IntJ FatiguelO No 4 (1988) pp 249-255

T h e e f f e c t of m i c r o s t r u c t u r e on the
m e c h a n i c a l b e h a v i o u r of a l o w
carbon, low alloy steel
M.T. Yu, T.H. T o p p e r and L. W a n g

The effect of microstructure on the mechanical behaviour of a low carbon, low alloy
steel was studied. The hot-rolled ferrite-pearlite showed low monotonic and cyclic
strengths with high ductility in terms of true fracture strain and a high threshold stress
intensity. The quenched and tempered low carbon martensite showed high monotonic
and cyclic strengths and high ductility. However, the threshold stress intensity was
significantly lower than that of the ferrite-pearlite. Both the strength and threshold
stress intensity of an austempered bainite and a duplex ferrite-martensite are greater
than those of the hot-rolled ferrite-pearlite. At a same strength level, the ductility and
the threshold stress intensity of the low carbon martensite are higher than those of
a medium carbon martensite.

Key words: microstructure; mechanical behaviour; cyclic strength; ferrite-pearlite;

ferrite-martensite; austempered bainite; low carbon, low alloy steel

Heat treatment is widely used in engineering to change the may be useful in finding applications for heat treated, low
microstructures of medium and high carbon steels. However carbon steels and in specifying heat treatment parameters
quenching-tempering or austempering is generally not ap- for carburized steels.
plied to low carbon steels (except after carburizing or carbo-
nitriding) since the increase in hardness or strength obtained
by heat treatments is considered to be insignificant. It has
been shown however, that a combination of high strength
M a t e r i a l and t e c h n i q u e s
and ductility can be obtained by quenching and tempering
low carbon steels, t An increase in both yield strength and The SAE8620 low carbon, low alloy steel used (Table 1s)
crack propagation threshold stress intensity has also been contains a small amount of alloy elements to increase harden-
observed 2 with a low carbon steel, heat treated to give a ability. Two types of specimen were used: cylindrical smooth
duplex ferrite-martensite microstructure. and centre notched plate. Blanks were heat treated and
The commercial potential of low carbon steel lies in machined to the dimensions shown in Figs. 1 and 2. To
its high ductility which allows commercial plastic forming account for engineering application and cost, the heat treat-
processes to be used rather than machining. If the overall ment of the specimens was made as simple as possible. The
mechanical properties of a heat treated low carbon steel are heat treatments and the resulting microstructures are given
similar or even better than those of a medium carbon steel, in Table 2. The quantities of the microstructure constituents
plastic forming processes may give significant cost reductions were roughly estimated using an optical microscope.
for a variety of components, especially for mass production All tests were performed in a laboratory using a closed-
car parts. loop servo-controlled electrohydraulic test system. A sinu-
The mechanical properties of carburized steels depend soidal waveform was used and the test frequency ranged
on the microstructures of the case and the core. H o w these from 10 to 70 Hz depending on the applied stress or strain
microstructures will affect the mechanical behaviour of the amplitude. All crack propagation tests were performed at
core is important in specifying heat treatment parameters a stress ratio R equal to zero. The K-decreasing procedure
for carburized steels. suggested by Bucci6 was used to minimize the high-to-low
The monotonic and fatigue properties of SAE8620 low load interaction effect. The sequence of load reduction
carbon, low alloy steel with various microstructures were followed by crack growth was repeated until no growth
investigated and compared with those of SAE1045 medium could be detected after 3-5 million cycles and the threshold
carbon steel which has been already studied. 3"4 The results stress intensity was determined. The load was then increased

Table 1. Chemical composition (weight %)

C Mn P S Me Si Cr Ni Fe
0.18-0.23 0.70-0.90 0.035 0.040 0.15-0.25 0.15-0.30 0.40,-0.60 0.40-0.70 Balance

