ELSEVIER Wear 197 (1996) 242-247
Experimental study on the abrasive wear in metal cutting
Departnmt of Marurfnctunrg Engineering, Teclmicnl University in Plovdiv, 4000 Plovdiv, Bulgaria
Received 18 October 1995; accepted 8 February 1996
Experiments were carried out m an attempt to analyze the influence of cutting conditions and parameters of the abrasive inclusions in the
work material on the amount of abrasive wear of the carbide cutting loads. The specimens were prepared~by powder metalbrgy from high
manganesesteei with a controlied structure and varied with the kind, concentration and granularity of the abrasive particles. The results show
that abrasive wear increases approximately linearly with the cutting temperature due to the change of the abrasive capability of some abrasive
inclusions with work hardness comparable with that of the tool material. The amount of abrasive wear w ABRwas found to depend non-linearly
on the particle concentration pa; wABR_ pa2’3. Another conclusion is that the amount of worn metal increases to some extent with the size of
the abrasive particles due to the larger size of the carbide conglomerates split off by the particles when they move along the tool surface.
Keywords: Abrasive wear; Cutting tool wear: Abrasive test
1. Introduction this definition: (a) the essential question whether the abrasive
particles, basically softer than carbides, are capable of causing
The nature of wear in metal cutting is not clear enough yet wear and if so, which way they act; and (b) the kind, con-
in spite of numerous investigations. The complexity of the centration, shape, properties and wear ability of these parti-
processes in the cutting zone hampers formulation of a thor- cles. The latter is a matter of materials science rather than
ough theory of cutting tool wear. Different theories have been metal cutting or tool engineering. Unfortunately, work in this
introduced hitherto to explain the wear mechanism but to the field is scanty and very limited data from the tool engineer’s
best of the author’s knowledge none gives a simple way of point of view are available.
predicting the amount of tool wear without initial experimen- The work materials contain exogenous and endogenous
tal work. These circumstances necessitate the development non-metal inclusions. The exogenous inclusions are compar-
of new and more sophisticated experimental methods in this atively scanty and do not play a significant role in wear. The
area. endogenous inclusions are silicates, carbides, oxides, phos-
It is accepted that cutting tool wear is induced by compli- phides, sulfides, etc., products of deoxidation and oxidation
cated physical, chemical and mechanical phenomena. Several in melted metals [ 21. The soft inclusions like phosphides and
“simple” mechanisms of wear (adhesion, abrasion, diffu- sulfides do not affect the abrasive wear, but the hard non-
sion, oxidation, etc.) can be defined that account for the wear metal inclusions such as carbides, oxides and silicates possess
of cemented carbides. The different mechanisms act simul- a strong abrasive ability even at the elevated temperatures
taneously with the predominant influence of one or more of typical for cutting. Some of these particles retain a hardness
them in different situations. In comparison with the other at high temperature comparable to or even higher than that
mechanisms of wear, abrasive wear is considered to be rela- of carbides. In machining, such particles can cut into the tool
tively insignificant, as it contributes to only about 10% of the surface, acting as a cutting tool and producing craters or
total tool wear [ I]. Nevertheless, when machining some grooves (so called hard abrasion following Larsen-Basse
materials that contain high concentrations of non-metallic [ 31) . In some circumstances they will split off grains or entire
inclusions, abrasive wear plays a significant role and is among grain conglomerates which on their part play the role of abra-
the main reasons for cutting tool damage. sive particles.
