Vol.

122 (2012)

No. 3

ACTA PHYSICA POLONICA A

Proceedings of the 12th International Symposium on Physics of Materials, Prague, September 48, 2011

Correlation between Microstructure/Fracture Surfaces

and Material Properties
a,

R. Bidulsk
a

, J. Bidulsk

and M. Actis Grande

Department of Applied Science and Technology, Politecnico di Torino, T. Michel 5, 15 100, Alessandria, Italy

Department of Metals Forming, Faculty of Metallurgy, Technical University of Koice

Letn 9, 042 00 Koice, Slovakia

The present paper deals with the evaluation of the eect of dierent vacuum heat treatments on the
microstructure and fracture surfaces of a low alloyed sintered Fe[1.5Cr0.2Mo]0.6C steel, in correlation with
the mechanical and plastic properties achieved. The heat treatment consists of the sintering process in vacuum
furnace at 1393 K for 1800 s, followed by dierent cooling conditions and an integrated nal tempering at 473 K
for 3600 s. The average cooling rates, calculated in the range of 1393 K to 673 K, were 0.1, 0.235, 3, and 6 K/s,
respectively. Vacuum heat treatment is supporting the bainitic-martensitic microstructure, the higher the faster
cooling rate applied. This provides a marked increase in strength coupled to a decrease in ductility. The decrease
of the impact energy after heat treatment is justied by the microstructural changes, especially if a part of
bainite is converted to brittle martensite. Samples with bainitic microstructure presented higher impact energies
than those with martensitic microstructure. This is fully conrmed by the fracture surface analysis at higher
magnication, revealing three main micromechanisms of fracture: brittle, ductile and quasi-cleavage.
PACS: 81.05.Bx, 81.20.Ev, 81.40.z, 62.20.mm, 62.25.Mn

The main aim of the paper was to study the eect

1. Introduction

Powder metallurgy (PM) is a well established technology for manufacturing parts to net or near net shape.
Some recent developments have been introduced, with
the goal of improving the PM parts performance.

of dierent vacuum heat treatments on the microstructural and fracture behaviour of Fe[1.5Cr0.2Mo]0.6C
PM steel.

In

2. Experimental conditions

terms of chemical compositions, systems using cheaper

alloying elements such as Cr and Mn [13] were introduced into the market in the last years. As for the density
improvements or porosity control, compacting techniques
have been developed, including various methods of severe
plastic deformation [47]. Sintering has also been modied, in order to get the full control of the processing
conditions, by means of a deep analysis of the temperature prole, sintering atmosphere purity, oxygen/carbon
ratio, dew points, cooling rates [813].
Dierent heat treatments can also be used. The vacuum heat treatment is a very successful secondary operation tool, due to its high exibility.
The nal microstructural constituents mainly depend
on the cooling rate, dimension of the part and, clearly,
on chemical composition.

In any case, heat treatments

are necessary to tune the mechanical properties to the nal requirements and consequently, the eect of the heat
treatments on the microstructure and mechanical properties has proven to be a very dynamic research eld during
the last years [14, 15].

The composition of the investigated system was Fe

[1.5Cr0.2Mo]0.6C. Graphite and AW wax powder (as
lubricant) were added to a commercial FeCrMo powder (namely Astaloy CrL, Hgans AB) and then homogeneised in a Turbula mixer for 20 min. Specimens with
3
a green density of approximately 7.0 103 kg m
were
obtained using a 2000 kN hydraulic press, applying a
pressure of 700 MPa. Two dierent specimen types were
prepared: dog-bone tensile (ISO 2740) and unnotched
3
3
impact energy 55 10 10 10
m (ISO 5754). Specimens were debinded before sintering in a Nabertherm
type furnace.
The heat treatment conditions consisted of the sintering process in a vacuum furnace at 1393 K for 1800 s
with an integrated nal tempering at 473 K for 3600 s.
The average cooling rates were calculated in the range
of 1393 K to 673 K and were 0.1, 0.235, 3, and 6 K/s,
respectively.
In vacuum furnaces, the cooling rate is generally determined by the pressure of the gas (N2 ) introduced into
the chamber. Dierent gas pressures were used 0 Pa,
50 kPa, 200 kPa, and 600 kPa.

corresponding author; e-mail:

[email protected]

The cooling rate was

monitored and recorded by means of thermocouples inserted in the central axis and close to the surface of the

(548)

Correlation between Microstructure/Fracture Surfaces . . .

