- Email: [email protected]
Fracture behavior of nanostructured heavily cold drawn pearlitic steel wires before and after annealing
Materials Science & Engineering A 707 (2017) 164–171
Contents lists available at ScienceDirect
Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea
Fracture behavior of nanostructured heavily cold drawn pearlitic steel wires before and after annealing ⁎
B.N. Jayaa, , S. Gotoa,b, G. Richterc, C. Kirchlechnera,d, G. Dehma,
MARK
⁎
a
Max Planck Institut fuer Eisenforschung GmbH, Duesseldorf, Germany Akita University, Tegata Gakuencho, Akita 010-8502, Japan c Max Planck Institute for Intelligent Systems, Stuttgart, Germany d Department of Material Physics, University of Leoben, Leoben, Austria b
A R T I C L E I N F O
A B S T R A C T
Keywords: Pearlitic steel Nanocrystalline Wire drawing Fracture Micromechanics
In situ micro-cantilever fracture testing is used to demonstrate changes in fracture behavior of nanostructured, heavily cold drawn pearlitic steel wires as a function of drawing strain and annealing conditions. It is shown that these steels exhibit a sharp transition in fracture behavior between a drawing strain of 320% and 520% with a drop in fracture toughness from 7.5 to 4 MPam1/2. This is confirmed from the nature of fracture which is stable with some degree of plasticity at drawing strains below 320% and changes to catastrophic cleavage fracture at drawing strains of 420% and above. This transition and associated brittleness is attributed to structural (cementite decomposition and strain induced increase in tetragonality) and microstructural (increasing nanocrystallinity and dislocation density) evolution that these steels undergo at higher drawing strains. On heat treating the 420% strained sample, brittle cleavage fracture continues for low temperature (200 °C) annealing with no visible changes in microstructure, while crack growth is suppressed and large-scale plasticity is recovered for high temperature (500 °C) annealing with accompanying grain coarsening, and re-precipitation of spherodized cementite at grain boundaries.
1. Introduction Cold drawn pearlitic steel, a nanolaminate composite structure consisting of alternate layers of ferrite and cementite, finds applications spanning from suspension bridge cables, rails and tire reinforcement cords to musical instruments and springs, due to its exceptional mechanical properties and the ability to be drawn into extremely thin wires [1]. The ductile α-Fe phase carries most of the strain while the harder Fe3C contributes to strengthening and work-hardening by acting as a barrier for dislocation movement. The mechanical performance of such a structure is dominated by the high density of ferrite-cementite interfaces. With increasing drawing strain, the individual colonies of pearlite become increasingly oriented along the drawing direction due to co-deformation with the ferrite matrix, developing a fiber texture along < 110 > . On further straining, the hard and brittle cementite phase breaks down and undergoes nano-scale refinement ([2] and references therein). The strength of cementite is known to be 10 times higher than that of the ferrite matrix in undeformed, as-patented pearlite [3]. Because of the widely different properties of ferrite and cementite, a very inhomogeneous stress distribution develops inside the pearlitic steels during wire drawing. ⁎
Severe plastic deformation (SPD) processes like high pressure torsion and cold drawing can produce steels with ultra-high tensile strengths, close to that of the constituent cementite [2,4]. Recently Li et al. [5], showed that tensile strengths close to 7 GPa could be achieved in heavily cold drawn pearlitic steel wire at drawing strains above 600%. Not surprisingly, these high strengths were accompanied by significant loss of ductility and linear elastic fracture when pulled in tension. Li et al. carried out correlative atom probe tomography (APT) and transmission electron microscopy (TEM), to establish that these steels undergo fundamental transformation in their structure following decomposition of the cementite phase and supersaturation of the excess carbon in the ferrite matrix at drawing strains above 300% [5]. Djaziri et al. were able to show using synchrotron based X-day diffraction that this supersaturation of carbon in the matrix results in an increase in its tetragonality, leading to formation of strain induced martensite [6]. In real applications like in suspension bridge cables, these cold drawn wires are subsequently subjected to various heat treatments in the process of galvanizing/coating them with anti-corrosive agents. This will expose the highly non-equilibrium cold drawn structure to temperatures high enough to bring about coarsening of their nanostructure and a loss of tensile strength. Thus, thermal stability evaluation
Correspondence to: Department of Metallurgical Engineering and Materials Science, IIT-Bombay, Mumbai, India. E-mail addresses: [email protected] (B.N. Jaya), [email protected] (G. Dehm).
http://dx.doi.org/10.1016/j.msea.2017.09.010 Received 27 March 2017; Received in revised form 3 July 2017; Accepted 2 September 2017 Available online 14 September 2017 0921-5093/ © 2017 Elsevier B.V. All rights reserved.
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
compound forms, to nanocrystalline Fe without any C in them. These comparisons are motivated by first principle calculations that indicate that C improves the interlamellar cohesion of ferrite, and its absence could enhance the brittle behavior [12].
is of paramount importance. On heat treating wires of drawing strain of 650%, Li et al. observed that the tensile strength is retained for low temperature annealing (< 250 °C) while there is a dramatic drop on annealing these wires at high temperatures (> 250 °C) [7]. In view of the evolving microstructure, the deformation and micro-fracture mechanism of these wires is expected to undergo significant transitions. These have immediate technological implications, when they are required to be exploited for their high strengths. Fracture behavior needs to be clearly understood as a function of drawing strain and annealing treatments to be able to prevent catastrophic fractures and provide sufficient damage tolerance at high drawing stains, while not compromising on strength. The micro-scale fracture behavior of cold drawn pearlitic steel wire in the longitudinal direction at high drawing strains were not explored earlier due to dimensional constraints of the limited wire diameter. The advent of in situ electron microscopy based micro-mechanical fracture testing [8] has changed this. Recently, Hohenwarter et al. [9], compared the damage tolerance of heavily cold drawn wires at two extreme drawing strains of 310% and 650%, both in the longitudinal and transverse directions. They identified and quantified the anisotropy in fracture behavior in the two directions, and observed catastrophic fracture in the longitudinal direction, and micro-cracking induced crack tip stress relaxation in the transverse direction. Their work provides valuable insights on the counter-intuitive contribution of the weak inter-columnar interface along the drawing direction enhancing the damage tolerance of these steel wires in the transverse direction. Our study expands on the work in the longitudinal direction, looking at intermediate drawing strains, and at different annealing conditions, where significant changes in microstructure have been reported [5,6,10,11]. Specifically, the objectives of the present work are to identify and quantify fracture behavior of cold drawn pearlitic steel wires in the drawing direction, in terms of their fracture toughness and fracture surface morphology a) at drawing strains of 320%, 420% and 520%, where a clear transition in microstructure has been observed from nanostructured lamellar pearlite composite, to nano-scaled sub-grained ferrite with supersaturated C [2,5,6], b) before and after low temperature (200 °C) and high temperature (500 °C) annealing for the transition drawing strain of 420% where a significant drop in tensile strength accompanied by grain coarsening has been recorded [10]. In addition to answering fundamental questions of structure-property correlations, quantification of fracture resistance in these wires will have a direct impact on critical applications where SPD steels are prone to fatigue, fracture and wear.
