Materials Characterization 46 (2001) 399 – 406

Characterization of the microstructure of high-chromium

cast irons using Mössbauer spectroscopy
K. Webera,*, D. Regenera, H. Mehnerb, M. Menzelb
a
Institute of Materials Engineering and Materials Testing, Otto-von-Guericke-University Magdeburg,
Große Steinernetischstraße 5, D-39104 Magdeburg, Germany
b
Federal Institute for Materials Research and Testing (BAM), Richard-Willstätter-Straße 11,
D-12489 Berlin-Adlershof, Germany
Received 7 January 2000; accepted 16 February 2001

Abstract

The finish grinding process for cement materials uses ball mills with differently sized grinding balls. Because
the grinding takes place through the impact and friction of the balls, balls with a high wear resistance are desirable
and the materials of choice are the high-chromium white cast irons. This article examines the behavior of these
cast irons when subjected to various heat treatments. The emphasis is on the dependence of the microstructure
properties on the chemical composition of the alloys and the heat treatment parameters. The content of retained
austenite and that of carbon in the martensite phase were determined by the use of Mössbauer spectroscopy. The
results verify that a higher proportion of retained austenite with low carbon content and a martensite with higher
hardness produce a material with good fracture toughness and resistance to abrasive wear. D 2001 Elsevier
Science Inc. All rights reserved.

Keywords: High-cromium cast irons; White cast irons; Grinding process of cement materials; Grinding balls; As-cast grinding
balls

1. Introduction bility of brittle failure, fracture toughness must also

be a consideration in selecting an appropriate mate-
In ball mills, the grinding process involves im- rial. The reason for this lies in the fact that grinding
pact and friction between the grinding balls them- balls frequently hit each other and not the mill feed
selves, the grinding medium, and the mill lining particles [2].
(Fig. 1). High-chromium white cast irons are often This article examines the effects of composition
selected as ball material because of their high abra- and heat treatment on the microstructure and
sion resistance [1]. However, because of the possi- mechanical properties of high-chromium white cast
irons. Experiments were carried out at selected
temperatures in order to determine the effect of heat
treatment on the abrasive wear and fracture tough-
* Corresponding author. Tel.: +49-391-67-14536; fax: ness properties of three high-chromium cast irons.
+49-391-67-14569. Quantitative phase analysis with the aid of Mössba-
E-mail address: [email protected] uer spectroscopy was used in order to provide a
(K. Weber). better understanding of the influence of microstruc-

1044-5803/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved.
PII: S 1 0 4 4 - 5 8 0 3 ( 0 1 ) 0 0 1 4 3 - 7
400 K. Weber et al. / Materials Characterization 46 (2001) 399–406

Fig. 1. Schematic representation of the movement of grinding balls in a ball mill.

ture on the wear resistance and impact toughness subspectra areas were evaluated as a percentage of
properties [3 – 6]. the total spectrum based on the assumption that the
recoil-free fractions are equal for different Fe sites
and phases. The isomer shift was relative to pure iron.
2. Experimental procedure The errors due to isomer shifts, quadrupole splitting,
and line width were between ± 0.002 and ± 0.01 mm/
The chosen variable parameters for the study were s. The errors for most of the areas were < 1%, but for
chemical composition and heat treatment. The chro- some phases of low concentration, they were up to
mium and carbon contents of the three white cast 8% of the area obtained.
irons are given in Table 1. The as-cast grinding balls Microstructural examinations were performed
were 80 mm in diameter. The balls were subjected to using a scanning electron microscope (SEM). Hard-
various isothermal annealing and tempering heat
treatments. Isothermal annealing temperatures ranged
from 950°C to 1150°C with the holding time of 4 h
for all balls. Tempering treatments were between
200°C and 400°C, with holding times at these tem-
peratures between 2 and 5.25 h.
The samples for Mössbauer spectroscopy were
measured in transmission geometry using an SM-
2201 spectrometer (Institute of Analytical Instrumen-
tation, St. Petersburg) as well as a WissEl Mössbauer
spectrometer. The activity of the Co-57 source was 1
GBq in a Cr matrix. The spectra were registered at
room temperature using an Fe resonance detector or a
proportional counter. Parameter fits were performed
using a standard least-squares fitting routine. The