0142-1123/88/040249-07 $3.00 © 1988 Butterworth & Co (Publishers) Ltd

Int J Fatigue October 1988 249
 1 2 . 7 0 - 2 0 UNF 13.05 R 13.05R Table 2. Heat t r e a t m e n t

i I
Designation Heat treatment Microstructure
hardness (Rc)
5.08 D
DAR As-received Ferrite- pearlite
T200 Austenized at 900°C Tempered
I
llIl I I l l l l l l l l l
f
I IIIIIIIllllllllllll for 20 min,
quenched in brine,
tempered at 200°C
martensite
44

for 60 mins, air

cooled,
41.91 = • 41.91
T400 As T200, tempered Tempered
at 400°C martensite
7.62
40
T600 As T200, tempered Tempered
at 600°C martensite
114.30 32.6
D730 Heated to 730°C for 30% martensite +
40 min, quenched in 70% ferrite
Fig. 1 Smoothspecimendesign.All dimensionsin mm brine, tempered at 26.3
200°C for 60 min, air
cooled
D750 Heated to 750°C for 50% martensite +
40 min, quenched in 50% ferrite
brine, tempered at 33
44.55 200°C for 60 min, air
2.0 cooled
D975 Heated to 900°C for 50% martensite +
20 min, cooled to 50% ferrite
--22.86 1.02 D 22.23 750°C and held for 32.4
50 min, quenched in
brine, tempered at
116.84
200°C for 60 min, air
r cooled

Fig. 2 Geometryand dimensionsof centre-notchedspecimen B440 Heated to 900°C for Bainite

20 min, quenched in 29.4
440°C salt bath for
30 min, air cooled
slightly (by 5-10%) and held constant while the test
continued in a K-increasing mode.
Two methods were used to measure crack closure: microstructure on strength and ductility is significant. The
yield strength of the quenched and tempered martensite T200
• a compliance determination by measuring the displace-
increased by a factor of almost three, and the ultimate tensile
ment of the crack borders at the centre of the specimen
strength increased by a factor of more than two with little
with a mechanical extensometer with a gauge length
loss of ductility compared to that of the hot-rolled ferrite-
of 2 mm;
pearlite (DAR). The yield strength and the ultimate tensile
• a compliance determination with a strain gauge bonded
strength of the T400 steel are significantly higher than those
on both sides of the crack. The load and displacement
of the D A R steel without any loss in ductility.
signals were input to a differential amplification circuit
It is generally true that the ductility of a steel is decreased
and then fed to a X - Y plotter. when the strength is increased and vice versa. However,
the low carbon martensite (LCM) T600 allows an increase
The crack closure levels measured from the exten- in both strength and ductility. Both strength and ductility
someter and strain gauges were in agreement (Fig. 3). Since of the austempered bainite increased compared with those
the strain gauge method is much more involved and gauges of the ferrite-pearhte. However the strength increase is not
sometimes break during a test, most of the measurements so significant as that of the low carbon martensites. A mild
were made using the clip-gauge extensometer. increase in strength and decrease in ductility was observed
with duplex phase ferrite-martensite microstructures.
The mechanical properties of the low carbon martensite
T200 are compared with those of a medium carbon marten-
Monotonic mechanical properties site (MCM) T482 (SAE1045 steel quenched and tempered
Monotonic tensile tests with cylindrical smooth specimens at 315, 482 and 649°C, designated as T315, T482 and T649
gave the yield strength, ultimate tensile strength, area respectively) in Table 4. Both the strength and ductility (in
reduction and true fracture strain for each microstructure terms of true fracture strain) of the low carbon martensite
(Table 3). From the data it can be seen that the effect of are higher than those of the medium carbon martensite.

250 Int J Fatigue October 1988

 Table 3. Mechanical properties
Material Yield Ultimate Area True Fatigue
strength strength reduction fracture limit
(M Pa) (M Pa) % strain (M Pa)
In(Ao/Af)
DAR 464 657 62.9 0.992 290
T200 1260 1477 54.8 0.794 670
T400 1142 1222 62.0 0.973 560
T600 853 897 70.0 1.21 430
D730 592 1023 39.6 0.504
D750 691 1123 46.0 0.616 440
D975 707 1179 27.4 0.32 430
B440 621 749 70.6 1.223
T315 1617 1984 11.4 0.103 593
T482 1049 1180 47.4 0.643 580
T649 644 794 63.5 1.007 380

50
Table 4. Comparison of mechanical properties

(-~a)MP Su ~, S_, A/~h

(MPa) In(Ao/Af) (MPa) (Mpax/~)
Extensometer
40- T482 1049 1180 0.643 580 3.5
Strain gauge T200 1260 1477 0.749 670 4.4
D750 691 1123 0.616 440 8.56

30-
Q.
:g
20001 . . . . . . .

uq

2°l
T000

°I
o
Differential displacement
iii 102 103 104
,

105
,

106 107
,

Cycles to failure
Fig. 3 Crack closure behaviour of DAR steel in the near threshold
region Fig. 4 Stress amplitude versus number of cycles to failure