549

specimen. The processes run fully automatically and the

documentation of the process is provided by the curves
of the programmer to give reproducible results.
Sintered specimens were tested on a ZWICK Z100 machine and in an impact testing apparatus ZWICK RKP
450.

Microstructures observations and fracture surface

analysis were carried out using a light optical and scanning electron microscopy. The apparent hardness (measured on the surfaces of tested specimen) was determined
by means of Vickers hardness indenter (HV10). Densities
were evaluated using the water displacement method.

3. Results

The microstructure of the studied materials is presented in Figs. 1 and 2.

Fig. 2. Microstructure of material cooled at 3 K/s (a)

and 6 K/s (b) rate, respectively.

Fig. 1. Microstructure of material cooled at 0.1 K/s

(a) and 0.235 K/s (b) rate, respectively.
At the lowest cooling rates, 0.1 K/s (Fig. 1a), the microstructure consists predominantly of pearlite.

In the

system cooled at 0.235 K/s (Fig. 1b), the microstructure

is a mix of bainite and pearlite.
Increasing the cooling rate results in an increased
amount of bainite/martensite.

In the system cooled

at 3 K/s (Fig. 2a), the microstructure consists of a

mix of martensite and bainite with a small amount of
tempered martensite.

In the system cooled at 6 K/s

(Fig. 2b), a dominantly martensitic microstructure with

small amount of tempered martensite can be observed.

Fig. 3. Fracture surface of material cooled at 0.1 K/s

and 0.235 K/s rate, respectively.

550

R. Bidulsk, J. Bidulsk, M. Actis Grande

The fracture surfaces of investigated materials are presented in Figs. 3 and 4.

All surfaces show a basically brittle behaving material, but at higher magnications, dierent fracture micromechanisms are detected. Figure 3a shows a ductile
fracture surface with localised plastic ow, presumably
formed principally of plastic deformation by slip.

The

pearlite microstructure features are dominant in the fracture surfaces. Figure 3b shows a ductile fracture surface.

Surfaces presented in Fig. 4a and Fig. 4b show a transgranular cleavage fracture mode. Figure 4a presents the
transgranular cleavage with river patterns features. Figure 4b presents two tilted transgranular cleavage planes,
with river patterns features near the isolated pore.

Detailed

information

about

microstructural

con-

stituents, the prevailing fracture surface mode and the

mechanical properties are presented in Table.

The results of the mechanical properties evidence that

with increasing values of transverse rupture strength
(TRS), the impact energy values decrease.

Several au-

thors [1618] suggest that the decrease of the impact energy after heat treatment is justied by the microstruc-

Fig. 4. Fracture surface of material cooled at 3 K/s

and 6 K/s rate, respectively.

tural changes, mainly if a part of bainite is converted to

brittle martensite.

TABLE
Microstructural constituents, prevailed fracture surface mode and mechanical properties.
Cooling
rate
[K/s]

Microstructure
constituent
P
F
B
M
[%]
[%]
[%]
[%]

0.1

97

0.235

77

23

32

68

18

82

Fracture surface

ductile, localised
plastic ow
ductile, localised
plastic ow occasionally
cleavage, localised
plastic ow occasionally
quasi-cleavage, cleavage,
shallow dimple

4. Discussion

Table focuses on the inuence of microstructure and

prevailing fracture surfaces on the mechanical properties.
Results in Table indicate that the microstructure and
fracture investigations properly match with the results of
strength and ductility. The strength of heat-treated PM
steels depends on the base powder, the carbon content,

Mechanical
properties
TRS
IE
HV10
[MPa]
[J]
934

17.0

191.2

1009

16.8

193

1114

13.9

220.4

1211

9.3

435.8

the cooling rate, the density and sintering; the combination of those variables may lead to the attainment of
a hardened steel preserving adequate ductility and fracture toughness. The admixed microstructures of bainite
and martensite predominantly aect the strength of the
material.

The obtained strength levels, coupled to the

microstructures and fracture surfaces, conrm that cool-

Correlation between Microstructure/Fracture Surfaces . . .

