178–88.