Anda di halaman 1dari 6

J. Mater. Sci. Technol., Vol.23 No.

5, 2007

659

Effect of Microstructure Refinement on the Strength and Toughness


of Low Alloy Martensitic Steel
Chunfang WANG , Maoqiu WANG, Jie SHI, Weijun HUI and Han DONG
National Engineering Research Center of Advanced Steel Technology, Central Iron and Steel Research Institute, Beijing
100081, China
[Manuscript received October 16, 2006, in revised form December 25, 2006]

Martensitic microstructure in quenched and tempered 17CrNiMo6 steel with the prior austenite grain size
ranging from 6 m to 199 m has been characterized by optical metallography (OM), scanning electron
microscopy (SEM) and transmission electron microscopy (TEM). The yield strength and the toughness of the
steel with various prior austenite grain sizes were tested and correlated with microstructure characteristics.
Results show that both the prior austenite grain size and the martensitic packet size in the 17CrNiMo6 steel
follow a Hall-Petch relation with the yield strength. When the prior austenite grain size was refined from
199 m to 6 m , the yield strength increased by 235 MPa, while the Charpy U-notch impact energy at
77 K improved more than 8 times, indicating that microstructure refinement is more effective in improving
the resistance to cleavage fracture than in increasing the strength. The fracture surfaces implied that the unit
crack path for cleavage fracture is identified as being the packet.
KEY WORDS: Martensitic steel; Grain refinement; Strength; Impact toughness; Cleavage fracture

1. Introduction
The lath martensite structure is one of the most
important structures in steels since commercial, heattreated low alloy structural steels usually have lath
type martensitic structures. The lath type martensitic structure is composed of packets of parallel laths
within the prior austenite grain[1,2] . A prior austenite
grain is divided into several packets.
It is well known that strength and toughness of
martensitic steels improve as the grain size is refined.
Because of complicated microstructures, the effective
grain size to the strength and toughness is not entirely clear for martensitic steels[36] . Many investigators have shown that the prior austenite grain size
plays a role in controlling the strength and toughness of steels having lath martensite. For example, a
Hall-Petch type relationship was observed to exist between the prior austenite grain size and the strength
of several steels[712] . However, some other investigators regarded that the yield strength and toughness are dependent on the packet size, but not on the
austenite grain size[1316] . Naylor once claimed that
the lath width is the basic parameter controlling the
yield stress and the ductile-to-brittle transition temperature of Fe-Mn and Fe-Mn-Cr low carbon steels
with lath martensite or baintic structures[17] .
The above survey of literature show the complex
of microstructure on the mechanical properties of lath
martensitic steels. Therefore, the purpose of the investigation is to examine the effect of microstructure
refinement on the strength and toughness of low carbon martensitic steels.
2. Experimental
A commercial 17CrNiMo6 steel was used in

Ph.D. Candidate, to whom correspondence should be addressed, E-mail: fang2004@126.com.

this investigation, with the chemical composition in


wt pct as follows: C 0.17, Si 0.21, Mn 0.55, P 0.011,
S 0.002, Cr 1.80, Ni 1.58, Mo 0.25. The steel was
provided as hot-rolled round bars with a diameter
of 70 mm. These round bars were further reheated
and forged into round bars with a smaller diameter
of 16 mm, followed by annealing to facilitate machining. Tensile specimens with a diameter of 5 mm
and Charpy U-notch specimens with dimensions of
10 mm10 mm55 mm were machined from the
round bars.
The specimens were austenitized at 11331473 K
for 1 h in air furnace and then quenched into water, followed by tempering at 443 K for 2 h. The
mean prior austenite grain sizes of the specimens that
were austenitized at 1133, 1213, 1323, 1373, 1423 and
1473 K were 11, 17, 26, 50, 129 and 199 m, respectively. Cyclic heat treatment of the specimens was
also carried out at 1133 K for 4 cycles in salt bath furnace, resulting in a fine austenite grain size of 6 m.
The tensile test was performed on a WE-300 tensile machine at room temperature (298 K) at a strain
rate of 2.5103 2.5104 /s in accordance with the
standard GB/T 228-2002. The impact test was performed on a JBN-300B impact machine with a hammer velocity of 5.5 m/s at 77 K in accordance with
the standard GB/T229-1994.
The microstructure was characterized by optical
metallography (OM), scanning electron microscopy
(SEM), and transmission electron microscopy (TEM).
A HITACHI S-4300 type scanning electron microscope and a HITACHI H-800 type transmission electron microscope were used. To reveal the prior
austenite grains, the specimens were etched with a
supersaturated picric acid aqueous solution after mechanical polishing. The prior austenite grain sizes in
the 17CrNiMo6 steel were determined by linear intercept measurements on optical micrographs. Then,
the specimens were etched with a 2% nital to delineate
the lath microstructure within the prior austenite

