You are on page 1of 12

International Journal of Fatigue 143 (2021) 106014

Contents lists available at ScienceDirect

International Journal of Fatigue


journal homepage: www.elsevier.com/locate/ijfatigue

Influence of plastic deformation in fatigue crack behavior in bainitic steel


M.C. Marinelli a, *, M.Balbi a, U. Krupp b
a
Instituto de Física Rosario – Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Universidad Nacional de Rosario, Bv. 27 de febrero 210 bis, 2000
Rosario, Argentina
b
IEHK - Institut für Eisenhüttenkunde / Steel Institute, RWTH Aachen University, Intzestraße 1, D-52072 Aachen, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: In the present paper, the behavior of fatigue microcracks is studied in a high-strength bainitic steel with the aim
LCF of identifying the microstructural features that influence the mode of initiation and propagation of microcracks.
Surface crack initiation For this study, a gradual monitoring of surface damage during low-cycle fatigue test at different plastic strain
Crack propagation
ranges was carried out by means of light optical and electron microscopy along with EBSD data analysis.
High strength bainitic steel
Moreover, the dislocation structure near the surface was analyzed and correlated with the crack initiation site.
The results showed a change with respect to cracking mechanisms from 0.2% plastic strain.

1. Introduction Sankaran et al. [20,21] revealed that microcracks nucleate in slip band
extrusions/intrusions on the surface under mixed mode (ductile and
There has been a considerable attention in the fabrication of high- brittle) in multiphase (ferrite/ bainite/ martensite) microstructures.
strength bainitic steels in recent years due to the many advantages Branco et al [18] reported that microcracks initiate along the slip band
that they put up in the marketplace, thanks to their good strength to and they grow through the grain boundaries and eventually form a long
weight ratio and good erosion resistance [1–3]. These steels are suitable crack by coalescence of short cracks.
for cold forming of structural and safety-related automotive compo­ Rementeria et al [22], using electron backscatter diffraction (EBSD),
nents. The high strength level gives potential for considerable weight reported that the crack initiation occurs at the surface on slip systems
reduction and a cost-effective way to produce energy-efficient vehicles. {1 1 0} 〈1 1 1〉 and identified the possible microstructural barriers for
In particular, high-strength bainitic steels are widely designed and crack propagation, such as block and packet boundaries. Meanwhile,
manufactured in bearing, gear and automobile industry as crash rein­ Mueller et al [23] identified the microstructural features governing the
forcement bars to protect against sidewise impact and for injection lines fatigue limit as the bainite block size, which is considered the main
(under pulsating loads) in common rail diesel engines [4–9]. obstacle for short fatigue crack growth.
Bainitic steels are processed using accelerated cooling in order to On the other hand, Nohava et al [24] analyzed secondary cleavage
obtain bainitic ferrite laths increasingly thin with a fine-scale dispersion cracks and they reported that those propagated mainly in the range of a
of retained austenite between the laths to achieve better mechanical bainite packet on cleavage crack planes {1 0 0} but also in slip planes and
properties and an extended fatigue life [10–14]. were arrested on high-angle twist type boundaries.
Although the fatigue life of high-strength bainitic steels with The aim of the present paper is to identify the microstructural fea­
different bainitic morphologies was analyzed by several authors tures that influence the mode of initiation and propagation of micro­
[15–19], fatigue damage mechanisms were not entirely understood. cracks during low-cycle fatigue at different plastic strain ranges. For this
It is well known that fatigue failure usually occurs as a result of purpose, a gradual monitoring of surface damage during the fatigue test
stress/strain concentration and crack development in a critical region. In is proposed by means of light optical and electron microscopy along with
previous studies, scanning electron microscopy (SEM) has been used to EBSD data analysis.
analyze the fracture surfaces, related to the fatigue behavior of the
bainitic steels. Some authors [9,17] have reported that fatigue micro­
cracks nucleate at the interface between the bainitic ferrite and retained
austenite regardless of the strain amplitude imposed. On the other hand,

* Corresponding author.
E-mail address: marinelli@ifir-conicet.gov.ar (M.C. Marinelli).

https://doi.org/10.1016/j.ijfatigue.2020.106014
Received 9 September 2020; Received in revised form 21 October 2020; Accepted 21 October 2020
Available online 26 October 2020
0142-1123/© 2020 Published by Elsevier Ltd.
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

2. Experimental methodology the misorientation angle is greater than 15◦ [25,26]. Fig. 1b represents
the misorientation angles corresponding to block and packet bound­
2.1. Material aries, resulting high-angle boundaries with misorientation angles of 45◦
and 60◦ . Each block is divided into sub-micrometer bainitic ferrite laths
The material employed in the present work is the bainitic steel with similar crystallographic orientation and an average width size of
16CrMnV7-7 with a nominal chemical composition as given in Table 1. 0.6 μm. This latter contains cementite particles and high dislocation
The material was supplied by Georgsmarienhütte GmbH (Germany) in density inside and a film-like retained austenite at lath/block boundaries
the form of cylindrical bars of 30 mm in diameter. According to the (Fig. 1c).
manufacturer, optimal material properties are achieved by isothermal Additionally, the X-ray diffraction (XRD) study revealed an amount
transformation to bainitic ferrite in the temperature range 300–500 ◦ C. of 5 vol% of retained austenite (Fig. 1d).