Abrasion is the process of tool wear during which hard Not only the temperature and material of inclusions affect
abrasive particles of the work material cut into the tool surface the abrasive wear. The results of many works show the sig-
and form wear particles. Two groups of problems arise from nificance of particle size and concentration as well. Ramal-
0043-1648/96/$15.00 0 1996 Elsevier Science S.A. Altkights reservea
V. Marinov/ Wear 197 (1996) 242-247 ‘43
ingham and Wright [ 41 have shown that crater wear increases normal solids [ 81. Cutting temperatures can reach values as
sharply with the particle concentration. In their paper they high as 1000 “C or more. In these circumstances-high tem-
presented experimental results obtained by Byrd and Fergu- peratures and enormous contact stresses-the tool surface
son, and Faulring. In both works the specimens were prepared layer acquires plasticity. That is why wear of cemented car-
by means of powder metallurgy. Byrd and Ferguson used a bides in metal cutting characterizes with some elements typ-
Fe-C-A&O3 mixture with 2 wt.% alumina and particle sizes ical of abrasion of steels (forming of plastic cut grooves)
1 p,m and 25 km. Faulring used a Fe-C-SiO, mixture vary- together with brittle splitting off of grains and grain conglom-
ing in size (5 km and 10 pm) and concentration (0.2 wt.% erates. These considerations make some experimental tests
and 0.4 wt.%) of the abrasive particles. Experiments have at low speeds and room temperature by means of abrasive
been carried out in constant cutting conditions and tool geom- papers [ 3,6] or indentors  more useful as qualitative
etry, i.e. at constant cutting temperature. According to the rather than quantitative methods. An experimental method
results of Byrd and Ferguson, with the increase in particle should take into account all features of the real process. The
size the wear land on the flank grows, but the depth of crater test should be carried out in conditions as close to the real
on the rake does not change. The results of Faulring show ones as possible. Ramalingham and Wright  used spe-
that the rise in particle concentration leads to more intensive cially prepared stainless steels with controlled characteristics
abrasive wear on the flank, but the amount of wear reduces The problem here is how to control exactly concentration,
when the particle size increases. Kramer  assumed that distribution, size and kind of the non-metallic inclusions.
the abrasive wear depends on the cutting speed (i.e. cutting Such control is necessary if one intends to draw conclusions
temperature) and the H,IH, ratio, where H, is the inclusion about the influence of these variables on the abrasive we&.
hardness and H, is the tool hardness. H, and H, are assumed The application of powder metallurgy gives an opportunity
to be temperature dependent. Low-speed tests at room tem- to produce appropriate specimens with controlled character-
perature were carried out by Larsen-Basse . The volu- istics. Usui and Shirakashi [ 71 used a bar with a longitudinal
metric removal rate was found to be proportional to the slot parallel to the axis that was filled with a mixture of
applied load, the type of abrasive material and the hardnesses binding agent, steel and aluminum oxide powder. Two types
of the particle and carbide materials. Similar conclusions of specimen were used varying with the size of powder grains.
were also formulated by Suh [ 61. Although the specimen technology is not fully described, it
The re<s cited above suggest that several important var- might be assumed from the information given that the mixture
iables dictate the mechanism of abrasive wear, namely cutting was a composite material with a non-metallic matrix, there-
temperature (i.e. cutting conditions) and abrasive particle fore it is difficult to accept as a material suitable for modeling
type, size and concentration. Although the importance of of a metal cutting process. The specimens of Byrd and Fer-
temperature is certain, only a few experimental tests on abra- guson, and Faulring [ 41 were prepared entirely by sintering
sive wear under different cutting temperatures have been of Fe-C-abrasive particle mixtures. This approach allows
carried out . The above results also allow us to analyze exact control of the process variables and mechanical prop-
the relations between abrasive particle characteristics and erties of the work material and should be considered as the
cutting tool life. Undoubtedly, cutting tool life is an easy-to- most appropriate for abrasive tests in metal cutting. Unfor-
apply in practice wear parameter but from the tribological tunately, in these experiments no different abrasive materials
point of view one should consider the volumetric amount of were tested and compared. Also, the cutting conditions (cut-
wear as a more precise measure. All these facts show the ting temperature) were kept constant, therefore no conclu-
necessity for an extensive experimental study to clarify the sions in this direction are reported.
influence of cutting conditions, abrasive particle concentra-
tion, size and type on the amount of abrasive wear. Some
experimental results along these lines are presented in this 3. Preparation of specimens
Powder metallurgy was used to prepare ten specimens with
different types, quantities and granularity of the abrasive par-
2. Experimental techniques ticles. The specimens varied with:
l the material of the abrasive particles-Al,O,, SiOZ and
As noted above, a number of different mechanisms account SIC;
for the wear of cemented carbides. Hence, each experimental l the size of the particles-8,40 and 90 pm;
method developed for investigation of the abrasive wear in l the concentration-0.0127, 0.0255, 0.127 and
metal cutting must be abrasive sensitive, i.e. give an exact 0.318 ~01%.
picture of the tool wear due only to abrasion, eliminating the A high manganese steel was chosen as the basic material.