660

J. Mater. Sci. Technol., Vol.23 No.5, 2007

Fig.1 Optical micrographs (a) and (c) showing the austenite grains and the corresponding SEM micrographs (b)
and (d) showing martensitic packets in (a) and (c) with the prior austenite grain sizes of 50 and 11 m,
respectively

Fig.2 TEM bright field images showing martensitic laths in the 17CrNiMo6 steel: (a) austenized at 1473 K and
(b) cyclic heat treated at 1133 K

grains. Packet sizes were measured by linear intercept


measurements on SEM images. Fracture surfaces of
the impact specimens were observed on a scanning
electron microscope. Thin-foil specimens for TEM observations were cut from the specimens and prepared
by twin-jet electrolytic polishing using a solution of
10% perchloric acid and 90% ethanol at 243 K.
3. Results and Discussion
3.1 Microstructure characterization
Figure 1 shows the optical and SEM micrographs
of the specimens with the prior austenite grain sizes of
50 m (Fig.1(a) and (b)) and 11 m (Fig.1(c) and (d))
for the 17CrNiMo6 steel. In the steel, packets were
distinctly present within the prior austenite grains, as

shown in Fig.1(b) and (d). The laths formed within a


prior austenite grain grouped themselves into several
packets. The martensitic packet size decreased with
the refinement of the prior austenite grains.
Figure 2 shows TEM bright field images of the
specimens which were austenized at 1473 K (Fig.2(a))
and cyclic heat treated at 1133 K (Fig.2(b)) to obtain
austenite grain size of 199 and 6 m, respectively.
Dislocations were observed within the laths of both
specimens. From TEM observations, the lath width
was estimated to be about 0.3 m, which was almost
independent of the prior austenite grain size, as reported by Roberts[7] .
Figure 3 shows the packet size distribution fraction for the specimens with the prior austenite grain
sizes of 6, 17, 26 and 199 m, respectively. About 200

661

J. Mater. Sci. Technol., Vol.23 No.5, 2007

Fig.3 Distribution of martensite packet size in 17CrNiMo6 steel with the prior austenite grain sizes: (a) 6 m,
(b) 17 m, (c) 26 m and (d) 199 m

packets in each specimen were measured from the


SEM images and the average packet size was then
obtained. Figure 4 shows the relation between the
packet size (dp ) and the prior austenite grain size
(d ). The martensite packet size decreased with the
refinement of the prior austenite size, with the average
martensite packet size decreased from 109 m to 4 m
as the austenite grain size decreased from 199 m to
6 m.

Fig.4 Relationship between the packet size and the prior


austenite grain size in the 17CrNiMo6 steel

3.2 Influence of microstructure refinement on the


yield strength
Figure 5 shows the true stress vs true strain curves
for the specimens with the prior austenite grain sizes
of 199, 50 and 6 m. It can be seen that the total
elongation increased with the refinement of the prior
austenite grains.
Figure 6 shows the Hall-Petch type plots of the
yield strength as a function of the prior austenite size
(Fig.6(a)) and the packet size (Fig.6(b)), respectively.
Generally the yield strength of metals is given by the
classic Hall-Petch relation:
y = 0 + ky d1/2

Fig.5 True stress vs true strain curves for the specimens


with the prior austenite grain sizes of 199, 50 and
6 m

(1)

where 0 and ky are constants with ky being the HallPetch slope, and d is the mean effective grain size.
It can be seen from Fig.6 that a linear relationship
existed between the yield strength (Rp0.2 ) and the
reciprocal of the square root of the prior austenite
1/2
grain size (d ) as well as between the yield strength
and the reciprocal of the square root of the packet
1/2
size (dp ), indicating the dependence of the yield
strength of the 17CrNiMo6 martensitic steel on the
prior austenite size and on the packet size.