551

ing rates of about 2 bar (3 K/s) are sucient to obtain

a mixed bainite and martensite microstructure.
Samples with bainitic microstructure presented higher
impact energies than those with martensitic microstructure. This is fully conrmed by the fracture surface mode
at higher magnication revealing two main micromechanisms of fracture: brittle (transgranular cleavage fracture) and ductile (dimples and plastic deformation by
slip). The brittleness is also caused by a complex eect
of microstructure heterogeneity around huge contaminations (complex refractory oxides, where chromium oxide
is dominant), mainly in places surrounding the original
powder particles.

Karlsson et al. reveal [19] that ox-

ides can be formed during the atomisation of low alloyed

chromium sintered steels. These oxides are much harder
to be reduced in subsequent processing. Such a layer of
oxides keeps covering the surfaces of the annealed powders and consequently oxides are found entrapped in the
sinter necks, Fig. 5 and Fig. 6.

Fig. 6. Trapped oxides inside pores in the fracture surface of material cooled at 3 K/s and 6 K/s rate, respectively.

ductile fracture occurred at the powder particle contacts.

Moreover, impurity segregation (oxides) and pores play
a role in the transition route from the interparticle ductile fracture to brittle mechanisms, mainly in the region
adjacent to the powder particles and/or the grain boundaries.
Another possible transition route for a quasi-cleavage
mechanism that was noticed ran along the interface between two dierent phases, that is to say between the
lower bainite (lath bainite) and transformed martensite
(tempered martensite). This could be attributed to the
dierent hardness at the two neighbour microstructure
phases, which caused dislocation pile-ups at the interface
and, according to [20], a concentration of stresses.

Ac-

cording to [21, 22], these fracture behaviours are called

Fig. 5. Trapped oxides inside pores in the fracture surface of material cooled at 0.1 K/s and 0.235 K/s rate,
respectively.

quasi-cleavage, especially in the fracture of quenched and

tempered steels.

In any case, Krauss [21] underlined

that the line between the terms cleavage and quasi-cleavage is somewhat arbitrary. Other authors stated

The pores cause the distribution of stress to be inhomo-

[2325] that quasi-cleavage is frequently decorated with

geneous across the cross-section and fractionate the frac-

ne lines called river markings running approximately

ture surfaces to the various fracture micromechanisms:

parallel to the crack propagation direction. They origi-

quasi-cleavage/cleavage and ductile with shallow dimple.

nate in the grain interior rather than at the grain bound-

Figure 5 and Fig. 6 reveal a microvoid coalescence frac-

ary and the river markings on opposing fracture surfaces

ture mechanism between adjacent powder particles. The

match ridge-to-ridge as opposed to mating step-to-step

formation of microvoids indicates that an interparticle

as in cleavage [25].

552

R. Bidulsk, J. Bidulsk, M. Actis Grande

However, the correlation of this surface with the un-

References

derlying microstructure is not fully understood and, consequently, the formation mechanism has not been identied so far. Nevertheless, such single event models are
not sucient to describe the quasi-cleavage mode often
observed in tempered martensitic steels.
Quasi-cleavage is characterized by single or small clusters of planar cleavage facets, separated from other co-planar facets or clusters of facets by ductile tear. Prior
austenitic grains contain several misoriented lath packets
which constitute the basic unit of crystallographic orientation. Quasi-cleavage involves frequent microcrack nucleation and propagation events, stopping at misoriented
lath packet or prior austenite grain boundaries.
The lack of understanding of the quasi-cleavage micromechanism is due to the limited knowledge of microstructure/fracture development for the studied low alloyed PM microstructures at high cooling rate (3 and
6 K/s).

The traditional

PM pecularities (sintered

necks, localized strain, grain boundaries, pores, original

powder particles, second phases) play a very important
role and these pecularities cause more dierences in microfracture modes if compared to the wrought materials.
The fracture behavior of vacuum heat treated low alloyed steels cooled at higher rates (more than 2 K/s) is
relatively unexplored.

Previous articles dealt with the

mechanical properties and microstructure development

[9, 10, 26, 27], whereas the proposed paper explores the
correlation between fracture modes, microstructure development and material properties. The obtained results
show that the vacuum heat treatments of low alloyed
sintered steel containing Cr and Mo is an eective instrument for microstructural hardening, to obtain high-strength sintered steels suitable for the production of
highly loaded mechanical parts.