2.2. Test specimen preparation and low-cycle fatigue test 3.2. Low-cycle fatigue behavior

The fatigue tests were carried out on cylindrical specimens for The cyclic behavior of the bainitic steel 16CrMnV7-7 at room tem­
studying the fatigue behavior and shallow-notched cylindrical speci­ perature under plastic strain control for Δεp = 0.1%, Δεp = 0.2% and
mens for the cracks study [27]. The former ones were prepared by Δεp = 0.3% is shown in Fig. 2.
standard mechanical polishing finished with 1 μm diamond paste to All three curves present a short initial cyclic hardening stage, fol­
achieve a smooth surface. On the other hand, the latter ones were me­ lowed by a cyclic softening response. The initial cyclic hardening
chanically polished down to 6 μm diamond paste and then electrolyti­ behavior is commonly observed during LCF in carbide-free and carbide-
cally polished using a solution of 10% perchloric acid in ethanol at T = bearing bainitic steels [17,19,27] as a result of the entanglement of
− 15 ◦ C, V = 15 V, I = 2A for 4 min to improve the observation of dislocations causing reduction in the mobile-dislocation density, while
microcracks with respect to nucleation and growth. The central part of the cyclic softening is the result of dislocation rearrangement.
the notch was monitored and digitally recorded, during the fatigue test, In order to obtain the Coffin-Manson curve for the studied steel, fa­
using an in-situ light optical microscopy (LOM) system consisting of a tigue tests were performed at different plastic strain ranges Δεp = 0.1%,
CCD camera, JAI model CM-140MCL with a 50X objective, ± 1 μm FD 0.15%, 0.2%, 0.25% and 0.3%.
and 13 mm WD and a 12X ultra zoom. The Fig. 3 shows a bilinear behavior at Δεp / 2 = 1 × 10− 3. This
Low-cycle fatigue tests (LCF) were performed at room temperature in phenomenon was previously noted in carbide-free and carbide-bearing
an electromechanical testing machine INSTRON 1362 under plastic bainitic steels [15,19]. In fact, the authors reported that this behavior
strain control with plastic strain ranges of Δεp = 0.1%, 0.15%, 0.2%, could be attributed to a change in the fracture mechanism and slip
0.25% and 0.3% with a fully reversed triangular wave and total strain system with increasing strain amplitude. However, there is no reliable
rate of 2 × 10− 3 s− 1. study on bainitic steels that shows it. Therefore, in this paper the crack
behavior with increasing plastic strain amplitude will be analyzed.
2.3. Microcracks analysis
3.3. Surface damage and microcracks behavior during LCF
The microstructural features and microcracks analysis were devel­
oped using automated electron backscatter diffraction (EBSD) in a FEG – 3.3.1. Δεp = 0.1%
SEM Quanta 200 scanning electron microscope and transmission elec­ The evolution of surface damage corresponding with Δεp = 0.1%
tron microscopy (TEM) using a Philips CM 200 transmission electron shows that the first slip lines start after 50 cycles as can be seen in
microscope operated at 160 kV. Fig. 4b. As cycling proceeds, these lines intensify causing extrusions,
TexSEM Laboratories software (TSL OIM) was used to perform EBSD which will give rise to microcracks after 200 cycles.
data acquisition and post-processing. EBSD maps were acquired either In order to determine the crack initiation and propagation sites, the
using a step size of 30 nm for the microstructural characterization of the slip systems associated with the microcrack and their Schmid factors
as-received steel or 100 nm to study the microcrack behavior. Grain (SF) were calculated using a computer analysis of selected EBSD-
dilatation cleanup type with a grain tolerance angle of 15◦ was characterized areas [28]. For this data processing, the points located
employed to clean up the data points in the fatigued specimens. in the proximity of cracks were selected. For each possible slip plane, the
Thin foils were prepared from slices taken parallel to the loading axis trace of the intersection of the plane with the sample surface was
of the fatigued shallow-notched cylindrical specimens. The discs of 3 calculated and compared with the crack line. The results showed that the
mm in diameter were subsequently thinned and electro polished in a cracks lie in {1 1 0} slip planes. However, to determine the slip system
twin-jet polishing unit using a 10% perchloric acid and 90% ethanol corresponding with the crack, other analysis is necessary. Thus, the SF
mixture at − 17 ◦ C and 20 V. for each slip system {1 1 0} 〈1 1 1〉 was calculated and compared with
those obtained using TSL OIM analysis, as shown in Fig. 5a. This figure
3. Results shows an SEM image overlapped with the Schmid factor map corre­
sponding to the zone indicated by dotted lines in Fig. 4f.
3.1. As-receved microstructure From the analysis carried out, it is observed that the crack is formed
by three microcracks f1, f2 and f3, which were initiated within bainitic
[ ]
This steel presents a distribution of lower bainite blocks identified by
ferrite laths orientated on the slip system (110) 111 with SF = 0.48 and
colors in Fig. 1a. Each group of blocks with the same habit plane form a [ ]
packet. In the literature, it is considered a bainite block boundary when SF 0.49 for f1 and f3 respectively, and (101) 111 with SF = 0.48 for f2.

Table 1
Nominal chemical composition of the 16CrMnV7-7 (in wt. %). Data supplied by the manufacturer Ms = 420 ◦ C, σy = 860 MPa, TS = 1200 MPa.
C Si Mn P S Al Sn

0.15–0.20 0.20–0.30 1.68–1.80 < 0.015 0.010–0.020 0.018–0.028 < 0.020


Cr Mo Ni V Nb N Cu
0.70–1.80 0.03–0.06 0.12–0.25 0.05–0.15 0.020–0.040 0.018–0.025 < 0.20

2
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 1. As received bainitic steel: a) grain map (without cleanup) obtained by EBSD indicating blocks of bainitic ferrite laths, b) misorientation distribution of bainite
blocks, c) TEM bright field image showing retained austenite (RA), bainitic ferrite (BF), cementite and carbide particles precipitate within bainitic ferrite laths, d) The
constituent phases were analyzed by XRD using Rietveld analysis by Powder cell 2.4.