Table 1
The chromium and carbon contents of the three high-
chromium white cast irons
Alloy Chromium content, % Carbon content, %
Fig. 2. Scanning electron micrograph of Alloy B after
A 17.50 2.66
annealing for 4 h at 1150°C and tempering for 2 h at 200°C.
B 18.74 2.50
(1) Primary carbides; (2) secondary carbides; (3) retained
C 23.86 3.07
austenite; (4) martensitic spikes.
 K. Weber et al. / Materials Characterization 46 (2001) 399–406 401

Fig. 3. Mössbauer spectrum of Alloy B after annealing for 4 h at 1150°C and tempering for 2 h at 200°C.

ness measurements were made using the Vickers The dynamic fracture toughness of the high-chro-
method and a 50-g load, the values reported later mium white cast irons was determined with an
being the average of 10 measurements. instrumented impact test machine using Charpy

Fig. 4. Mössbauer spectrum of Alloy B after annealing for 4 h at 1000°C and tempering for 2 h at 200°C.
402 K. Weber et al. / Materials Characterization 46 (2001) 399–406

Table 2
Percentage of iron atoms in the retained austenite phase and martensite with carbon in interstitial sites for two heat treatments of
Alloy B
Fe atoms in Fe atoms in Fe atoms in Fe atoms in Fe atoms in
retained retained neighbor neighbor neighbor s
austenite austenite sphere sphere phere
depleted of with more 1 of C atoms in 2 of C atoms in 3 of C atoms in
carbon (singlet), carbon (doublet), martensite (Sextet 1), martensite (Sextet 2), martensite (Sextet 3),
Heat treatment % % % % %
1150°C, air, 36 25 8 12 19
2 h at 200°C
1000°C, air, 11 16 14 19 40
2 h at 200°C

specimens (10
10
55 mm3). The impact hammer secondary carbides precipitated preferentially in cer-
velocity was 1 m/s at impact. All reported fracture tain areas of the prior austenite matrix during austen-
toughness values represent the average of four tests. ite destabilization (Fig. 2, from Alloy B). This also
Following the tests, the fracture surfaces of the speci- indicates that the chromium and carbon diffused out
mens were examined in the SEM. of the supersaturated solid solution of austenite.
The abrasive wear investigation employed an Figs. 3 and 4 are Mössbauer spectra from Alloy B
impact crusher, in which a ball segment of the white in two different heat treatment conditions and provide
cast irons is rotated at 1200 rpm. Measurement of the a measure of the retained austenite content. Overall,
wear that resulted from the impact and friction the proportion of the retained austenite in the exam-
between the cement particles and the test pieces ined microstructure varied between 27% at the
was determined by the mass loss of the specimens annealing temperature of 1000°C and 61% at
after crushing at 17,000 contacts. 1150°C; that is, it gradually increased with increasing
isothermal annealing temperature. However, there are
two components to the retained austenite — one that
3. Results and discussion is carbon-depleted and the other that contains more
carbon. In all the spectra, there is a single line caused
3.1. Microstructure and quantitative phase analysis by the carbon-depleted retained austenite. At the
same time, the carbon atoms still in the austenite
The high-chromium white cast irons were sub- destroy the strict cubic symmetry of the face-centered
mitted to different kinds of heat treatment in order cubic structure and produce an electric field gradient
to obtain different microstructures and to examine on their neighboring iron atoms. These atoms give
their influences of the impact toughness and hard- rise to a doublet superimposed on the singlet of the
ness properties. austenite line. This latter component, the retained
The SEM examination revealed that interdendritic austenite with more carbon, consists of 11% of the
regions consist of rod and plate-like eutectic carbides total austenite in the balls heat-treated at 1000°C and
embedded in the secondary carbide-free matrix. The 25% in those heat-treated at 1150°C.