Smooth specimen fatigue behaviour

were tested. The crack propagation rate da/dn is plotted
Smooth specimen fatigue tests were performed under strain against stress intensity range, AK, in Fig. 5. The threshold
control at a stress ratio R = - 1 (Fig. 4). At all lives stress intensity is about 8.4 MPax/~. The crack propagation
from low cycle to high cycle fatigue, the fatigue strength rates for the other microstructures versus stress intensity
of the T200 steel is increased by a factor of more than two range are plotted in Fig. 6.
compared with the D A R steel. The fatigue strengths of the The B440 bainite and D730 and D750 duplex phase
T400, T600, D750 and D975 steels are also increased. The martensite-ferrite showed a higher threshold stress intensity
fatigue limit of the T200 low carbon martensite is higher than that of the as-received ferrite-pearlite. However the
than that of the T315 medium carbon martensite 4 even though tempered martensites showed lower threshold stress inten-
the monotonic strength and hardness of the latter are higher. sities than that of the as-received ferrite-pearlite. The
The fatigue strength of T400 steel low carbon martensite threshold stress intensity of the duplex phase D975 steel
is comparable to that of the T482 medium carbon steel. is lower than the other duplex-phase steels but higher than
those of the tempered martensites.
Crack closure was measured near the threshold region
Threshold stress intensity and crack and the measured threshold stress inensity, the crack opening
closure stress intensity and the calculated effective threshold stress
intensity are listed in Table 5. It can be seen that the high
Four specimens of the as-received hot-rolled ferrite-pearlite threshold stress intensities of DAR, D730, D750 and B440

Int J Fatigue October 1988 251

 10 -8 10 -8

10 .9 u 10-9
E
"D
"o
"D

O D730
• D750
10 -10 10 -10 D975
1"3 B440
• DAR I • T600
O DAR 2 T400
• DAR 3 I T200
~7 D A R 4

I I I I i I~1] I I I I I III
10 100 1 10 100

A K (MPa ¢'m) K (MPa / m )

Fig. 5 Crack propagation rate versus stress intensity range for the Fig. 6 Crack propagation rate versus stress intensity range for various
as-received steel microstructure conditions

T a b l e 5. T h r e s h o l d stress i n t e n s i t y ( M Pa~/-m) I separation at the crack tip of the material involving a

ductile tearing process.
Material AKth AKop AKeff The fracture surfaces of DAR, D730, D750 and D440
are rougher than those of T200 and T400. A rough fracture
DAR 8.4 4.2 4.2 surface may contribute to a so-called roughness-induced
D730 9.0 4.4 4.6 crack closure. 7's The fracture surface of B440 shows
D750 8.5 4.0 4.5
orientations of bainite packages in the fracture surface plane.
D975 6.2 2.6 3.6
B440 8.5 4.2 4.3 Tearing of acicular ferrite is distinguishable (Fig. 7c and d).
T200 4.4 0.5 3.9 The fracture surface of the tempered martensite (Fig.
T400 4.2 7g and h) is smoother than those of all the duplex phases
T600 4.6 and bainite; this results in a lower roughness-induced crack
closure stress intensity 7.
T315 3.76 0 3.76 Facets, or microcleavage, 9 can be intergranular or trans-
T482 3.5
T649 3.86 granular. Transgranular facets develop along certain
crystallographic planes, ie (001) in the FCC and (0001) in
the HCP crystalsA° The models proposed for the formation
are related to their high crack opening stress intensities. of facets involve alternating slip activity. When the reverse
The variation in effective threshold stress intensity for these plasticity at the crack tip is contained within individual grains,
materials in various microstructural conditions is not the anisotropy of slip activity will produce a facet with a
significant. plane determined by the orientation of the grain. In a poly-
crystalline material with a random grain orientation, facets
of varying orientations to the macroscopic crack plane would
be expected. The development of the facets is associated
Fractography with mode II fracture concurrently with the change from
Fracture surfaces were examined with a scanning electron 'microstructurally insensitive' to 'microstructurally sensitive'
microscope. The fracture surfaces near the threshold stress crack growth.
intensity region of all material conditions show a combination Transgranular facets were observed on D A R and 13440
of 'ripples', 'furrows' and 'facets' (Fig. 7). Ripples refer to steel specimens. Intergranular facets, which increased with
a shallow waved and furrows to a deep waved appearance the stress intensity, were observed on the tempered marten-
of a fracture surface. These characteristics may relate to mode site (Fig. 7h). The presence of intergranular facets may be