5. Conclusion

1. Vacuum heat treatment is supporting the bainitic-martensitic microstructure, the higher the faster cooling
rate applied. This provides a marked increase in strength
coupled to a decrease in ductility.
2. Dierent micromechanisms of fracture were identied in the evaluated systems: brittle (quasi-cleavage and
cleavage) and ductile (dimples and plastic deformation
by slip).
3. Results indicate the strong correlation between microstructure, fracture surfaces and the mechanical properties, in terms of strength and ductility.

Acknowledgments

J. Bidulsk thanks Slovak national projects VEGA

1/0385/11.

[1] E. Hryha, L. ajkov, E. Dudrov,

Prog. 7, 181 (2007).
[2] E. Hryha, E. Dudrova, L. Nyborg,
Trans. A 41, 2880 (2010).

Powder Metall.
Metall. Mater.

[3] M. Seleck, A. alak, Int. J. Powder Metall. 46, 29

(2010).
[4] J. Bidulsk, T. Kvakaj, R. Bidulsk, M. Actis
Grande, High Temp. Mater. Process. 27, 203 (2008).
[5] J. Bidulsk, T. Kvakaj, R. Bidulsk, M. Actis
Grande, T. Doni, M. Martikn, Acta Metall. Slovaca
16, 4 (2010).
[6] R. Bidulsk, J. Bidulsk, M. Actis Grande, High
Temp. Mater. Process. 28, 337 (2009).
[7] M. Kvakaj, T. Kvakaj, A. Kovov, R. Koiko,
J. Bacs, Acta Metall. Slovaca 16, 84 (2010).
[8] S.C. Mitchell, A. Cias, Powder Metall. Prog. 4, 132
(2004).
[9] M. Zendron, L. Girardini, A. Molinari, Powder Metall. 51, 237 (2008).
[10] S. Hatami, A. Malakizadi, L. Nyborg, D. Wallin,
J. Mater. Process. Technol. 210, 1180 (2010).
[11] E. Hryha, C. Gierl, L. Nyborg, H. Danninger, E. Dudrova, Appl. Surf. Sci. 256, 3946 (2010).
[12] M. Hrubovcakova, E. Dudrova, Acta Metall. Slovaca
15, 248 (2009).
[13] J. Bidulsk, R. Bidulsk, M. Actis Grande, Acta Metall. Slovaca 16, 146 (2010).
[14] R. Bidulsk, M. Actis Grande, High Temp. Mater.
Process. 27, 249 (2008).
[15] R. Bidulsk, M. Actis Grande, Z. Brytan, M. Rosso,
Mater. Sci. Forum 672, 293 (2011).
[16] N. Candela, F. Velasco, J.M. Torralba, Mater. Sci.
Eng. A 259, 98 (1999).
[17] M. Campos, L. Blanco, J.M. Torralba, Powder Metall.
Prog. 7, 12 (2007).
[18] R. Bidulsk, M. Actis Grande, J. Bidulsk,
T. Kvakaj, Mater. Tekhnol. 27, 249 (2008).
[19] H. Karlsson, L. Nyborg, S. Berg, Powder Metall. 48,
51 (2005).
[20] M.W. Wu, K.S. Hwang, H.S. Huang, Metall. Mater.
Trans. A 38, 1598 (2007).
[21] G. Krauss, Metall. Mater. Trans. A 32, 861 (2001).
[22] V. Kerlins, A. Phillips, in: Fractography, 9th ed., Metals Handbook, ASM International, Ohio 1987, p. 12.
[23] S.P. Lynch, Acta Metall. 32, 79 (1984).
[24] A.W. Thompson, Fatig. Fract. Eng. Mater. Struct.
19, 1307 (1996).
[25] M.L. Martin, J.A. Fenske, G.S. Liu, P. Sofronis,
I.M. Robertson, Acta Mater. 59, 1601 (2011).
[26] U. Engstrm, D. Milligan, A. Klekovkin, Adv. Powder
Metall. Particul. Mater. 7, 21 (2006).
[27] T. Marcu Puscas, M. Signorini, A. Molinari, G. Straffelini, Mater. Charact. 50, 1 (2003).