Table 3
Percentage of iron atoms in C-neighbor spheres in the martensite phase and corresponding hardness for five heat treatments of
Alloy A
Fe atoms in C Fe atoms in C Fe atoms in C
neighbor sphere neighbor sphere neighbor sphere
Heat treatment 1 (Sextet 1), % 2 (Sextet 2), % 3 (Sextet 3), % Hardness, HV50
950°C, air 15 24 33 846
1000°C, air, 15 20 32 797
2 h at 200°C
1000°C, air, 13 23 31 770
2 h at 400°C
1000°C, air, 12 22 31 748
5.25 h at 400°C
1050°C, air, 10 18 28 784
2 h at 200°C
 K. Weber et al. / Materials Characterization 46 (2001) 399–406 403

Fig. 5. Mössbauer spectrum of Alloy C after annealing for 4 h at 1000°C and tempering for 2 h at 200°C.

The need for the grinding balls to be resistant to the proportions of the various phases in Alloy B
abrasive wear requires that they have a matrix of high following the two different heat treatments.
hardness. The hardness is dependent on the carbon Alloy A, which contained slightly less chromium
content of the martensite phase, the other phase found but slightly more carbon than Alloy B, was subjected
in the grinding balls. Martensite is ferromagnetic and to several different heat treatments. Table 3 provides
has a Mössbauer spectrum similar to that of pure iron, a summary of the proportions of the three sextets in
the well-known six-line spectrum. Figs. 3 and 4 the martensite phase of this alloy and the correspond-
display a martensitic spectrum with three sextets. ing hardness values produced by the different heat
These three sextets and the intensities of their sub- treatments. The overall percentage of martensite is
spectra correspond to three different surroundings of greatest after the 950°C isothermal anneal (72%) and
the iron atoms. As can be seen in Figs. 3 and 4, the lowest following the 1050°C/200°C treatment (56%).
martensite content is higher (73%) in the balls iso- All three treatments based on the 1000°C isothermal
thermally annealed at 1000°C. Table 2 summarizes anneal yielded martensite percentages of 65 – 67%.

Table 4
Percentage of iron atoms in C-neighbor spheres in the martensite phase and corresponding hardness for five heat treatments of
Alloy C
Fe atoms in Fe atoms in Fe atoms in Fe atoms in
C neighbor C neighbor C neighbor C neighbor
sphere 1 sphere 2 sphere 3 sphere 4
Heat treatment (Sextet 1), % (Sextet 2), % (Sextet 3), % (Sextet 4), % Hardness, HV
950°C, air 25 13 17 22 740
1000°C, air, 20 18 10 29 766
2 h at 200°C
1000°C, air, 21 13 14 27 761
2 h at 400°C
1000°C, air, 20 15 9 28 779
5.25 h at 400°C
1050°C, air, 23 8 6 33 784
2 h at 200°C
404 K. Weber et al. / Materials Characterization 46 (2001) 399–406

Table 5
Hardness and reciprocal mass loss of Alloys A and C after various heat treatments
Alloy A Alloy C
Abrasive wear Abrasive wear
Heat treatment Hardness, HV50 resistance, 1/g Hardness, HV50 resistance, 1/g
950°C, air 846 8.28 740 6.87
1000°C, air, 797 7.85 766 6.94
2 h at 200°C
1000°C, air, 770 7.45 761 6.92
2 h at 400°C
1000°C, air, 748 5.25 779 6.08
5.25 h at 400°C
1050°C, air, 784 8.13 784 7.24
2 h at 200°C