2.2. Micro-scale fracture testing and characterization The wires being in the range of tens to hundreds of microns in diameter, fracture toughness determination in the longitudinal drawing direction requires micro-mechanical testing. The validity of micro-scale fracture test geometries in determining the fracture toughness of brittle and semi-brittle systems has already been established [13–15]. Focused ion beam machined and notched (Zeiss AURIGA®) micro-cantilevers were tested in bending in situ in the scanning electron microscopy (SEM) (ASMEC UNAT-2 in JEOL JSM 2000) till fracture. Beam dimensions for all drawing strains were chosen to be in the same order of magnitude ~8 × 2 × 2 µm3 with a crack length to width (a/W) ratio ~0.3 (Table 1). All beams were Ga+ FIB machined at 4 nA and polished at 240 pA and eventually notched from the top at 10pA current at 30 kV. All micro-beams for a given drawing strain were machined on wires drawn from the same batch, but owing to the limited wire diameter, each wire piece could support 3–4 microbeams only (Fig. 1a-b). The loading was initially monotonic for all cases (Fig. 1c) but lead to different responses as a function of drawing strain or annealing conditions. For example, high drawing strains (≥ 420%) showed a purely elastic-brittle behavior and lower drawing strains (320%) showed significant deviation from linear elasticity (Fig. 1d). In such cases, cyclic loading with a series of load-partial unload-load sequence was carried out (see supplementary information). The latter was to enable J-integral calculations by recording the changing compliance during the unloading sequences of the load (P)-deflection (d) curve. Linear elastic fracture mechanics (LEFM) was used to determine KC for higher drawing strains (≥ 420%) (Eqs. (1) and (2)) [14], while it was used to determine only the lower limit of fracture toughness for lower drawing strain (320%). Fracture energy G and plastic zone size rp was evaluated according to Eqs. (3) and (4) respectively. Results were averaged over 5 or more micro-cantilevers for each case except for the annealed condition. Post-mortem high resolution imaging of the fracture surface and crack morphology was carried out in the Zeiss AURIGA® SEM.
Kc =
Pcrit L a f⎛ ⎞ BW 3/2 ⎝ W ⎠
a a a 2 a 3 f ⎛ ⎞ = 1.46 + 24.36 ⎛ ⎞ − 47.21 ⎛ ⎞ + 75.18 ⎛ ⎞ ⎝W ⎠ ⎝W ⎠ ⎝W ⎠ ⎝W ⎠
2. Experimental procedure 2.1. Material
G=J= Hypereutectoid steel wires of nominal composition Fe-0.98 wt%C (Fe–0.98C–0.31Mn–0.20Si–0.20Cr–0.01Cu–0.006P–0.007 S in wt%) supplied by Suzuki Metal Industry Co., Ltd. of drawing strains of 320%, 420% and 520% were used for the present study. Corresponding average lamellar thickness/grain sizes were 18, 15 and 10 nm respectively [2]. A second series of tests were carried out on annealed wires (200 °C and 500 °C for 30 min) of 420% drawing strain to evaluate the role of recovery and grain coarsening on the fracture behavior of these steels in comparison to the as-drawn state. Corresponding mean subgrain sizes were 19 and 130 nm respectively [10]. Table 1 shows the drawing strains, diameters and heat treatment conditions of the corresponding wires used. In addition, magnetron sputtered Fe thin films of 2.2 ± 0.06 µm thickness on SiO2 substrates were also used to evaluate the fracture behavior of pure unalloyed nanocrystalline α-Fe. The average in-plane grain size of these thin films was 200 nm ± 33 nm and these were columnar in nature with the grains oriented primarily along the < 110 > . These were used to compare the fracture toughness of pearlitic steels that contain C at the grain boundaries in either elemental or
KC 2 (1−ν 2) E
(1)
(2)
(3)
2
rp =
1 ⎛ KC ⎞ ⎜ ⎟ 2π ⎝ σy ⎠
(4)
where Kc is the stress intensity factor, a is the crack length, σ is the applied stress, Pcrit is the fracture load, and L, B, and W correspond to the cantilever length, breadth and width respectively, G is the fracture energy, J is the J-integral, E is the elastic modulus, ν is the Poisson's ratio, σy is the tensile yield strength. J-integral measurements were carried out by repeated load-unload cycles in the case of the 320% asdrawn samples, which showed some plasticity during the cantilever fracture experiments. The procedure is detailed below. The elasto-plastic fracture toughness in terms of JC values for the 320% drawing strain sample, was determined by the procedure as established by Wurster et al. [16] on two micro-cantilever specimens. Quantification of fracture toughness in the classical sense in more ductile systems at small length scales remains a challenge [8]. This is a problem for pearlitic wires of lower drawing strains. ASTM like standards of plane strain and small scale yielding are impossible to be 165
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Table 1 Specimen details listed in terms of cold drawing strains and heat treatment conditions, specimen dimensions and number of specimen tested for every condition. L: length, B: breadth, W: width and a: crack length of the cantilever beam, as indicated in Fig. 1. Sample Unit
Wire diameter µm
Cantilever dimensions: L×B×W µm
a/W
Statistics
As drawn: ε = 320% As drawn: ε = 420% As drawn: ε = 520% Annealed: 200 °C, 30 min ε = 420% Annealed: 500 °C, 30 min ε = 420% Fe thin film
115 66 42 66 66 2.2 (thickness)
8.2 × 1.5 × 2.2 7.8 × 2.2 × 1.8 8.5 × 1.7 × 1.9 10.5 × 1.9 × 2 11.5 × 2.2 × 2.3 15 × 2 × 2.5
0.25 0.3 0.3 0.28 0.25 0.28
5 8 6 4 4 5
of the cold drawn specimen of drawing strains 320% and 520% strain respectively while Table 2 summarises the KC values obtained from LEFM calculations for all of them. The as-drawn specimen corresponding to drawing strains of 420% and 520% show linear elastic deformation and catastrophic fracture with a fracture toughness of 5.2 MPam1/2 and 4 MPam1/2 respectively (Fig. 3d-f, Fig. 4a). The specimen corresponding to 320% drawing strain and lower show deviation from linear elastic behavior and a gradual drop in load during fracture, following plastic deformation and stable crack growth (Fig. 3a-c) with a corresponding LEFM fracture toughness of 7.5 MPam1/2. Fracture toughness calculations for these samples were repeated using J-integral measurements. The JC and JR were found to be 488 N/m and 3203 N/ m, respectively employing Eq. (6). Converting JC and JR to K values using Eq. (3) gives an initiation Kc of 10.6 MPam1/2 and a propagation KR of 27 MPam1/2 for a drawing strain of 320% (Fig. 2) at the end of the test. The procedural details of such a calculation are given in Wurster et al. [16]. This increase from 7.5 MPam1/2 at initiation (from LEFM) to 27 MPam1/2 after propagation (using J-integral) accompanied with stable crack growth hints at possible R-curve effect for this drawing strain, which remains to be systematically quantified. It is important to J note that the requirement for B > 25 σ is not met in this case and hence
complied at micro- and nano- dimensions. CTOD measurements could not be carried out due to lack of sufficient imaging resolution in our SEM. Even if J-integral measurements are carried out, only a lower limit of fracture energy can be ascertained due to specimen size limitations. Fig. 2a-b shows the P-d curve of a 320% strain sample subjected to cyclic loading with ~10% increments in displacement after each cycle. The specimen attained its maximum load at the fourth loading cycle and recorded gradual increase in compliance after that, corresponding to stable crack growth (Fig. 2c-d). The stiffness change was calculated from the unloading slopes after every cycle to correlate it to the change in crack length using Eq. (5) [16]. The change in crack length (W-ai) was in turn used to estimate both the initiation JC and propagation JR according to Eq. (6).