Fig. 3. Coffin - Manson curve showing a change in slope to Δεp = 0.2%.


Fig. 2. Stress amplitude vs number of cycles at several plastic strain amplitudes
of bainitic steel 16CrMnV7-7.
twist (α) and tilt (β) angles between the crack plane and the slip plane of
neighboring block (Fig. 6). According to Zhai et al [29], the smaller the
To determine the crack growth behavior and the crack plane effect
twist angle and the tilt angle between two slip planes belonging to two
on crack retardation at block boundaries, the coplanarity between the
contiguous grains are, the lower is the resistance to microcrack
slip planes of adjacent bainite blocks will be analyzed calculating the

3
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 4. Optical images showing the evolution of surface damage during LCF at Δεp = 0.1%.

Fig. 5. Crack analysis for Δεp = 0.1%. a) Schmid factor map corresponding to slip systems {1 1 0} 〈1 1 1〉, b) grain shape orientation map showing crack propagation
almost 135◦ to the loading axis.

On the other hand, analyzing the opposite crack tip, the angles be­
tween the crack plane f2 and the possible slip planes of the neighboring
block were calculated. In this case, the lowest tilt angle (β = 6◦ ) corre­
sponds to the plane (101), but for this plane the twist angle is high (α =
59◦ ) and it is known that a certain region on the grain boundary plane
needs to be broken because of the crack plane twist, which presents
significant resistance to crack growth [30]. Therefore, f2 did not grow on
this plane and it was arrested in the packet boundary producing a high
stress concentration. After a few cycles f2 merges with f1 in a non-
crystallographic coalescence of microcracks.
Fig. 5b shows the grain shape orientation map obtained by TSL OIM
analysis. On this map each point is shaded according to the shape
Fig. 6. Twist (α) and tilt (β) angles of the crack plane deflection at orientation of the grain to which it belongs. The orientation in this
packet boundary. instance is the corresponding orientation with the major axis of an el­
lipse fit to the grain makes with the loading axis.
propagation. This map will allow us to analyze the influence of the orientation of
Calculating the angles between the crack plane of f2 and the crack bainite blocks on the crack propagation. Definitely, it can be seen that
plane of f3 the results show α = 9◦ and β = 2◦ indicating, in this case, that the cracks are located on bainite blocks orientated between 120◦ and
the block boundary exerts a certain resistance to the crack propagation. 140◦ with respect to the loading direction (represented in orange in

4
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 5b). In particular, the crack plane traces are almost 135◦ to the 3.3.3. Δεp = 0.3%
loading axis. Fig. 9 shows the evolution of microcracks with the number of cycles
at Δεp ¼ 0.3%. The first microcracks initiate at 40 cycles and as cycling
3.3.2. Δεp = 0.2% proceeds, a high density of microcracks (2.6 × 10− 3 μm− 2) remain
At Δεp = 0.2%, slip lines were observed after 20 cycles (indicated by arrested at block/packet boundaries as seen in Fig. 9b-d, while others
arrows in Fig. 7b) which intensify during ongoing cycling, leading to the microcracks coalesce forming a long crack.
initiation of microcracks after 100 cycles, as shown in Fig. 7c. Using the EBSD grain map and the misorientation profile (Fig. 10),
In a previous study the microcrack nucleation sites were identified at microcracks were mainly identified at high-angle block boundaries.
Δεp = 0.2% [27]. Observing the Fig. 8a, microcracks can be divided into Fig. 11a shows the crack growth curve corresponding to the long
two main types: microcracks initiated on slip systems (highlighted by crack shown in Fig. 9d. This crack is formed by five small cracks iden­
white dotted lines) and the microcracks nucleated at block boundaries tified as F1, F2, F3, F4 and F5. In particular, we will analyze the behavior
(highlighted by white lines). The misorientation profile allows us to of each microcrack that intervenes in the formation of this long crack.
identify the different types of microcracks as shown in Fig. 8b and 8c. Fig. 11b shows in more detail the crack growth curve for each micro­
The red and blue lines indicate the point-to-point as well as the point-to- crack analyzed.
origin misorientation measurements along lines 1 and 2 respectively, According to the crack growth curve and the analysis carried out by
indicated in Fig. 8a. In this way, in Fig. 8b the point-to-origin profile EBSD, the microcrack identified as F1 initiates at a block boundary and
indicates that the microcrack initiated within the bainite block while in propagates until a packet boundary where it remains arrested. After 160
Fig. 8c the microcrack nucleated between highly misoriented bainite cycles, it propagates in the adjacent block on the slip system
( )
blocks. Therefore, with the purpose of explaining the different types of 1 0 1 [1 1 1] with SF = 0.33 until another packet boundary and
microcracks initiation, the Taylor factor (TF) will be considered for this
analysis. This factor shows the predicted yield response of a grain finally, at 300 cycles, the crack grows on the slip system (1 0 1)[1 1 1]
relative to the stress state and its interaction with the surrounding grain (SF = 0.3) and it is slow at the adjacent packet boundary.
structures. Grains with low TF value are associated with low yield Regarding the microcrack F5, it initiates on the slip system
( )
strength and can be deformed at lower stresses, while those with high 1 1 0 [1 1 1] with the highest SF = 0.4 and propagates following the
TFvalue tend to resist yielding process. Fig. 8d shows the TF map where ( )
slip systems (0 1 1)[1 1 1] and 0 1 1 [1 1 1] with the highest SF of
it is possible observing that microcracks initiated at block boundaries
are related to blocks with a high difference in TF values. For example, 0.36 and 0.37, respectively.
the microcrack indicated in Fig. 8c is between a blue block, indicating On the other hand, the microcracks F2, F3 and F4 nucleate at high-
low TF, and a red block showing high TF. On the other hand, micro­ angle block boundaries, as seen in Fig. 10, and remain arrested until
cracks initiated on slip systems correspond to system with low TF. enough energy is accumulated to surpass the adjacent boundary and
Moreover, it is interesting to note that most of the microcracks merge with the closest microcracks.
remain arrested in block or packet boundaries and only propagate those According to Fig. 11b, it is interesting to note that F5 grows linearly
that initiated on slip planes whose trace is almost 135◦ with respect to with the number of cycles. Calculating the angles α and β between crack
the loading axis as shown in Fig. 8e. These microcracks named as f1, f2 planes for F5 the results show low values (α = 2◦ ; β = 1◦ for the first pair
and f3 were identified on the slip system (110)[111] with SF = 0.48 and of plains and α = 3◦ ; β = 2◦ for the second). Therefore, as long as the
SF = 0.45 for f1 and f3 respectively, and (110)[111] with SF = 0.46 for f2. crack planes are in a close-proximity coplanar orientation, the block
Considering α and β angles between crack planes corresponding to f1 boundaries are not longer important for microcrack growth and the
and f2 (α = 3◦ , β = 6◦ ) and between the crack planes f3 and f2 (α = 30◦ , β crack grows at the same rate.
= 20◦ ) it is inferred that the crack grows from f1 to f2 and merges with f3 On the other hand, calculating the twist and tilt angles for crack
at the packet boundary as shown in Fig. 8e. planes corresponding to F1, high values are obtained, such as α = 41◦ ; β
= 16◦ . Therefore, in this case, the packet boundaries represent an effi­
cient barrier to crack propagation, decreasing its rate.