influence of other wear mechanisms. Metal cutting is a unique The chemical composition of the mix before adding the non-
environment from a tribological point of view. At the very metal inclusions was 1.19% C, 16.3% (Fe-Mn) , and 82.15%
cutting edge the friction is internal and occurs between the Fe with 3.5 wt.% zinc stearate. The non-metal inclusions
layers of the work material as typical for fluids rather than prepared for each specimen were blended with the mixture
244 V. Murinov/ Wear 197 (1996) 242-247
prior to compaction. After compaction at 700 MPa, the green (b) The total volumetric wear WE was calculated by graph-
compacts were sintered in a controlled-atmosphere (disso- ical integration (Fig. 1) :
ciated ammonia) three-chamber furnace to 1150 “C. The
mechanical properties of the sintered materials were: den- W,=O.l 2s~;lc~;;)$vIX n=[l,11 I=lOb (1)
sity-6260 kg rnp3, hardness-HB 185. Special measures
were provided to minimize the influence of stochastic and where VB, is the width of the wear land measured at each
uncontrollable factors: 0.1 mm, CQis the clearance angle, ‘yeis the rake angle, and b
the specimens were produced from the same preliminarily is the length of wear land along the cutting edge in mm, equal
blended basic mix to keep constant the basic chemical to the width of cut.
composition; (c) The amount of wear per unit width of cut, wE, was
the process conditions of blending, compaction and sin, calculated by dividing WE by b.
tering were kept constant for all specimens; (d) The graphs wX vs. cutting time, 7, were plotted and
to avoid stratification and to obtain a uniform volume den, the rate of wear W was estimated according to Fig. 2 as
sity, the specimens were produced as compound parts,
each consisting of 11 rings positioned on a mandrel. The ti=AwlAr (2)
ring die dimensions were: outer diameter-54.55 mm, (e) For each specimen, the amount of wear in constant
inner diameter-23.4 mm, length-25 mm. cutting time r= 3 min was defined as
Specimen #l did not contain abrasive particles and was
Gz= T& .
used to estimate the amount of tool wear due only to abrasion. (3)
(f) The amount of abrasive wear wABRwas estimated as
wABR, = w2, - w_iY, (4)
4. Experimental procedure
where the indexes refer to the number of the specimen, the
The specimens were machined on a CNC tool lathe with a first one having no abrasive inclusions.
PSPNR 3225-12 cutting tool provided with a K20 carbide
insert SNMM 120408. The cutting conditions varied as fol-
lows: cutting speed V-from 0.33 to 0.75 m s-r, feed f 5. Results and discussion
-from 0.15 to 0.33 mm tr-‘, and depth of cut d-from 1.2
5. I. Cutting conditions
to 2.5 mm. The experimental data were processed in the fol
lowing sequence: The wear mechanisms are temperature induced. The nat-
(a) The flank wear land was photographed by a tool micro ural way to analyze the influence of the cutting conditions on
scope at various periods of time. the wear, therefore, is to represent their effect through the
cutting temperature. Fig. 3 shows the influence of the cutting
0. lmm Y
IF Rake face
Fig. 1 Graphical integration of the worn metal volume,
840 850 860 870 880 890 T,*C
Fig. 3. The average total amount of wear 1;~ and average abrasive wear
Fig. 2. Defimtlon of the rate of wear. WA,, as a function of the cutting temperature.
V. Murinov/ Weur 197 (1996) 242-247 245
. I 0"mnl' .:;:.;
:.:.:.: ..:.:.:.:.:.....:.......:.: ..
.............. .: .....
::::::‘:.< :...:: :.:::y.y:r::_
.: :.:.:.:.: :,:,
: :.:: ::::,:.:,:.:,:,::::~:,:,~:.:.:
I o- :_:_:.:.:.:
)~:: :,:.:: “Y:: ..................
,:.: ............ i,i’,i:~:::i:‘j:~:~:~ii:~jiii’:
A&O, Sic SiO
Fig. 5. The amount of abrasive wear as a function of the abrasive material.
the basic hypothesis that the abrasive wear depends strongly
on the ratio HA/H,, where HA is the particle hardness and
HM is the tool material hardness, both at ambient temperature.