W − ai =
Ji =
kL3 BE
2Apl (i) (Ki )2 (1−ν 2) + E B (W − a 0)
(5)
(6)
where i represents increments in unloading cycles, Apl represents the plastic area under the P-d curve.
ys
these cannot be considered standard J-integral measurements. Post-mortem analysis of the fracture surface shows interface dominated cleavage fracture in all cases (Fig. 3c & f). The 320% drawn specimen shows a mixed transgranular and intergranular fracture with an average fracture feature spacing of 200 nm, while the 420% and
3. Results Fig. 3a & d shows the P-d curves (from monotonic loading) and Fig. 3b-c & e-f the corresponding crack trajectory and fracture surfaces
Fig. 1. (a)-(b) FIB micro-machined micro-cantilevers on cold drawn wire of 420% drawing strain. (c) Insitu micro-cantilever bending of the cantilever with the crack plane aligned along the longitudinal drawing direction, using a conical indenter, inside the SEM. (d) Load displacement curves showing responses of cold drawn wires of 320% and 520% drawing strains. A linear elastic response is seen for the 520% drawing strain while considerable deviation from elastic behavior is seen for the 320% drawing strain before fracture.
166
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Fig. 2. (a) A 320% drawn wire specimen with L = 7.09 µm, W = 2.2 µm, B = 1.55 µm and a0 = 0.62 µm, loaded cyclically to determine the elastoplastic fracture toughness JC. (b) Corresponding load-displacement response, (c) stiffness plot as a function of the number of unloading cycle showing a progressive increase in compliance, and (d) incremental increase in crack growth as a function of unloading cycle, calculated using Eq. (5) from the unloading stiffness. The final crack length measured from SEM is 0.8 µm which comes close to the crack length of 0.75 µm calculated from compliance changes.
While the 200 °C annealed specimen shows linear elastic brittle fracture like the as-drawn specimen, the 500 °C annealed specimen shows significant plastic deformation and no visible crack propagation even when subjected to large strains. While cleavage fracture features are part of the 200 °C annealed sample (Fig. 4-II), dimpled ductile fracture
520% drawn specimen show increasing tendency for fracture along the inter-columnar boundary parallel to the drawing direction with an average fracture feature spacing of 107 nm and 100 nm respectively. Fig. 4 compares the P-d curves and fracture features of the 420% asdrawn specimen with those annealed at 200 °C and 500 °C (30 min).
Fig. 3. Fracture response of cold drawn wires of drawing strain (a-c) 320% and (d-f) 520% showing the load (P)-displacement (d) characteristics, crack trajectory and fracture surface respectively. A distinct difference in fracture behavior is seen with the lower drawing strains showing elastic-plastic fracture with a stable and tortuous crack growth and mixed fracture, and the higher drawing strains showing brittle fracture with a straight path along the inter-columnar boundary. (b) and (e) are side surface views showing the crack trajectory. (c) and (f) are inclined views along the beams showing the initial notch (ao) and fracture morphology in the longitudinal direction.
167
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Table 2 Summary of fracture results of all tested specimen with calculated plastic zone size and fracture energy and associated microstructural parameters. Sample
daverage (Li et al. [5,7])
KC
σy (Li et al. [5,7])
rp
G
Unit As drawn: ε = 320% As drawn: ε = 420% As drawn: ε = 520% Annealed 200 °C: ε = 420% Annealed 500 °C: ε = 420% Fe thin film
nm 18
MPam1/2 7.5 ± 0.85
GPa 3.8
nm 620
J/m2 243
15
5.2 ± 0.94
5.2
159
117
10
4.0 ± 0.27
6.2
66
69
19
5.1 ± 0.65
5.2
153
113
~130
N. A.
1.9
N.A.
N.A.
200
2.0 ± 0.2
1.7*
220
51
Fig. 5. Fracture toughness of all the tested wires as a function of their corresponding yield strength showing a clear trend of brittle fracture and low fracture toughness for wires of high yield strengths corresponding to SPD states. Data from other works [4] have been added for completion. The number in brackets (n) is the average grain size.
stable crack growth occurred for the drawing strain of 320%, pointing to a larger process zone and activation of additional toughening mechanisms at lower drawing strains. Consequently, the plastic zone size of 620 nm is a large fraction of the specimen width of ~ 2 µm for the 320% strained sample (Table 2), implying that the Kc is only a lower estimate for the same. For the heavily cold drawn samples, the plastic zone size is nearly 1/10th the beam width (Table 2) and can be assumed to satisfy plane strain. The average fracture toughness of the 200 °C annealed specimen was 5.1 MPam1/2. This is very close to the as-drawn condition, while that of the 500 °C annealed specimen could not be evaluated due to extensive plastic deformation. It is clear from Fig. 5 that the fracture toughness follows an inverse relationship with yield strength and grain size in general for a given alloy system. The fracture toughness of the nanocrystalline sputtered pure Fe thin film (with average grain size of 200 nm) was much lower at 2 MPam1/2, making it the most brittle among all the specimens tested in the present work (Table 1). The crack path was intergranular, while the P-d response was linear elastic (Fig. 6). This shows that even for a comparatively larger grain size, and with the crack being oriented in a direction perpendicular to the columnar grains, pure sputtered Fe films exhibit a lower fracture resistance compared to nanostructured heavily cold drawn ferrite, with C segregated at its sub-grain boundaries. Although possessing very different defect structures, these results hint at C playing an important role in improving grain boundary cohesion in steels as compared to pure Fe.