Fig. 7. Optical images showing the evolution of surface damage during LCF at Δεp = 0.2%.

5
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 8. a) Schmid factor map corresponding to the zone indicated in Fig. 7f , b) -c) misorientation profile along line 1 and 2 showing microcrack nucleation within the
bainite block and at the block boundary respectively, d) Taylor factor map, d) Grain shape orientation map indicating crack growth at bainite blocks orientated
between 120◦ and 140◦ to loading axis.

Fig. 9. Optical images showing the short fatigue crack evolution at Δεp = 0.3%.

6
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 10. EBSD grain map along with misorientation profiles indicating microcrack initiation at high- angle block boundaries. Microcracks that are circled in white
remain arrested at block/packet boundaries while those identified as F1, F2, F3 and F4 coalesce forming a long crack.

Fig. 11. a) Evolution of microcracks and their coalescence into a long crack, b) Magnification of the growth curves before the formation of the long crack. PB (packet
boundary), BB (block boundary).

It is interesting to mention that after 800 cycles, when the crack dislocations within thin laths, while Fig. 12b shows a rearrangement of
length reaches almost 300 μm (Fig. 11a), the crack growth rate increases dislocations in the {1 1 0} planes within larger laths.
significantly and packet boundaries are no longer strong barriers to At Δεp = 0.2%, the characteristic dislocation arrangement corre­
crack propagation. sponds to loop patches of dislocations. However, near microcracks, some
bainitic ferrite laths show a dislocation arrangement in the form of cells
(Fig. 13a).
3.4. Dislocation structure
On the other hand, it is interesting to note an accumulation of
dislocation at lath boundaries enclosed by dotted lines as shown in
To better understand the microcrack behavior the dislocation
Fig. 13b, causing a stress concentration.
structure developed near the surface was analyzed for the three plastic
Finally, the characteristic dislocation structure developed within the
strain ranges studied.
bainitic ferrite laths at Δεp = 0.3% corresponds to dislocation cells with
At Δεp = 0.1%, a heterogeneous distribution of dislocations within
a mean diameter of 0.3 μm as well as wall dislocation structures at the
laths can be seen in Fig. 12. Fig. 12a presents a high density of

7
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 12. TEM bright field image of dislocation structure near surface at Δεp = 0.1%. a) High density of dislocations within thin bainitic ferrite laths, b) dislocation
arrangement within large bainitic ferrite laths.

Fig. 13. TEM bright field image of dislocation structure near the surface at Δεp = 0.2%, a) dislocation arrangement forming cell, b) accumulation of dislocations
at boundaries.