Alloy C contained both the most chromium and of the matrix must be investigated together because
the most carbon of the three white cast irons exam- of the complexity of the working stress. Moreover,
ined. With the increase in chromium content, more of it is well known that a correlation exists between the
the carbon became bound up in chromium carbides hardness and the abrasive wear resistance.
and a martensite was produced that is depleted in The results of the present study of the abrasive
carbon. Consequently, the Mössbauer spectrum of wear resistance of grinding balls heat-treated under
this alloy (Fig. 5) contained an additional, fourth, different conditions then used for grinding cement
sextet. Table 4 provides a summary of the proportions particles in an impact crusher are summarized in
of all four sextets in the martensite phase of this alloy Table 5. The abrasive wear resistance parameter is
and the corresponding hardness values produced by the reciprocal value of the mass loss through wear
the different heat treatments. The overall martensite after crushing at 17,000 contacts. Thus the smaller the
proportions range from 70% to 77%, higher than mass loss, the higher is the parameter.
those produced by equivalent heat treatments on As was discussed in Section 3.1, the higher
Alloy A. However, because of the carbon depletion, amount of chromium carbides in Alloy C led to a
the hardness values measured were generally lower decrease in the carbon content of the martensite phase
than those produced in Alloy A. in this material compared with Alloy A. Particularly
at the lower heat treatment temperatures, there was a
3.2. Abrasive wear of the heat-treated grinding balls concomitant loss of hardness that, in turn, resulted in
a greater mass loss due to abrasive wear. However,
It has been established that the properties of the increasing the tempering temperature and time has
matrix have a decisive influence on the wear proper- a drastic negative effect on the wear resistance of
ties of white cast irons [7,8]. The microstructure of both alloys.
the high-chromium white cast irons consists of a
mixture of carbides inside a matrix of martensite and 3.3. Fracture toughness of the heat-treated grinding
retained austenite. The hardness and the toughness balls

The values of the dynamic fracture toughness of

Table 6 the heat-treated Alloys A and C are given in Table 6.
Dynamic fracture toughness K1d of Alloys A and C after
various heat treatments
K1d of Alloy A, K1d of Alloy C, Table 7
p p Dynamic fracture toughness K 1d and abrasive wear
Heat treatment MPa m MPa m
resistance of Alloy B after three different heat treatments
950°C, air 30.1 35.0
1000°C, air, 35.8 36.6 Abrasive wear
p
2 h at 200°C Heat treatment K1d, MPa m resistance, 1/g
1000°C, air, 27.0 34.7 1000°C, air, 29.4 8.66
2 h at 400°C 2 h at 200°C
1000°C, air, 25.8 28.7 1000°C, air, 26.1 6.05
5.25 h at 400°C 2 h at 400°C
1050°C, air, 33.4 40.9 1150°C, air, 38.0 4.97
2 h at 200°C 2 h at 200°C
 K. Weber et al. / Materials Characterization 46 (2001) 399–406 405

Table 8
Percentage of retained austenite in the matrices of Alloys A and C, based on the Mössbauer subspectra
Alloy A Alloy C
Retained austenite Retained austenite Retained austenite Retained austenite
Heat treatment depleted in carbon, % with more carbon, % depleted in carbon, % with more carbon, %
950°C, air 12 16 7 15
1000°C, air, 16 17 10 14
2 h at 200°C
1000°C, air, 14 19 8 16
2 h at 400°C
1000°C, air, 13 21 7 19
5.25 h at 400°C
1050°C, air, 23 21 15 16
2 h at 200°C