Fig. 4. (a) Fracture response of 420% drawn wire under cold drawn and annealed states showing linear-elastic response for the cold-drawn and low-temperature annealed states and showing elasto-plastic deformation for the high-temperature annealed state. The fracture surfaces reveal brittle inter-columnar boundary fracture for the former two states, while exhibiting dimpled, ductile fracture for the latter. (b) FIB polished and exposed surfaces of as-drawn and high temperature annealed samples under channeling contrast, showing evidence of grain coarsening.
4. Discussion
features were prominently visible for the 500 °C annealed sample (Fig. 4-III) (when fractured artificially after the loading sequence). This clearly shows the difference in fracture behavior due to microstructural changes occurring during the 500 °C annealing [10]. Fig. 5 shows the initiation fracture toughness calculated for all the specimens tested in the present work, superimposed with the results from equivalent strains of SPD specimens (as reported earlier by Hohenwarter et al. [4]), plotted as a function of their yield strength. It was seen that wires with 200% drawing strain or lower showed large scale ductility due to which their fracture behavior could not be quantified using existing micro-mechanical test methods and inapplicability of LEFM. Micro-scale fracture tests carried out for drawing strains of 320%, 420% and 520% revealed a significant drop in the initiation KC from 7.5 to 5.2 to 4 MPam1/2 respectively (Table 2 and Fig. 5). The corresponding yield strength data from [5] is also repeated in Table 2. Fracture was unstable and catastrophic for drawing strains ≥ 420% but
Our micro-cantilever fracture toughness evaluation of cold drawn pearlitic steel wire tested in the longitudinal drawing direction shows a clear drop in fracture toughness with increasing drawing strain, transitioning from stable, elastic-plastic, semi-brittle crack growth up to 320% to linear-elastic, brittle fracture above it (Fig. 3a & d and Fig. 4a). The fracture is mixed transgranular-intergranular in the case of the 320% sample while it proceeds along the inter-columnar boundary in the case of 420% and 520% drawing strain (Fig. 3). These results match very well with those reported by Hohenwarter et al. [9]. Compared to the as-drawn state, wires of drawing strain (420%) annealed at high temperatures (500 °C) exhibit substantial plastic deformation and no evidence of cracking (Fig. 4). Sputtered pure Fe thin film fail by brittle intergranular fracture (Fig. 6) and show a lower fracture toughness than the drawn wires. This is attributed to the absence of C in the pure unalloyed film in contrast to the cold drawn pearlite, where C is 168
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Fig. 6. (a) Snapshots of different stages of bending of sputtered Fe thin film micro-cantilever before final fracture. The crack plane is aligned perpendicular to the film plane in the crack arrestor orientation. (b) Corresponding load-displacement curve and (c) Fracture surface revealing the columnar nature of the film in the sputtering direction and subsequent brittle fracture. (d) Schematic showing orientation of the cantilever with respect to the Fe film/substrate combination. This cantilever orientation results in the crack-plane encountering multiple grains in its path.
cold drawn state is weak and can crack easily. This in turn can provide additional damage tolerance and toughening, when the wire is loaded in the transverse direction, as shown by Hohenwarter et al.[9]. A clear difference in the fracture behavior at the lower drawing strain of 320% was observed in comparison to the higher drawing strains of 420% and 520%. The load-displacement response showed a deviation from linear elastic deformation (Fig. 3a) and on loading cyclically, showed a gradual increase in compliance signaling slow crack growth (Fig. 2). In fact the J-integral measurements show that due to stable crack growth, the fracture toughness KC of the 320% as-drawn sample rises from ~10 to 27 MPam1/2 while the specimen records an increment in crack length from 0.63 to 0.76 µm (Fig. 2), in the specimen that was subjected to repeat load-unload cycles, signaling an Rcurve effect. Such stable crack growth was absent in the specimen subjected to higher drawing strains, they instead recording a near-catastrophic fracture at a single fracture load. Correspondingly, the microstructure at drawing strains of 320% and 420% as recorded independently by Li et al. [5] and Djaziri et al. [6] also showed significant differences. The wires which initially have distinct α-Fe and θ-Fe3C constituting phases with an interlamellar spacing of ~200 nm undergo extensive refinement to reach spacing of 20 nm up to a drawing strain of 320% [5]. Li et al., concluded that the extremely high dislocation density (> 1016/m2) accumulating at the ferrite/cementite interlamellar boundaries during severe cold drawing creates an avalanche effect sweeping the excess carbon from the decomposed cementite into the ferrite matrix [5]. Djaziri et al. showed that there is a sharp increase in tetragonality of the ferritic matrix as the drawing strain is increased from 320% to 420%, followed by a plateau value [6]. They correlated the C supersaturation in the ferritic matrix (due to cementite decomposition) and the high dislocation density to be responsible for the deformation induced martensite formation in heavily cold drawn pearlitic wires. With the help of atomistic simulations, they showed that such deformation induced martensite is stabilized at large drawing strains that the steels are subjected to during wire drawing owing to their high elastic limits.
enriched at the grain boundaries. The following discussion attempts to discern the role of different microstructural and compositional components (grain size, dislocation density, C as a solid solution strengthener vs segregator) in influencing the fracture properties of severely plastically deformed pearlitic steel in the as-drawn and annealed states. Factors are addressed with respect to their influence on the fracture behavior alone while their role on strengthening has already been extensively discussed by Li et al. [2].