662

J. Mater. Sci. Technol., Vol.23 No.5, 2007

Fig.6 Dependence of the yield strength on (a) the prior austenite grain size and (b) the martensite packet size
for the 17CrNiMo6 steel

Fig.7 Effect of (a) the prior austenite grain size and (b) the martensite packet size on the impact energy AKU2
at 77 K for the 17CrNiMo6 steel

Fig.8 Fracture surfaces of the impact specimens tested at 77 K for the prior austenite grain sizes of (a) 6 m,
(b) 17 m, (c) 26 m and (d) 199 m

663

J. Mater. Sci. Technol., Vol.23 No.5, 2007

Table 1 Relationship between the grain size and packet size and
arithmetic mean of the cleavage facet size for the 17CrNiMo6 steel
Grain size, d /m
199
129
50
26
17
11
6

Packet size, dp /m
109
31
19
12
9
7
4

Fig.9 Profile fractograph of the fractured impact specimen in the 17CrNiMo6 steel with the prior austenite grain size of 199 m

In a previous study of Fe-0.20C-Ni-Cr-Mo steel,


Tomita regarded the packet size as the effective grain
size for the lath martensite from the results that
there was a Hall-Petch relationship between the yield
strength and the packet size, but no such relationship
between the yield strength and the prior austenite
grain size[15] . The present results showed that it was
difficult to tell that for the 17CrNiMo6 steel whether
the packet size or the prior austenite grain size was
the effective grain size simply from the Hall-Petch plot
since both sizes had a Hall-Petch relationship with the
yield strength.
As the prior austenite grain size in the 17CrNiMo6 steel was refined from 199 m to 6 m, the yield
strength increased by 235 MPa, being only a 25% increase, indicating that the grain refinement was not
very effective in increasing the strength. The data for
the Fe-0.2C martensitic steel[13] and the Fe-0.20C-NiCr-Mo martensitic steel[15] are also included in Fig.6.
There was good agreement between the results for
the 17CrNiMo6 steel and for the Fe-0.20C-Ni-Cr-Mo
steel, which was presumable due to similar carbon
and alloy contents in the two steels. The dependence
of the yield strength on the packet size was not so
strong for the two low alloy steels, as indicated by
the low slope of the Hall-Petch plot, although the Fe0.2C steel showed weaker dependence. Accordingly,
the microstructure refinement was not so effective in
increasing the strength for lath martensitic steels.
3.3 Influence of microstructure refinement on the impact toughness
Figure 7 shows the relationship between the prior
austenite grain size and the Charpy impact energy
(AKU2 ) at 77 K for the 17CrNiMo6 steel. The impact

Cleavage facet size/m


96
26
17
13
9
7
5

energy value increased with decreasing prior austenite grain size and it increased more than 8 times from
5.7 J to 46.7 J when the prior austenite grain size
was refined from 199 m to 6 m, indicating that
grain refinement contributed largely to the toughness
of martensitic steels.
Figure 8 shows the fracture surfaces of the impact
specimens with the prior austenite grain sizes of 6, 17,
26 and 199 m. The fractured surfaces of the 17CrNiMo6 steel consisted of quasi-cleavage facets, and complex river patterns consisting of small cleavage steps,
were also observed within the facets. In this case,
a quasi-cleavage facet was defined as the spacing divided by heavy tear lines. The arithmetic mean of
the observed facet size was determined by linear intercept method on SEM images, and the results is
given in Table 1. For the specimens with the prior
austenite grain sizes ranging from 199 m to 6 m,
the arithmetic mean of the quasi-cleavage facet size
was decreased from 96 m to 5 m, which was corresponding to the measured values of the martensite
packet size of from 109 m to 4 m, respectively. It
implied that the dominant microstructural feature in
the toughness of the martensite was the packet size.
Further evidence of the direct relationship between
the packet size and the facet size is shown in Fig.9,
where the broken impact specimen with the prior
austenite grain size of 199 m was observed along the
longitudinal direction after it was cut, polished and
etched by 2% nital. The cleavage crack usually largely
changed its direction at the packet boundaries, due to
the fact that a local cleavage cracking event across a
large packet could lead to an instant load drop, but
when the crack attempted to run from one packet
to another, the crystallographic orientation and microstructure changed, so that more work had to be
done to cross the packet boundary. As a result, the
local cleavage crack might be arrested.
From the above results, it follows that grain refinement was very effective in improving the resistance to
the cleavage fracture, but was relatively ineffective in
increasing the strength.
The basis on the classic pile-up model of the
Hall-Petch relation gives,

ky 3

2Gbb
q

1/2
(2)

where q is a geometric factor of order unity; G is the


shear modulus; b is the Burgers vector; and b is the
critical shear stress for the transmission of slip across
a grain boundary. The Burgers vector is fixed by the
crystal and the shear modulus can only be changed
by significantly changing the composition of the steel.