end of the fatigue life (Fig. 14a). Fig. 14b shows low dislocation density to the nucleation of microcracks. Fig. 13a shows a formation of dislo­
in the bainitic ferrite laths interior and higher dislocation density near cation cell walls near the cracks which indicates a high local strain in
boundaries. this region.
Using EBSD misorientation analysis, the parameter kernel average
4. Discussion misorientation (KAM) can be applied to evaluate the local straining
level, monitoring the local misorientations [32,33,34,35]. In KAM, the
4.1. Crack initiation mean misorientation between each measurement point and its first
neighbors is calculated, excluding any high-angle boundaries. KAM
At Δεp = 0.1% and 0.2% microcracks orientated on slip systems considers only misorientations in a small local neighborhood within a
{1 1 0} 〈1 1 1〉 with the highest SF within bainite blocks were observed. grain and the values are independent of the grain size [35]. Fig. 15
Analyzing the dislocation structure, the Fig. 12b shows a rearrangement shows KAM map for Δεp = 0.2% where high-angle block boundaries
of dislocations in the {1 1 0} planes within bainitic ferrite laths. (40◦ –60◦ ) are marked with black lines. From this, it can be inferred that
Furthermore, it is known that each bainite block contains bainitic ferrite the plastic strain is mainly accumulated in the principal crack zone
laths with parallel {1 1 0} planes [31] and misorientation less than 15◦ where dislocation cells were observed. Moreover, KAM map reveals a
[26]. In particular, for the current steel, the results showed high-angle misorientation gradient near bainite block boundaries, this could be
block/packet boundaries with misorientation of 45◦ and 60◦ (Fig. 1b) indicating an accumulation of dislocations at bainitic ferrite lath
and also film-like retain austenite at lath/block boundaries. This fact boundaries as shown in Fig. 13b. In addition, at this deformation, the
suggests that dislocations are limited to moving within the bainitc ferrite results showed microcracks nucleated at bainite block boundaries with a
lath contributing to the formation of extrusion/intrusion which will lead high difference in TF values. Indeed, the block with a lower TF will yield

8
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 14. TEM bright field image of dislocation structure near the surface at Δεp = 0.3%, a) Formation of dislocation walls and cells, b) low dislocation density in the
bainitic ferrite laths interior and accumulation of dislocation near boundaries enclosed by dotted lines.

determined. Moreover, the KAM map in Fig. 16c shows greater local
misorientation due to a relatively severe accumulation of plastic strain
in these zones. By observing an enlarged area in Fig. 16d, sub-grains can
be identified within the blocks, which could be related to dislocation
cells and walls as shown in Fig. 14a. In a recent study about the evolu­
tion of dislocation density during plastic loading in a bainitic steel [37],
it was demonstrated that at high plastic deformation the dislocation
density near bainitic ferrite lath boundaries increases much more than
that in the interior, as shown in Fig. 14b. Moreover, according with Liu
et al [38], the high local plastic strain at lath boundaries could form cells
or sub-grains, which can act as damage initiation site during fatigue.

4.2. Crack propagation

With respect to cracks propagation the analysis considering the twist


and tilt angles of the crack plane at block boundaries allowed under­
standing the behavior of the crack against a microstructural barrier such
as a block or packet boundary. The smaller the twist and tilt angles, the
lower is the resistance of the block or packet boundary and the crack
grows linearly, as observed in Fig. 11b for F5. It is interesting to mention
the importance of twist angle with respect to crack propagation because
Fig. 15. KAM map from EBSD analyses applied for indicating the local of both the crack path and the growth rate are more sensitive to twist
straining level at Δεp = 0.2%. Microcracks nucleated at block boundaries are angle compared with a tilt angle [39]. In other words, a high twist angle
circled in white lines while microcracks on slip system in dotted lines. presents much larger resistance to crack propagation than a tilt angle
and the crack can be arrested or be deflected at boundaries. In Fig. 11b
prior to the “stronger” block with high TF and as it deforms a stress for F1 it is possible to appreciate a deceleration in the crack growth
concentration will build up near the boundary between them [27]. through a high twist angle boundary (α = 41◦ ).
On the other hand, at large plastic strain range (Δεp = 0.3%) a high It is noteworthy that, at low plastic strain, the longest crack is formed
density of microcracks nucleated at high-angle block boundaries was by microcracks on {1 1 0} plane surface traces orientated almost 135◦ to
observed. In this case, the cracks did not necessarily initiate between the loading axis within blocks orientated between 120◦ and 140◦ , as
blocks with high difference of Taylor factor values (see Fig. 16a). Hence, shown in Fig. 5b and 8e. In particular, the slip systems of these crack
for explaining the crack initiation, the lattice rotation due to plastic planes showed low TF and very high SF allowing easy slip for disloca­
deformation is considered using the EBSD misorientation parameter tions in that direction and therefore a preferential direction of crack
Grain Reference Orientation Deviation (GROD) since this parameter growth [40].
considers local variations in grain scale deformation, while KAM con­ On the other hand, at large plastic strain, although the main crack
siders only misorientations in a small local neighbourhood within a path remains almost normal to the principal axis of tension, this crack
grain [33,35]. GROD reveals the angular deviation of each point relative propagates by linking up with small cracks (<80 μm) orientated almost
to a given reference orientation [36]. In this case, the reference orien­ to 135◦ , 45◦ and almost normal to the loading axis (Fig. 9). The first two
tation is set to the lowest average kernel misorientation in the scanned correspond to transgranular crack growth on {1 1 0} 〈1 1 1〉 slip systems
area. in bainite packets with an orientation close to 45◦ or 135◦ with respect to
The GROD map in Fig. 16b shows that lattice rotation within blocks the loading axis as seen for F1 and F5 crack, while the latter is related to
contributes to microcrack initiation even when high TF values are the crack growth on packet boundaries. This kind of intergranular crack
is originated by a high stress concentration due to the accumulation of

9
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Fig. 16. EBSD analysis applied to microcrack initiation at Δεp = 0.3%. a) Microcracks at block boundaries with high Taylor factor, b) GROD map indicating lattice
rotation within blocks, c) KAM map showing greater local misorientation due to an accumulation of plastic deformation, d) KAM map enlarged area where sub-grains
can be identified within the blocks.