4.1. Cold drawn The microstructure of cold drawn wires retain the fiber texture and columnar nature even at very high drawing strains [2,5]. The grain size (lamellar thickness) along the wire cross-section decreases from 18 to 10 nm for drawing strains from 320% to 520% as previously reported [5,6]. Because of the large volume fraction of nano-scaled interfaces/ sub-grains, fracture properties of these wires are primarily interface dominated. Hohenwarter et al. carried out macro-scale fracture tests on bulk high pressure torsion samples [4]. They found that the orientation where the pre-notch was parallel to the deformation direction and hence to the ferrite-cementite lamellae, showed a sharp drop in fracture toughness with increasing drawing strain, exhibiting a near cleavage fracture at higher strains. This is much like the cold drawn wires in the present study. Hohenwarter et al. [4,9] also observed large anisotropy in fracture behavior in SPD samples where the crack plane was aligned perpendicular to the lamellae, which showed at least a 10 times higher fracture toughness and extensive crack tortuosity due to crack kinking along the interfaces. This, along with our results indicates that a low fracture toughness in the longitudinal direction is a direct result of the weak inter-lamellar grain boundaries in cold drawn pearlitic steel wire coupled with the elongated grain structure offering a smooth and easy path for crack propagation. Micro-cracking parallel to the drawing direction seems to accompany the growth of the primary crack in our experiments for drawing strains of 320% and 420% (Fig. 7). This is further evidence that the inter-columnar ferritic grain boundary in the 169
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Fig. 7. (a) Micro-cracking along the inter-columnar boundaries, parallel to the drawing direction, in the 320% drawn wire, (b) schematic showing the crack propagation path in red, along the weak cohesion paths of the ferrite-cementite and ferrite-ferrite interfaces. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
drawing strain of 420%. This drawing strain represents a significant change in microstructure due to the nearly complete decomposition of cementite, and the onset of linear elastic brittle fracture along the intercolumnar boundaries. The critical annealing temperature at which the drop in tensile strength occurs is higher for lower drawing strains [18], which means the thermal stability of 420% drawing strain wires is expected to be intact up to 300 °C. Hence we choose two annealing temperatures: 200 and 500 °C for 30 min, at either ends of the spectrum, for our study. Fracture propagates along the inter-columnar boundaries in both the as-drawn sample and the sample annealed at 200 °C (Fig. 4). In contrast, crack propagation was not observed even after extensive plastic deformation in the sample annealed at 500 °C. The fracture toughness of samples annealed at 500 °C (average grain size of 130 nm) therefore could not be evaluated. In the former two cases, fracture along inter-columnar grain boundaries was continuous and clean, as though individual columns were peeled off, while in the latter, it was sluggish with several barriers preventing crack propagation. The yield strength of the samples after high temperature annealing drops considerably from 4.5 to 2 GPa following coarsening of the dislocation substructure, while the dislocation density itself is also lower (~1014/m2) after recovery [10]. The cantilevers were broken open post-mortem for fracture surface analysis. The fracture surface bears a clear signature of brittle cleavage fracture in the as-drawn and low temperature annealed sample, whereas void coalescence and dimpled ductile rupture is clearly visible on the high temperature annealed sample (Fig. 4). Microstructurally, there is no significant difference between the as-drawn and low temperature annealed wires. This is reflected in their very similar KC values and nearly identical fracture surfaces. It is also probable that the deformation induced tetragonality [6] is retained during low temperature annealing since the matrix remains supersaturated with C, as shown by APT measurements [10]. This will add to its brittle nature. In case of the high temperature annealed sample, re-formed cementite particles act as points of stress concentration for void nucleation but increased re-segregation of C to the grain boundaries enhances boundary cohesion, promoting ductile fracture. The sub-grain coarsening helps to promote dislocation activity in the grain interior, leading to extensive plasticity before fracture. This results in the low yield strength and a large plastic zone around the crack tip, rendering both plane-strain LEFM conditions as well as J-integral calculations invalid in micro-scale specimen. Fig. 5 summarizes the results of the above two sections, showing that fracture toughness follows an inverse relationship with yield strength in the case of cold drawn pearlitic steel wires. This implies that increasing nanocrystallinity leads to lower fracture toughness and brittle fracture, as is observed here. Samples with grain size above 200 nm show ductile fracture and fracture toughness above 50 MPam1/ 2 while those in the grain size regime of 20 nm and below exhibit brittle
It must be noted that there is still a small volume fraction of crystalline cementite in these heavily drawn wires, but they are non-stoichiometric and predominantly located at the ferritic grain boundaries [2]. In addition, the heavily cold drawn wires were reported to have a high vacancy concentration of 10−5−10−4 (compared to equilibrium vacancy concentration of 10−20 in as-patented Fe-C alloys), which agglomerates into clusters inside α-Fe [17]. These could operate as crack initiation sites. While the microstructure transitions from lamellar pearlite to nanostructured α-Fe with sub-grain boundaries, it inherits the basic grain orientation of the drawing process, albeit with an increasing fraction of low angle grain boundaries (LAGBs) formed by rearrangement of high density of dislocations at the former ferrite/cementite interfaces. A combination of a highly work hardened matrix with a nanocrystalline grain size therefore not only leads to high strength but severe loss in ductility and low fracture toughness, making these highly cold drawn steels increasingly prone to cleavage fracture and an expected change in fracture micro-mechanism from stable crack growth to catastrophic cracking. 4.2. Annealed Li et al. carried out microstructural investigations into the strength and ductility variations in heavily cold drawn wires (drawn to 650%) after several annealing treatments [7]. They observed that the tensile strength of these wires remained constant for low temperature annealing (< 200 °C) and showed a consistent and steady drop at higher (> 200 °C) annealing temperatures. Surprisingly, ductility does not monotonically increase as expected, and instead shows a rise up to the annealing temperature of 350 °C and then drops significantly above it. Li et al. [7,18] argue that the low temperature microstructural stability was a consequence of two opposing effects: slight reduction in dislocation density by dislocation annihilation and rearrangement, compensated by relaxation of internal stresses at the sub-grain boundaries, making them stronger obstacles for dislocation penetration. Carbon resegregation at the grain boundaries begins to occur at temperatures above 250 °C, resulting in formation of stoichiometric cementite again. C is a grain boundary stabilizer, known to improve grain boundary cohesion in α-Fe by reducing its energy [12]. Both together lead to achieving high temperature thermal stability to some extent. While the loss in strength was attributed to grain coarsening and spherodized carbide formation at the grain boundaries and triple junctions, the loss in ductility did not have satisfactory explanation. The loss in both strength and ductility will eventually dictate whether the fracture is brittle or ductile. We wanted to determine if annealing can bring about an enhancement in damage tolerance while not suffering a significant loss in strength, by examining the impact of low and high temperature annealing on the fracture behavior of the cold drawn steel at the transition 170
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
fracture and a Kc below 10 MPam1/2. Interfaces in pure nanocrystalline metals are prone to embrittlement, their high energies offering easier paths for crack propagation. Coarse-grained bcc-Fe has a fracture toughness above 200 MPam1/2 [19] but that of nanocrystalline SPD-Fe with an average 200 nm grain size is around 14 MPam1/2 [19]. The fracture toughness measurements on pure nc-Fe thin films (sputtered to 2.2 µm thickness) with an average in-plane grain size of 200 nm obtained within this study (Fig. 6) showed linear-elastic response with a KC as low as 2 MPam1/2 and unstable brittle fracture. This is the lowest fracture toughness reported for a nanocrystalline Fe based system, which is otherwise believed to be moderately ductile. Taking all these results into account it must be concluded that the composition and morphology of interfaces play an important role in the fracture behavior of nanocrystalline Fe based alloys. The individual contributions to fracture could not be separately quantified and require further microstructural characterization around the crack tip.