Table 7 lists the dynamic fracture toughness values fracture toughness. At the same time, the imposition
obtained from Alloy B in three different heat-treated of tempering treatments affects the distribution of
conditions, together with the corresponding abrasive carbon within the retained austenite. That is, increas-
wear resistance values. It is readily apparent that ing either the tempering temperature or the temper-
the fracture toughness of all three alloys is highly ing time reduces the amount of retained austenite
dependent on the heat treatment parameters involved. with low carbon. This brings about a reduction in
These results can be explained by consideration of the the fracture toughness and a lower resistance to
Mössbauer spectra and, specifically, of the retained brittle fracture.
austenite content and the determination of the carbon SEM examination of the fracture surfaces sup-
atom positions in the martensite. The quantitative ports the above. Fig. 6 is taken from a fractured
information obtained from the Mössbauer spectra of specimen of Alloy B that had been isothermally
Alloys A and C were presented earlier, in Tables 3 annealed at a high temperature (1150°C) then tem-
and 4, and those specifically pertaining to the retained pered for 2 h at 200°C. Examination of the fracture
austenite content are collected together in Table 8 for surfaces of this sample revealed that the fracture
ease of comparison. advanced along the interfaces of the massive eutectic
It is immediately evident from Table 8 that carbides in the interdendritic region and the retained
increasing the temperature of the isothermal anneal austenite. By contrast, the fracture surface shown
leads to an increase in the retained austenite content. in Fig. 7 is completely dominated by cleavage frac-
These higher contents, in turn, tend to increase the ture of the carbide colonies. This sample, also from

Fig. 6. Scanning electron micrograph of the fracture surface Fig. 7. Scanning electron micrograph of the fracture surface
of Alloy B after annealing for 4 h at 1150°C and tempering of Alloy B after annealing for 4 h at 1000°C and tempering
for 2 h at 200°C. for 2 h at 400°C.
406 K. Weber et al. / Materials Characterization 46 (2001) 399–406

Alloy B, was isothermally annealed at a much lower sifies the resistance to brittle fracture of these materi-
temperature (1000°C) but was subjected to a more als. The combination of an annealing treatment at
intense tempering treatment (2 h at 400°C). The 1000 – 1050°C (4 h) and a tempering treatment at
combination of a low isothermal anneal temperature 200°C (2 h) produced an optimized microstructure
and a more substantial temper treatment resulted in a that had good fracture toughness and was resistant to
lower resistance to brittle fracture than that evident in abrasive wear.
the sample of Fig. 6.

References
4. Conclusion
[1] Zum Gahr K-H. Abrasive wear of cast irons. Wear
Mössbauer spectroscopy is a powerful and unique 1980;64:175 – 94.
[2] Duda WH. Cement-data-book. Wiesbaden: Bauverlag,
tool for the study of metallurgical phenomena. The
1985.
method is suitable for investigating dilute polycrys-
[3] Cohen RL. Applications of Mössbauer spectroscopy,
talline iron alloys and for identifying complex mix- vol. 1. New York: Academic Press, 1976. pp. 37 – 81.
tures of iron phases. Mössbauer spectra show both a [4] Cranshaw TE, Dale BW, Longworth GO, Johnson CE.
superposition of many six-line subspectra, each of Mössbauer spectroscopy and its applications. Cam-
them with hyperfine magnetic fields and isomer shift bridge (UK): Cambridge Univ. Press, 1985.
determined by different surroundings (martensite) [5] Horton GK, Maradudin AA. Dynamical properties of
and a singlet (pure g iron) and a doublet (correspond- solids, vol. 5. Amsterdam: Elsevier, 1984.
ing to austenite containing carbon). [6] Lesoille M, Gielen PM. Mössbauer spectroscopy study
In the present study on high-chromium white cast of iron – carbon austenite and virgin martensite. Metall
Trans 1972;3:2681 – 9.
irons, it was determined that the hardness is not an
[7] Ki JT, Zhou QD, Liu SH, Song GS. Influence of re-
accurate indicator of the wear resistance of grinding
tained austenite on the wear resistance of high chromi-
balls in ball mills. More important was the carbon um cast irons under various impact loads. Wear 1993;
content of the martensite: more carbon in the mar- 162 – 164:83 – 8.
tensite led to lower abrasive wear losses. At the same [8] Zum Gahr K-H, Doane DV. Optimizing fracture tough-
time, a higher content of retained austenite depleted ness and abrasion resistance in white cast irons. Metal
in carbon increased the fracture toughness and inten- Trans 1980;11A:613 – 20.