252 Int J Fatigue October 1988

Fig. 7 Fractography: (a) DAR, AK = 9 MPa~/'~ (b) DAR, high AK = 25 MPa~/-m; (c) B440, AK~ = 8.8 MPa~/-m; (d)
16 MPa~/'~" (e) D730, AK = 9.18 MPa~/-~ (f) 0730, AK = 14.3 MPa~//~ (g) T400, A/~ = 4.4 M P a ~ " (h) T400, BK440-AK =
M Pa~/'~ - 13.32

Int J Fatigue O c t o b e r 1 988 253

related to impurity induced grain boundary segregation, tl SAE1045 medium carbon steel. 3 It is generally accepted that
The weakened grain boundary decreases the resistance of when yield strength is increased, the threshold stress intensity
the material to Km,~, ie normal separation occurs before shear is decreased. This general trend is not observed in the three
separation, resulting in a low threshold stress intensity. low carbon martensites. The yield strength of the T200 steel
is 50% higher than the T600 steel, but the threshold stress
intensity is almost same for the two steels.
An overall improvement in mechanical properties is
Discussion obtained by austempering. The yield strength, ultimate
tensile strength, true fracture strain and the threshold stress
A combination of high strength and high ductility can be intensity of the B440 steel are increased compared with those
obtained by quenching and tempering a low carbon steel, of the D A R steel. The strength increase is minor. Martemper-
The mechanical properties, including yield strength, ultimate ing will increase strength more significantly although the
tensile strength, true fracture strain, fatigue limit and thres- effect of martempering on threshold stress intensity is not
hold stress intensity of the low carbon martensite T200 and clear. Further research is needed.
the dual phase steel D750 are compared with those of the The monotonic and cyclic strengths for the D750 and
medium carbon martensite T482 in Table 4. The medium D975 steels are roughly equal. However the threshold stress
carbon martensite T482, which represents a typical heat intensity of the D750 steel is higher than the D975 steel.
treatment condition for many machinery and automobile This suggests that the threshold stress intensity is more sensi-
components, shows a good combination of strength and tive to the distribution of microstructure constituents than
ductility. Compared with the T482 medium carbon marten- the strength. More distinguishable continuous martensite was
site, the low carbon martensite T200 gives not only a higher observed in the D750 steel than in the D975 steel.
monotonic strength and fatigue limit, but also has a higher In summary, a high strength and a high ductility can
ductility and threshold stress intensity. The ultimate tensile be achieved by quenching and tempering a low carbon steel.
strength and the true fracture strain of the dual phase D750 At a given level, low carbon martensite has a higher ductility
steel are comparable to those of the T482 medium carbon and a higher threshold stress intensity than medium carbon
martensite, but the yield strength and fatigue limit are lower. martensite.
However the dual phase D750 steel has a significantly higher Overall mechanical properties can also be improved by
threshold stress intensity than the T482 medium carbon austempering. A dual phase structure obtained by heating
martensite. into the ferrite-austenite region followed by quenching and
The practical importance of the results lies not so much tempering increases strength while retaining a high threshold
in the somewhat improved mechanical properties obtained stress intensity. There is no direct relationship between
with the low carbon martensite or dual-phase steel as with macroplasticity and threshold stress intensity.
the high ductility of the low carbon steel in the as-received
ferrite-pearlite condition. This allows low carbon steel to
be made into components by inexpensive commercial plastic
forming processes. Medium carbon steels which have a lower
Conclusions
as-received ductility cannot be plastically formed and must
be machined at a higher fabrication cost. 1) A combination of high strength and high ductility was
The crack propagation threshold stress intensity of the obtained by quenching and tempering a low carbon,
low carbon martensite is lower than that of the ferrite-pearlite low alloy steel. At the same strength level, the low
or the ferrite-martensite. There is no direct relationship carbon martensite showed a higher ductilty and thres-
between threshold stress intensity and macroplasticity, in hold stress intensity than a medium carbon martensite.
terms of area reduction or true fracture strain, which is 2) Through low cycle to high cycle fatigue, the fatigue
thought to decrease stress concentration and increase crack strength of tempered low carbon martensite T200 was
propagation resistance. The true fracture strain of the increased by a factor of two compared with that of
tempered T600 low carbon martensite is much higher than the as-received ferrite-pearlite. The fatigue limit of the
that of the dual phase steels D730, D750 and DAR. The T200 low carbon martensite was higher than that of
threshold stress intensity of the T600, however, is only about the T315 medium carbon martensite even though the
half that of these dual-phase steels. The threshold stress inten- hardness and the monotonic strength of T315 steel is
sity of the D975 steel is significantly higher than that of higher than that of the T200 steel.
the low carbon martensites even though the true fracture 3) The duplex phase ferrite-martensite obtained by heating
strain of the former is only about one third of the latter. the low carbon steel to the ferrite-austenite region and
The free ferrite in the as-received D A R steel and the then quenching it gave a higher strength than that of
heat treated dual-phase steels D730, D750 and D975 seems the as-received ferrite-pearlite. It also retained a high
to be a controlling factor of threshold crack propagation. threshold stress intensity.
A possible role played by free ferrite is that of blunting 4) Increases in strength, ductility and crack propagation
the crack tip and increasing fracture surface roughness, hence threshold stress intensity were achieved by austempering.
increasing crack closure. The data in Table 4 show that 5) Microstructure has a strong effect on crack closure level,
the effective threshold stress intensities for the various and hence, on the threshold stress intensity. Ferrite-
material conditions are about the same. The high nominal pearlite, ferrite-martensite and bainite showed a high
threshold stress intensity of the dual-phase steels comes from crack closure stress intensity. However, the tempered
their high opening stress intensity. martensite showed a low crack closure stress intensity.
The effect of tempering temperature on strength and 6) There is no direct relationship between macroplasticity
ductility is notable. However the effect of tempering tempera- and threshold stress intensity. The effect of yield
ture on threshold stress intensity is insignificant. This is strength (or tempering temperature) on threshold stress
in agreement with the results obtained from a heat treated intensity of the low carbon martensite is insignificant.