Acknowledgements B. N. J would like to thank Prof R. Pippan (Erich Schmid Institute of Materials Science, Austrian Academy of Sciences, Leoben, Austria.), Dr. A. Hohenwarter and Dr. B. Volkert (Department of Materials Physics, Montanuniversität Leoben, Austria) for useful discussions. References [1] M. Elices, Influence of residual stresses in the performance of cold-drawn pearlitic wires, J. Mater. Sci. 39 (12) (2004) 3889–3899. [2] Y.J. Li, P. Choi, C. Borchers, S. Westerkamp, S. Goto, D. Raabe, R. Kirchheim, Atomic-scale mechanisms of deformation-induced cementite decomposition in pearlite, Acta Mater. 59 (10) (2011) 3965–3977. [3] T.A. Zubkova, I.L. Yakovleva, L.E. Kar’kina, I.A. Veretennikova, Study of the hardness and elastic modulus of cementite in the structure of granular pearlite by the nano-indentation method, Metal. Sci. Heat. Treat. 56 (5) (2014) 330–335. [4] A. Hohenwarter, A. Taylor, R. Stock, R. Pippan, Effect of large shear deformations on the fracture behavior of a fully pearlitic steel, Metall. Mater. Trans. A 42 (6) (2011) 1609–1618. [5] Y. Li, D. Raabe, M. Herbig, P.-P. Choi, S. Goto, A. Kostka, H. Yarita, C. Borchers, R. Kirchheim, Segregation stabilizes nanocrystalline bulk steel with near theoretical strength, Phys. Rev. Lett. 113 (10) (2014) 106104. [6] S. Djaziri, Y. Li, G.A. Nematollahi, B. Grabowski, S. Goto, C. Kirchlechner, A. Kostka, S. Doyle, J. Neugebauer, D. Raabe, G. Dehm, Deformation-induced martensite: a new paradigm for exceptional steels, Adv. Mater. 28 (35) (2016) 7753–7757. [7] Y.J. Li, P. Choi, S. Goto, C. Borchers, D. Raabe, R. Kirchheim, Evolution of strength and microstructure during annealing of heavily cold-drawn 6.3 GPa hypereutectoid pearlitic steel wire, Acta Mater. 60 (9) (2012) 4005–4016. [8] B.N. Jaya, V. Jayaram, Fracture testing at small-length scales: from Plasticity in Si to brittleness in Pt, JOM 68 (1) (2016) 94–108. [9] A. Hohenwarter, B. Völker, M.W. Kapp, Y. Li, S. Goto, D. Raabe, R. Pippan, Ultrastrong and damage tolerant metallic bulk materials: a lesson from nanostructured pearlitic steel wires, Sci. Rep. 6 (2016) 33228. [10] Y.J. Li, A. Kostka, P. Choi, S. Goto, D. Ponge, R. Kirchheim, D. Raabe, Mechanisms of subgrain coarsening and its effect on the mechanical properties of carbon-supersaturated nanocrystalline hypereutectoid steel, Acta Mater. 84 (2015) 110–123. [11] Y. Ivanisenko, W. Lojkowski, R.Z. Valiev, H.J. Fecht, The mechanism of formation of nanostructure and dissolution of cementite in a pearlitic steel during high pressure torsion, Acta Mater. 51 (18) (2003) 5555–5570. [12] R. Wu, A.J. Freeman, G.B. Olson, Effects of carbon on Fe-grain-boundary cohesion: first-principles determination, Phys. Rev. B 53 (11) (1996) 7504–7509. [13] B.N. Jaya, C. Kirchlechner, G. Dehm, Can microscale fracture tests provide reliable fracture toughness values? A case study in silicon, J. Mater. Res. 30 (05) (2015) 686–698. [14] K. Matoy, H. Schönherr, T. Detzel, T. Schöberl, R. Pippan, C. Motz, G. Dehm, A comparative micro-cantilever study of the mechanical behavior of silicon based passivation films, Thin Solid Films 518 (1) (2009) 247–256. [15] S. Brinckmann, C. Kirchlechner, G. Dehm, Stress intensity factor dependence on anisotropy and geometry during micro-fracture experiments, Scr. Mater. 127 (2017) 76–78. [16] S. Wurster, C. Motz, R. Pippan, Characterization of the fracture toughness of microsized tungsten single crystal notched specimens, Philos. Mag. 92 (14) (2012) 1803–1825. [17] Y.Z. Chen, G. Csiszár, J. Cizek, C. Borchers, T. Ungár, S. Goto, R. Kirchheim, On the formation of vacancies in α-ferrite of a heavily cold-drawn pearlitic steel wire, Scr. Mater. 64 (5) (2011) 390–393. [18] J. Godet, F.A. El Nabi, S. Brochard, L. Pizzagalli, Surface effects on the mechanical behavior of silicon nanowires: consequence on the brittle to ductile transition at low scale and low temperature, Phys. Status Solidi (a) 212 (8) (2015) 1643–1648. [19] A. Hohenwarter, R. Pippan, Fracture and fracture toughness of nanopolycrystalline metals produced by severe plastic deformation, Philos. Trans. R. Soc. A: Math. Phys. Eng. Sci. 373 (2038) (2015).
5. Conclusion High strength in conjunction with sufficient damage tolerance is a necessary condition for advanced applications of cold drawn pearlitic steel wires as suspension bridge cables, rail tracks and escalators. The evaluation of micro-mechanism of fracture and its quantification in the weaker drawing direction is an essential first step in determining the damage tolerance of these steels in their operating conditions. In this work, we show that cold drawn pearlitic steels become increasingly brittle in the longitudinal direction with increasing drawing strain. This transition in fracture behavior is correlated to the structural and microstructural changes that the steel undergoes between drawing strains of 320–420%. The continuous evolution of carbon from its crystalline cementite phase into its elemental form during wire drawing and its resegregation into spheroidal cementite during subsequent heat treatment is reflected in its changing fracture behavior. The low fracture resistance exhibited by these steels at higher drawing strains can be attributed to the increasing degree of nanocrystallinity in combination with highly C supersaturated, increasingly tetragonal matrix. The fracture resistance of these steels is lowest in the longitudinal orientation since the crack propagates along the weaker inter-columnar boundary. A weak interface in the drawing direction acts as a precursor to enhanced damage tolerance in the transverse direction via initiation of toughening mechanisms like crack deflection, kinking and microcracking in such highly anisotropic materials. Low temperature annealing does not change either the microstructure or the fracture behavior of drawn wires compared to its cold drawn state. In contrast, high temperature annealing results in significant improvement in damage tolerance, accompanied by large-scale plasticity, although associated with a drastic reduction in strength. These results have technological implications in the optimal design of ultra-high tensile strength cold drawn wires with sufficient fracture resistance, necessary for sustainable design and safe service operations.