664

J. Mater. Sci. Technol., Vol.23 No.5, 2007

The controllability of the Hall-Petch coefficient in this


equation is the variable parameter b . This parameter is affected by the nature of the grain boundary.
In iron, the primary slip planes are the 110 planes. It
is expected to increase with the misorientation of the
dominant <110> slip planes. Increasing the typical
misorientation should increase b , hence ky , while decreasing misorientation should decrease ky . To maximize the strengthening that can be accomplished by
grain refinement, ky should be as large as possible.
For iron cleaves on {100} planes, as has been
proved[18,19] , the appropriate measure of the grain
size for cleavage should be the coherence length on
{100}. Toughness should increase with the misorientaion of {100} planes across the boundaries. Since
the lath boundaries tend to lie parallel to {110}, the
laths share common {100} planes that cut through
the boundaries. In present study it shows that impact energy increased more than 8 times and the
yield strength increased only 25%. So transformation
for refining the grain size in present study benefited
against cleavage as the misorientation along {100}
planes was significant, however it might have only
slight misorientation of the [110] slip planes. As a
result, grain refinement was very effective in improving the resistance to the cleavage fracture, but was
relatively ineffective in increasing the strength.
4. Conclusions
(1) There is a Hall-Petch relationship between the
yield strength and the prior austenite grain size and
the packet size for the 17CrNiMo6 steel. The strength
and impact toughness of the 17CrNiMo6 steel can be
improved with the refinement in the prior austenite
size and the packet size.
(2) Microstructure refinement is more effective in
improving the resistance to the cleavage fracture than
in increasing the strength. The prior austenite grain
size in the 17CrNiMo6 steel was refined from 199 m
to 6 m, the yield strength increased by 235 MPa, being only a 25% increase and the impact energy value
increased more than 8 times.

(3) The cleavage crack usually largely changed its


direction at the packet boundaries. The unit crack
path for cleavage fracture was identified as being the
packet.

Acknowledgements
The authors would like to thank Mr. Wenhan Zhang
for help in the experiment of revealing the prior austenite grain size. This work was supported by the General
Armaments Department Beforehand Research Foundation
(No. 9140A12050306QT0901).
REFERENCES
[1 ] T.Maki, K.Tsuzaki and I.Tamura: Trans ISIJ, 1980,
20(4), 207.
[2 ] S.Morito, H.Tanaka, R.Konishi, T.Furuhara and
T.Maki: Acta Mater., 2003, 51(6), 1789.
[3 ] G.Krauss: Mater. Sci. Eng. A, 1999, 273-275, 40.
[4 ] G.Krauss: ISIJ Int., 1995, 35(4), 349.
[5 ] L.Ryde: Mater. Sci. Tech., 2006, 22(11), 1297.
[6 ] C.K.Yao and Z.Xu: Mater. Chem. Phys., 1986, 14(6),
559.
[7 ] M.J.Roberts: Metall. Trans., 1970, 1(12), 3287.
[8 ] A.Di.Schino and J.M.Keny: Mater. Lett., 2003,
57(12), 1830.
[9 ] M.Y.Liu, B.Shi, C.Wang, S.K.Ji, X.Cai and
H.W.Song: Mater. Lett., 2003, 57(19), 2798.
[10] R.Ishibashi, H.Arakawa, T.Abe and Y.Aono: ISIJ Int.,
2000, 40(Suppl.), 169.
[11] C.C.Anya and T.N.Baker: Mater. Sci. Eng., 1989,
A118, 197.
[12] H.J.Rack: Mater. Sci. Eng., 1978, 34(3), 263.
[13] T.Swarr and G.Krauss: Metall. Trans. A, 1976,
7A(1), 41.
[14] L.A.Norstrom: Scand. J. Metall., 1976, 5(4), 159.
[15] Y.Tomita and K.Okabayashi: Metall. Trans. A, 1986,
17A(7), 1203.
[16] T.Inoue, S.Matsuda, Y.Okamura and K.Aoki: Trans.
JIM, 1970, 11(1), 36.
[17] J.P.Naylor: Metall. Trans. A, 1979, 10A(7), 861.
[18] V.Randle and P.Davies: Mater. Sci. Tech., 2005,
21(11), 1275.
[19] R.Ayer, R.R.Mueller and T.Neeraj: Mater. Sci. Eng.
A, 2006, 417(1-2), 243.

Anda mungkin juga menyukai