dislocations at the block boundary causing severe plastic deformation propagation ability.
and lattice rotation in the blocks as shown in Fig. 14b and Fig. 16 for F2, In the current study, the change of slope at Δεp = 0.2% may be
F3 and F4. However, the growth rate of these microcracks is not critical attributed to the change in the crack initiation mode as well as to the
compared to those that grow by slip systems as seen in Fig. 11b. change crack propagation mode. The results showed that at low plastic
Rementeria and Mueller et al [22,23] have reported that the crack path strain range (Δεp = 0.1%), the microcraks initiate in bainitic ferrite laths
and its velocity are highly influenced by the size of bainite blocks and by orientated on slip systems {1 1 0} 〈1 1 1〉 with the highest SF and a
microstructural barriers at which the crack can be stopped, decelerated transgranular propagation takes place on {1 1 0} slip planes with surface
or deflected. Furthermore, in the current work, the importance of the traces oriented almost 135◦ with respect to loading axis. Conversely, at
bainite packet orientation with respect to loading direction is remark­ high plastic strain range (Δεp = 0.3%), the microcracks preferably
able. If the crack path to the next bainite packet is through highly nucleate at high-angle block boundaries and the crack propagation oc­
misoriented packet, the crack may deviate and grow intergranularly curs as a combination of two modes, transgranular and intergranular as
until it reaches to a less energy state lattice rotation to activate the seen in the previous section.
available slip systems [30]. On the other hand, it is noteworthy to mention that at Δεp = 0.2% the
microcracks could initiate either at bainitic ferrite laths on the primary
slip system {1 1 0} 〈1 1 1〉 or at high-angle block boundaries. In this case,
4.3. Plastic strain/ life behavior both crack initiation mechanisms are present. Hence, this plastic strain
range is considered a transition range with respect to cracking
The Coffin-Manson curves exhibit a bilinear relationship in the mechanisms.
log–log coordinates, and the transition occurs at the plastic strain
amplitude of around Δεp/2 = 0.1%, as seen in Fig. 3. This phenomenon 5. Conclusions
was previously reported by Mediratta et al [41] in dual-phase steels as a
consequence of a change in fracture mode from dimple fracture at low The mechanisms of initiation and propagation of microcracks during
strain amplitudes to cleavage fracture at high strain amplitudes. More­ low-cycle fatigue in a bainitic steel 16CrMnV7-7 were characterized
over, Zhang et al [42] attributed the bilinearity in the Coffin-Manson through LOM, EBSD and TEM. The study of both microcracks and fatigue
relationship for Nickel-based alloy to the dependence of deformation behavior was carried out at different plastic strain ranges: Δεp = 0.1%,
mode (from twinning to slip bands with increasing the strain amplitude) 0.2% and 0.3% concluding that:
as well as the change in fatigue crack initiation mode from transgranular - Fatigue behavior shows initial hardening response with subsequent
mode at low strain amplitude to intergranular when the strain amplitude cyclic softening for all plastic strain ranges studied.
is high. On the other hand, in bainitic steels with low-temperature - Coffin-Manson plot shows a bilinear behavior at Δεp = 0.2%, which
bainite, Zhang et al [15] associated this phenomenon with the micro­ is related to a change in the crack initiation mode as well as to the
structure stability, the compatible deformation ability and the crack