171
Contents lists available at ScienceDirect
Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea
Fracture behavior of nanostructured heavily cold drawn pearlitic steel wires before and after annealing ⁎
B.N. Jayaa, , S. Gotoa,b, G. Richterc, C. Kirchlechnera,d, G. Dehma,
MARK
⁎
a
Max Planck Institut fuer Eisenforschung GmbH, Duesseldorf, Germany Akita University, Tegata Gakuencho, Akita 010-8502, Japan c Max Planck Institute for Intelligent Systems, Stuttgart, Germany d Department of Material Physics, University of Leoben, Leoben, Austria b
A R T I C L E I N F O
A B S T R A C T
Keywords: Pearlitic steel Nanocrystalline Wire drawing Fracture Micromechanics
In situ micro-cantilever fracture testing is used to demonstrate changes in fracture behavior of nanostructured, heavily cold drawn pearlitic steel wires as a function of drawing strain and annealing conditions. It is shown that these steels exhibit a sharp transition in fracture behavior between a drawing strain of 320% and 520% with a drop in fracture toughness from 7.5 to 4 MPam1/2. This is confirmed from the nature of fracture which is stable with some degree of plasticity at drawing strains below 320% and changes to catastrophic cleavage fracture at drawing strains of 420% and above. This transition and associated brittleness is attributed to structural (cementite decomposition and strain induced increase in tetragonality) and microstructural (increasing nanocrystallinity and dislocation density) evolution that these steels undergo at higher drawing strains. On heat treating the 420% strained sample, brittle cleavage fracture continues for low temperature (200 °C) annealing with no visible changes in microstructure, while crack growth is suppressed and large-scale plasticity is recovered for high temperature (500 °C) annealing with accompanying grain coarsening, and re-precipitation of spherodized cementite at grain boundaries.
1. Introduction Cold drawn pearlitic steel, a nanolaminate composite structure consisting of alternate layers of ferrite and cementite, finds applications spanning from suspension bridge cables, rails and tire reinforcement cords to musical instruments and springs, due to its exceptional mechanical properties and the ability to be drawn into extremely thin wires [1]. The ductile α-Fe phase carries most of the strain while the harder Fe3C contributes to strengthening and work-hardening by acting as a barrier for dislocation movement. The mechanical performance of such a structure is dominated by the high density of ferrite-cementite interfaces. With increasing drawing strain, the individual colonies of pearlite become increasingly oriented along the drawing direction due to co-deformation with the ferrite matrix, developing a fiber texture along < 110 > . On further straining, the hard and brittle cementite phase breaks down and undergoes nano-scale refinement ([2] and references therein). The strength of cementite is known to be 10 times higher than that of the ferrite matrix in undeformed, as-patented pearlite [3]. Because of the widely different properties of ferrite and cementite, a very inhomogeneous stress distribution develops inside the pearlitic steels during wire drawing. ⁎
Severe plastic deformation (SPD) processes like high pressure torsion and cold drawing can produce steels with ultra-high tensile strengths, close to that of the constituent cementite [2,4]. Recently Li et al. [5], showed that tensile strengths close to 7 GPa could be achieved in heavily cold drawn pearlitic steel wire at drawing strains above 600%. Not surprisingly, these high strengths were accompanied by significant loss of ductility and linear elastic fracture when pulled in tension. Li et al. carried out correlative atom probe tomography (APT) and transmission electron microscopy (TEM), to establish that these steels undergo fundamental transformation in their structure following decomposition of the cementite phase and supersaturation of the excess carbon in the ferrite matrix at drawing strains above 300% [5]. Djaziri et al. were able to show using synchrotron based X-day diffraction that this supersaturation of carbon in the matrix results in an increase in its tetragonality, leading to formation of strain induced martensite [6]. In real applications like in suspension bridge cables, these cold drawn wires are subsequently subjected to various heat treatments in the process of galvanizing/coating them with anti-corrosive agents. This will expose the highly non-equilibrium cold drawn structure to temperatures high enough to bring about coarsening of their nanostructure and a loss of tensile strength. Thus, thermal stability evaluation
Correspondence to: Department of Metallurgical Engineering and Materials Science, IIT-Bombay, Mumbai, India. E-mail addresses: [email protected] (B.N. Jaya), [email protected] (G. Dehm).
http://dx.doi.org/10.1016/j.msea.2017.09.010 Received 27 March 2017; Received in revised form 3 July 2017; Accepted 2 September 2017 Available online 14 September 2017 0921-5093/ © 2017 Elsevier B.V. All rights reserved.
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
compound forms, to nanocrystalline Fe without any C in them. These comparisons are motivated by first principle calculations that indicate that C improves the interlamellar cohesion of ferrite, and its absence could enhance the brittle behavior [12].
is of paramount importance. On heat treating wires of drawing strain of 650%, Li et al. observed that the tensile strength is retained for low temperature annealing (< 250 °C) while there is a dramatic drop on annealing these wires at high temperatures (> 250 °C) [7]. In view of the evolving microstructure, the deformation and micro-fracture mechanism of these wires is expected to undergo significant transitions. These have immediate technological implications, when they are required to be exploited for their high strengths. Fracture behavior needs to be clearly understood as a function of drawing strain and annealing treatments to be able to prevent catastrophic fractures and provide sufficient damage tolerance at high drawing stains, while not compromising on strength. The micro-scale fracture behavior of cold drawn pearlitic steel wire in the longitudinal direction at high drawing strains were not explored earlier due to dimensional constraints of the limited wire diameter. The advent of in situ electron microscopy based micro-mechanical fracture testing [8] has changed this. Recently, Hohenwarter et al. [9], compared the damage tolerance of heavily cold drawn wires at two extreme drawing strains of 310% and 650%, both in the longitudinal and transverse directions. They identified and quantified the anisotropy in fracture behavior in the two directions, and observed catastrophic fracture in the longitudinal direction, and micro-cracking induced crack tip stress relaxation in the transverse direction. Their work provides valuable insights on the counter-intuitive contribution of the weak inter-columnar interface along the drawing direction enhancing the damage tolerance of these steel wires in the transverse direction. Our study expands on the work in the longitudinal direction, looking at intermediate drawing strains, and at different annealing conditions, where significant changes in microstructure have been reported [5,6,10,11]. Specifically, the objectives of the present work are to identify and quantify fracture behavior of cold drawn pearlitic steel wires in the drawing direction, in terms of their fracture toughness and fracture surface morphology a) at drawing strains of 320%, 420% and 520%, where a clear transition in microstructure has been observed from nanostructured lamellar pearlite composite, to nano-scaled sub-grained ferrite with supersaturated C [2,5,6], b) before and after low temperature (200 °C) and high temperature (500 °C) annealing for the transition drawing strain of 420% where a significant drop in tensile strength accompanied by grain coarsening has been recorded [10]. In addition to answering fundamental questions of structure-property correlations, quantification of fracture resistance in these wires will have a direct impact on critical applications where SPD steels are prone to fatigue, fracture and wear.