254 Int J Fatigue October 1988

Acknowledgement 6. Bucci, R. T. 'Development of a proposed standard practice
for near threshold fatigue crack growth rate measurement'
Financial support of this work by GKN Technology Limited ASTM E-9/E-24 Symp of Fatigue Crack Growth Measure-
and the government of PRC is gratefully acknowledged. ment and Data Analysis Pittsburgh, PA, USA 29 October 1979
7. McEvily, A. J. "Current aspects offatigue" Met Sci11 (1977)
pp 274-284
References 8. Beeven=, C. J., Bell, K., Carson, R. L. end Starke, E.
A. 'A model for fatigue crack closure" Engng Fract Mech
1. Sherman, A.M. and Davies R. G. 1"he effect of martansite 19 (1984) pp 93-100
content on the fatigue of a dual-phase steel' Int J Fatigue
31 (1951) pp 36-40 9. Ritcherde, C. E. and Lindley, T. C. ~ influence of stress
intensity and microstructure on fatigue crack propagation in
2. Minakawe, K. and McEvily, A. J. 'On the influence of ferritic materials' Engng Fract Mech 4 (1972) pp 951-978
microstructura on the threshold level for fatigue crack growth
in steels' Proc 5th Int Conf on Strength of Metals and Alloys, 10. Beeven=, C, J. 'Fatigue crack growth characteristics at low
Aachen, FRG, August 1979 ed P. Hassen et al (Pergamon stress intensities of metals and alloys' Fatigue 1977, University
Press, New York, 1979) of Cambridge, UK 28-30 March 1977 (EMAS, Warley, UK,
1977)
3. Yu, M. T. and Topper, T. H. 'The effects of material strength,
stress ratio and compressive overloads on the threshold 11. Ritchie, R. O. and Knott, J. F. 'Mechanism of fatigue
behaviour of a SAE1045 steel" ASME TransJEMT 107 (1985) crack growth in low alloy steel' Acta MetallZI (1973)
pp 19-25
4. Yu, M. T., Topper, T. H. and DuQuesne,/, D. L. "The
effect of microstructure on the fatigue behavior of a SAE1045 Authors
steel' in Proc Symp on Microstructure and Mechanical
Behavior of Materials (Xian, PRC, 21-24 October 1985) M.T. Yu and T.H. Topper are with the Civil Engineering
5. Metal Handbook ed. H. E. Boyer and T. L. Gall (ASM, Metals Department , University of Waterloo, Waterloo, Ontario,
Park, OH, USA, 1985). Canada and L. Wang is with Sian Jiaotong University, PRC.

Int J Fatigue O c t o b e r 1 988 255