10
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

change in the crack propagation mode. [11] Garcia-Mateo C, Caballero FG, Bhadeshia HKDH. Acceleration of Low-temperature
Bainite. ISIJ Int 2003;43(11):1821–5. https://doi.org/10.2355/
- The Taylor factor analysis allowed understanding microcrack
isijinternational.43.1821.
nucleation at low strains, while at high strains lattice rotation within [12] Zhang FC, Wang TS, Zhang P, Zheng CL, Lv B, Zhang M, et al. A novel method for
bainite blocks due to plastic strain accumulated contributes to micro­ the development of a low-temperature bainitic microstructure in the surface layer
crack initiation even when high Taylor factor values are determined. of low-carbon steel. Scr Mater 2008;59(3):294–6. https://doi.org/10.1016/j.
scriptamat.2008.03.024.
- At low plastic strain (Δεp < 0.2%), microcraks initiate in bainitic [13] Wang TS, Li XY, Zhang FC, Zheng YZ. Microstructures and mechanical properties of
ferrite laths orientated on primary slip systems {1 1 0} 〈1 1 1〉 with low 60Si2CrVA steel by isothermal transformation at low temperature. Mater Sci Eng,
Taylor factor. The crack propagation takes place on {1 1 0} slip planes A 2006;438-440:1124–7. https://doi.org/10.1016/j.msea.2006.02.077.
[14] Rementeria R, García I, Aranda MM, Caballero FG. Reciprocating-sliding wear
with surface traces oriented at 135◦ with respect to loading axis. behavior of nanostructured and ultra-fine high-silicon bainitic steels. Wear 2015;
- At high plastic strain (Δεp = 0.3%), the microcracks preferably 338-339:202–9. https://doi.org/10.1016/j.wear.2015.06.011.
nucleate at high-angle block boundaries as a consequence of a high local [15] Zhang FC, Long XY, Kang J, Cao D, Lv B. Cyclic deformation behaviors of a high
strength carbide-free bainitic steel. Mater Des 2016;94:1–8. https://doi.org/
plastic strain and a lattice rotation. The crack propagation occurs by 10.1016/j.matdes.2016.01.024.
coalescence of small cracks that grew both transgranular and [16] Kang J, Zhang FC, Long XY, Lv B. Low cycle fatigue behavior in a medium-carbon
intergranular. carbide-free bainitic steel. Mater Sci Eng, A 2016;666:88–93. https://doi.org/
10.1016/j.msea.2016.03.077.
- The analysis of crack propagation considering the twist and tilt [17] Zhou Q, Qian L, Meng J, Zhao L, Zhang F. Low-cycle fatigue behavior and
angles of the crack plane at block boundaries allowed a better under­ microstructural evolution in a low-carbon carbide-free bainitic steel. Mater Des
standing of crack behavior against microstructural barriers. Although 2015;85:487–96. https://doi.org/10.1016/j.matdes.2015.06.172.
[18] Branco R, Costa JD, Antunes FV. Low-cycle fatigue behaviour of 34CrNiMo6 high
both parameters control the crack path and the growth rate, the twist
strength steel. Theor Appl Fract Mech 2012;58(1):28–34. https://doi.org/10.1016/
angle proved to be the most significant. j.tafmec.2012.02.004.
- At Δεp = 0.2% microcracks can nucleate either at bainitic ferrite [19] Long X, Zhang F, Yang Z, Lv Bo. Study on microstructures and properties of
laths on the primary slip system {1 1 0} 〈1 1 1〉 or at high-angle block carbide-free and carbide-bearing bainitic steels. Mater Sci Eng, A 2018;715:10–6.
https://doi.org/10.1016/j.msea.2017.12.108.
boundaries between bainite blocks with high difference of Taylor factor [20] Sankaran S, Sarma VS, Gouthama SS, Padmanabhan KA. Low cycle fatigue
values. The microcracks on the surface plane traces oriented almost 135◦ behaviour of a multiphase medium carbon microalloyed steel processed through
to loading axis, coalesce forming a dominant fatigue crack which results rolling. Scr Mater 2003;49:503–8. https://doi.org/10.1016/S1359-6462(03)
00363-4.
in the specimen failure. [21] Sankaran S, Sarma VS, Padmanabhan KA. Low cycle fatigue beha v ior of a
multiphase microalloyed medium carbon steel : comparison between ferrite Á
pearlite and quenched and tempered microstructures 2003;345:328–35.
Declaration of Competing Interest [22] Rementeria R, Morales-Rivas L, Kuntz M, Garcia-Mateo C, Kerscher E, Sourmail T,
Caballero FG. On the role of microstructure in governing the fatigue behaviour of
nanostructured bainitic steels. Mater Sci Eng, A 2015;630:71–7. https://doi.org/
The authors declare that they have no known competing financial 10.1016/j.msea.2015.02.016.
interests or personal relationships that could have appeared to influence [23] Mueller I, Rementeria R, Caballero FG, Kuntz M, Sourmail T, Kerscher E.
A constitutive relationship between fatigue limit and microstructure in
the work reported in this paper. nanostructured bainitic steels. Materials (Basel) 2016;9:1–19. https://doi.org/
10.3390/ma9100831.
[24] Nohava J, Haušild P, Karlı ́k M, Bompard P. Electron backscattering diffraction
Acknowledgments
analysis of secondary cleavage cracks in a reactor pressure vessel steel. Mater
Charact 2002;49(3):211–7. https://doi.org/10.1016/S1044-5803(02)00360-1.
This work was supported by Agencia Santafesina de Ciencia, Tec­ [25] Hu H, Xu G, Zhou M, Yuan Q. Effect of mo content on microstructure and property
nología e Innovación by under grant Res 133/18 Proy IO 2017-00092, of low-carbon bainitic steels. Metals (Basel) 2016;6. https://doi.org/10.3390/
met6080173.
Universidad Nacional de Rosario (UNR) under grant number RCS 201 [26] Guo B, Fan L, Wang Q, Fu Z, Wang Q, Zhang F. The role of the bainitic packet in
19 and Consejo Nacional de Investigaciones Científicas y Técnicas control of impact toughness in a simulated CGHAZ of X90 pipeline steel. Metals
(CONICET). (Basel) 2016;6. https://doi.org/10.3390/met6110256.
[27] Marinelli MC, Alvarez-Armas I, Krupp U. Cyclic deformation mechanisms and
microcracks behavior in high-strength bainitic steel. Mater Sci Eng, A 2017;684:
References 254–60.
[28] Marinelli MC, Moscato MG, Signorelli JW, El Bartali A, Alvarez-Armas I. K-S
relationship identification technique by EBSD. vol. 465. 2011. https://doi.org/
[1] Babu SS, Vogel S, Garcia-Mateo C, Clausen B, Morales-Rivas L, Caballero FG.
10.4028/www.scientific.net/KEM.465.415.
Microstructure evolution during tensile deformation of a nanostructured bainitic
[29] Zhai T, Wilkinson AJ, Martin JW. Crystallographic mechanism for fatigue crack
steel. Scr Mater 2013;69(11-12):777–80. https://doi.org/10.1016/j.
propagation through grain boundaries. Acta Mater 2000;48:4917–27. https://doi.
scriptamat.2013.08.026.
org/10.1016/S1359-6454(00)00214-7.
[2] Long XY, Zhang FC, Kang J, Lv B, Shi XB. Low-temperature bainite in low-carbon
[30] Krupp U. Fatigue Crack Propagation in Metals and Alloys. n.d.
steel. Mater Sci Eng, A 2014;594:344–51. https://doi.org/10.1016/j.
[31] Beladi H, Adachi Y, Timokhina I, Hodgson PD. Crystallographic analysis of
msea.2013.11.089.
nanobainitic steels. Scr Mater 2009;60:455–8. https://doi.org/10.1016/j.
[3] Zhang FC, Lv B, Zheng CL, Zou Q, Zhang M, Li M, Wang TS. Microstructure of the
scriptamat.2008.11.030.
worn surfaces of a bainitic steel railway crossing. Wear 2010;268(11-12):1243–9.
[32] Wang X, Chen J-G, Su G-F, Li H-Y, Wang C. Plastic damage evolution in structural
https://doi.org/10.1016/j.wear.2010.01.016.
steel and its non-destructive evaluation. J Mater Res Technol 2020;9(2):1189–99.
[4] Zhao J, Zhao T, Hou CS, Zhang FC, Wang TS. Improving impact toughness of high-
[33] Rui S-S, Shang Y-B, Su Y, Qiu W, Niu L-S, Shi H-J, Matsumoto S, Chuman Y. EBSD
C–Cr bearing steel by Si–Mo alloying and low-temperature austempering. Mater
analysis of cyclic load effect on final misorientation distribution of post-mortem
Des 2015;86:215–20. https://doi.org/10.1016/j.matdes.2015.07.055.
low alloy steel: A new method for fatigue crack tip driving force prediction. Int J
[5] Li Y, Zhang F, Chen C, Lv Bo, Yang Z, Zheng C. Effects of deformation on the
Fatigue 2018;113:264–76. https://doi.org/10.1016/j.ijfatigue.2018.04.016.
microstructures and mechanical properties of carbide–free bainitic steel for railway
[34] Yoda R, Yokomaku T, Tsuji N. Plastic deformation and creep damage evaluations of
crossing and its hydrogen embrittlement characteristics. Mater Sci Eng, A 2016;
type 316 austenitic stainless steels by EBSD. Mater Charact 2010;61(10):913–22.
651:945–50. https://doi.org/10.1016/j.msea.2015.09.117.
https://doi.org/10.1016/j.matchar.2010.05.006.
[6] Caballero FG, Santofimia MJ, García-Mateo C, Chao J, de Andrés CG. Theoretical
[35] Schayes C, Bouquerel J, Vogt J-B, Palleschi F, Zaefferer S. A comparison of EBSD
design and advanced microstructure in super high strength steels. Mater Des 2009;
based strain indicators for the study of Fe-3Si steel subjected to cyclic loading.
30(6):2077–83. https://doi.org/10.1016/j.matdes.2008.08.042.
Mater Charact 2016;115:61–70. https://doi.org/10.1016/j.matchar.2016.03.020.
[7] Pointner P. High strength rail steels—The importance of material properties in
[36] Wright SI, Nowell MM, Field DP. A Review of Strain Analysis Using Electron
contact mechanics problems. Wear 2008;265(9-10):1373–9. https://doi.org/
Backscatter Diffraction. Microsc Microanal 2011;17(3):316–29. https://doi.org/
10.1016/j.wear.2008.03.015.
10.1017/S1431927611000055.
[8] Long XY, Branco R, Zhang FC, Berto F, Martins RF. Influence of Mn addition on
[37] He SH, He BB, Zhu KY, Huang MX. Evolution of dislocation density in bainitic steel:
cyclic deformation behaviour of bainitic rail steels. Int J Fatigue 2020;132:105362.
Modeling and experiments. Acta Mater 2018;149:46–56. https://doi.org/10.1016/
https://doi.org/10.1016/j.ijfatigue.2019.105362.
j.actamat.2018.02.023.
[9] Bhadeshia HKDH. Steels for bearings. Prog Mater Sci 2012;57(2):268–435. https://
[38] Liu S-D, Zhu M-L, Zhou H-B, Wan Di, Xuan F-Z. Strain visualization of growing
doi.org/10.1016/j.pmatsci.2011.06.002.
short fatigue cracks in the heat-affected zone of a Ni–Cr–Mo–V steel welded joint:
[10] Caballero FG, Bhadeshia HKDH. Very strong bainite. Curr Opin Solid State Mater
Sci 2004;8(3-4):251–7. https://doi.org/10.1016/j.cossms.2004.09.005.