T The abrasive capability of an abrasive particle Ca can be
Fig. 4. Influence of the cutting temperature on the wear mechanisms presented as a function of this ratio by a simple expression,
assuming that if the ratio is greater than some constant 5; then
the abrasive particle always causes wear (Ca = 1) ; if this ratio
temperature on the average total amount of wear Wx and
is smaller than 1 there is no wear (Ca= 0)) and if l>
average abrasive wear WABR.The cutting temperature T is
H,/H, > 1, the amount of abrasive wear changes linearly
calculated from an empirical equation for the same work
from 0 to the maximum value in these conditions ( 1 > Ca >
material obtained by means of a general factorial experiment
0). Thus Ca defines the possibility of an abrasive particle
b;Filonenko et al. [ lo] :
causing wear. Since the hardnesses depend on the cutting
temperature, Ca is also temperature dependent. The constant
5 has values of 1.2-2.0 [ 3,11,12]. Fig. 6 shows experimental
The equation is valid in the following range of cutting con-
results for the abrasive wear of specimens with different abra-
ditions: V= 15-65 m min- ‘, f= 0.1 l-O.34 mm tr- i. and
sive inclusions together with plots of Ca in l= 1.8. The hard-
d= OS-3 mm.
ness of Al,O, particles in the range of the work temperature
The high manganese steel is a typical difficult-to-cut mate-
is always higher than lHM and Ca = 1. The amount of abrasive
rial and the cutting temperature reaches high values (up to
wear is high, approximately constant, and not temperature
900 “C) at a cutting speed of about 1 m s-‘. In these condi-
dependent. In the case of SIC inclusions the abrasive capa-
tions the predominant wear mechanism is diffusion accom-
bility of the particle increases with the temperature since the
panied by abrasive wear. Adhesive wear prevails at low
hardness ratio increases as well. The wear dependence is
temperatures as shown in the well-known dependence
affected by Ca and the abrasive wear increasess at high tem-
between tool wear mechanisms and temperature (Fig. 4).
peratures. At very high cutting temperatures S102 is softer
The experimental results show good coincidence with the
than the tool material carbides, Ca = 0, and theoretically no
theoretical ones in the last third of the diagram. In our exper-
abrasive wear should exist. Nevertheless, some wear was
iments the abrasive wear is about 20-25% of the total amount
registered. One possible reason is that in the quartz sand from
of wear. This differs considerably from the value of 10%
which the SiOZ was inserted there were impurities of some
cited by Ho and Chen [ 1 ] and from the common opinion that
other considerably harder abrasive materials
the abrasive wear is of secondary importance and might be
neglected. If it is true for most of the work materials, abrasion
is one of the main reasons for tool failure when cutting mate- 5.3. Particle me
rials with elevated concentrations of abrasive particles or
materials like reinforced plastics The influence of the particle size is shown in Fig. 7. Assum-
ing that the abrasive wear model can be described as a simple
5.2. Abrasive muterials sliding of abrasive particles along a deformable tool surface,
the size of the particles should not affect the amount of worn
Experimental results for specimens with miscellaneous material if in the specimens the volumetric contents of the
abrasive inclusions are shown in Fig. 5. Differences in wear abrasive particles with different sizes are the same. Some
are significant, more than ten times between the hardest experimental results with tool material steel reveal a negli-
(A1203) and softest (SiO?) materials. The results confirm gible role of the particle size in abrasive wear [ 121, others
K Mrrrinov / Wear 197 (1996) 242-247
840 850 860 870 880 890 840 850 860 870 880 890 840 850 860 870 880 T, “c
AN, Sic SiO,
Fig. 6. The influence of the cutting temperature on the abrasive wear for different abraswe materials.
I O’lmm3 .10-3mm3 .
. 20. : I
lo- ’ f
0 0.05 0.1 0.15 0.2 0.25
0 20 4b 60 80 Da,nm Fig. 8. Abrasive wear as a function of the abrasive particle concentration.