11
M.C. Marinelli et al. International Journal of Fatigue 143 (2021) 106014

Intergranular cracking and crack closure. Int J Press Vessels Pip 2019;178:103992. [41] Mediratta SR, Ramaswamy V, Rao PR. Influence of ferrite-martensite
https://doi.org/10.1016/j.ijpvp.2019.103992. microstructural morphology on the low cycle fatigue of a dual-phase steel. Int J
[39] Fang W, Xie H, Yin F, Li J, Khan DF, Fang Q. Molecular dynamics simulation of Fatigue 1985;7(2):107–15. https://doi.org/10.1016/0142-1123(85)90041-6.
grain boundary geometry on crack propagation of bi-crystal aluminum. Mater Sci [42] Zhang X-C, Li H-C, Zeng Xu, Tu S-T, Zhang C-C, Wang Q-Q. Fatigue behavior and
Eng, A 2016;666:314–9. https://doi.org/10.1016/j.msea.2016.04.077. bilinear Coffin-Manson plots of Ni-based GH4169 alloy with different volume
[40] Azar AS, Svensson L-E, Nyhus B. Effect of crystal orientation and texture on fatigue fractions of δ phase. Mater Sci Eng, A 2017;682:12–22. https://doi.org/10.1016/j.
crack evolution in high strength steel welds. Int J Fatigue 2015;77:95–104. https:// msea.2016.11.040.
doi.org/10.1016/j.ijfatigue.2015.03.008.

12

You might also like