Fig. 7. Abrasive wear as a function of the particle size.
imply that the bigger the particle is the less is the abrasive plot of wABRvs. pa should not be a straight line, as confirmed
wear (Faulring, in [ 41) . Experiments with carbides indicate in our experiments.
that the more intensive wear occurs if the particle size is
bigger (Byrd and Ferguson, in [ 4,7] ). Similar results are
shown in Fig. 7. One possible explanation is that the abrasive
particles not only cut grooves in the tool surface as in the
case of high speed steels but they also split off carbide grains
or entire grain conglomerates. The big abrasive particles will
snlit off bigger carbide grain conglomerates. An experimental method and results of abrasive wear
when cutting high-manganese steel with a carbide cutting
5.4. Abrasive concentration tool are presented. Powder metallurgy was used to prepare
the specimens with different types, quantities and granu-
The-experimental results shown in Fig. 8 confirm the larity of abrasive particles.
reports of other authors about the influence of the volumetric The results obtained show an approximately linear rela-
concentration of abrasive particles on the amount of abrasive tionship between the amount of abrasive wear and cutting
wear. The amount of worn material is in direct proportion to temperature. The abrasive wear is about 20-25% of the
the number II of abrasive particles which possess abrasive total amount of wear for the specimens with an artificially
potential and are in contact with the tool surface, i.e. elevated concentration of abrasive particles.
+vABR-~r. This number is defined as II = C,(~Z*)*‘~, where The results confirm that the amount of abrasive wear
n* is the total number of abrasive particles per unit volume depends on the ratio of particle and tool material hard-
of the work material, n* = pa/V,, pa being the volumetric nesses at ambient temperature. The increase in the wear
percentage of abrasive particles and VP the volume of an with the cutting temperature is controlled by the presence
individual particle. The coefficient C, depends on Ca and of abrasive particles with a hardness equal to or higher
geometrical characteristics of the particles and is a constant than that of the total material, as their coefficient of abra- -__-
in defined cutting conditions. Hence w,n,-pa”’ and the sive capability is greater than zero and not a constant.
V. Marinov / Wear I97 (1996) 242-247 247
When the size of the abrasive particles increases, the  E. Usui and T. Shirakashi, Analytical prediction of cutting tool wear,
Wear, 100 (1984) 129-151.
amount of worn metal increases too, due to the larger size
[ 8 J E.M. Trent, Metal cutting and the tribology of seizure: Movement of
of the carbide conglomerates split off by the particles work material over the tool in metal cutting, Wear, 128 (1988) 47-
when they move along the tool surface. 64.
The amount of abrasive wear wABRwas found to depend  T.O. Mulheam and L.E. Samuels, The abrasion of metals: A model of
non-linearly on the particle concentration pa; the process, Wear. 5 (1962) 478-498.
[IO] S.N. Filonenko, N.S. Molodtzov and V.P. Luca, Turning of High
Manganese Steel GlSLand HASTELOJDAiloy, Technica, Kiev, 1970
[ 1 l] D.N. Gorkunov, Tribotechnics, Mashinostroenie, Moscow, 1985 (in
[ 121 M.M. Hrushtchov and M.A. Babitchev, Abrasive Wear, Nauka,
[l] CF. Ho and N.N.S. Chen, Prediction of wear of carbide cutting tools, Moscow, 1970 (in Russian).
ht. .I. Prod. Rex, 15 (1977) 277-290.
 S. Ramalingham and J.D. Watson, Tool life distributions. Part4: Minor
phases in work material and multiple-injury tool failure, Trans. ASME, Biography
J. Eng. Indust.. 100 (1978) 241-253.
 J. Larsen-Basse, Effect of composition, microstructure, and service
conditions on the wear of cemented carbides, J. Metals, 35 (1983)
Dr. Marinov received an MSc from the Higher Technical
35-42. School in Rousse, Bulgaria in 1979, and a PhD in Manufac-
 S. Ramalingham and P.K. Wright, Abrasive wear in machining: turing Engineering from the Technical University in Sofia,
Experiments with materials of controlled microstructure, Tmns. Bulgaria in 1992. He joined the Technical University in Plov-
ASME, J. Eng. Mater. Technol., 103 ( 1981) 77-84.
div in 1987. His research interests include tribology of cut-
[S] B.M. Kramer, A comprehensive tool wear model, Am. CIRP, 35
ting, finite element modeling of metal cutting, smart materials
 N.P. Suh, New theories of wear and their implications for tool and applications. He is a member of the Union of Scientists
materials, Wear, 62 (1980) l-20. in Bulgaria and the Society of Bulgarian Tribologists.