2.2. Micro-scale fracture testing and characterization The wires being in the range of tens to hundreds of microns in diameter, fracture toughness determination in the longitudinal drawing direction requires micro-mechanical testing. The validity of micro-scale fracture test geometries in determining the fracture toughness of brittle and semi-brittle systems has already been established [13–15]. Focused ion beam machined and notched (Zeiss AURIGA®) micro-cantilevers were tested in bending in situ in the scanning electron microscopy (SEM) (ASMEC UNAT-2 in JEOL JSM 2000) till fracture. Beam dimensions for all drawing strains were chosen to be in the same order of magnitude ~8 × 2 × 2 µm3 with a crack length to width (a/W) ratio ~0.3 (Table 1). All beams were Ga+ FIB machined at 4 nA and polished at 240 pA and eventually notched from the top at 10pA current at 30 kV. All micro-beams for a given drawing strain were machined on wires drawn from the same batch, but owing to the limited wire diameter, each wire piece could support 3–4 microbeams only (Fig. 1a-b). The loading was initially monotonic for all cases (Fig. 1c) but lead to different responses as a function of drawing strain or annealing conditions. For example, high drawing strains (≥ 420%) showed a purely elastic-brittle behavior and lower drawing strains (320%) showed significant deviation from linear elasticity (Fig. 1d). In such cases, cyclic loading with a series of load-partial unload-load sequence was carried out (see supplementary information). The latter was to enable J-integral calculations by recording the changing compliance during the unloading sequences of the load (P)-deflection (d) curve. Linear elastic fracture mechanics (LEFM) was used to determine KC for higher drawing strains (≥ 420%) (Eqs. (1) and (2)) [14], while it was used to determine only the lower limit of fracture toughness for lower drawing strain (320%). Fracture energy G and plastic zone size rp was evaluated according to Eqs. (3) and (4) respectively. Results were averaged over 5 or more micro-cantilevers for each case except for the annealed condition. Post-mortem high resolution imaging of the fracture surface and crack morphology was carried out in the Zeiss AURIGA® SEM.
Kc =
Pcrit L a f⎛ ⎞ BW 3/2 ⎝ W ⎠
a a a 2 a 3 f ⎛ ⎞ = 1.46 + 24.36 ⎛ ⎞ − 47.21 ⎛ ⎞ + 75.18 ⎛ ⎞ ⎝W ⎠ ⎝W ⎠ ⎝W ⎠ ⎝W ⎠
2. Experimental procedure 2.1. Material
G=J= Hypereutectoid steel wires of nominal composition Fe-0.98 wt%C (Fe–0.98C–0.31Mn–0.20Si–0.20Cr–0.01Cu–0.006P–0.007 S in wt%) supplied by Suzuki Metal Industry Co., Ltd. of drawing strains of 320%, 420% and 520% were used for the present study. Corresponding average lamellar thickness/grain sizes were 18, 15 and 10 nm respectively [2]. A second series of tests were carried out on annealed wires (200 °C and 500 °C for 30 min) of 420% drawing strain to evaluate the role of recovery and grain coarsening on the fracture behavior of these steels in comparison to the as-drawn state. Corresponding mean subgrain sizes were 19 and 130 nm respectively [10]. Table 1 shows the drawing strains, diameters and heat treatment conditions of the corresponding wires used. In addition, magnetron sputtered Fe thin films of 2.2 ± 0.06 µm thickness on SiO2 substrates were also used to evaluate the fracture behavior of pure unalloyed nanocrystalline α-Fe. The average in-plane grain size of these thin films was 200 nm ± 33 nm and these were columnar in nature with the grains oriented primarily along the < 110 > . These were used to compare the fracture toughness of pearlitic steels that contain C at the grain boundaries in either elemental or
KC 2 (1−ν 2) E
(1)
(2)
(3)
2
rp =
1 ⎛ KC ⎞ ⎜ ⎟ 2π ⎝ σy ⎠
(4)
where Kc is the stress intensity factor, a is the crack length, σ is the applied stress, Pcrit is the fracture load, and L, B, and W correspond to the cantilever length, breadth and width respectively, G is the fracture energy, J is the J-integral, E is the elastic modulus, ν is the Poisson's ratio, σy is the tensile yield strength. J-integral measurements were carried out by repeated load-unload cycles in the case of the 320% asdrawn samples, which showed some plasticity during the cantilever fracture experiments. The procedure is detailed below. The elasto-plastic fracture toughness in terms of JC values for the 320% drawing strain sample, was determined by the procedure as established by Wurster et al. [16] on two micro-cantilever specimens. Quantification of fracture toughness in the classical sense in more ductile systems at small length scales remains a challenge [8]. This is a problem for pearlitic wires of lower drawing strains. ASTM like standards of plane strain and small scale yielding are impossible to be 165
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Table 1 Specimen details listed in terms of cold drawing strains and heat treatment conditions, specimen dimensions and number of specimen tested for every condition. L: length, B: breadth, W: width and a: crack length of the cantilever beam, as indicated in Fig. 1. Sample Unit
Wire diameter µm
Cantilever dimensions: L×B×W µm
a/W
Statistics
As drawn: ε = 320% As drawn: ε = 420% As drawn: ε = 520% Annealed: 200 °C, 30 min ε = 420% Annealed: 500 °C, 30 min ε = 420% Fe thin film
115 66 42 66 66 2.2 (thickness)
8.2 × 1.5 × 2.2 7.8 × 2.2 × 1.8 8.5 × 1.7 × 1.9 10.5 × 1.9 × 2 11.5 × 2.2 × 2.3 15 × 2 × 2.5
0.25 0.3 0.3 0.28 0.25 0.28
5 8 6 4 4 5
of the cold drawn specimen of drawing strains 320% and 520% strain respectively while Table 2 summarises the KC values obtained from LEFM calculations for all of them. The as-drawn specimen corresponding to drawing strains of 420% and 520% show linear elastic deformation and catastrophic fracture with a fracture toughness of 5.2 MPam1/2 and 4 MPam1/2 respectively (Fig. 3d-f, Fig. 4a). The specimen corresponding to 320% drawing strain and lower show deviation from linear elastic behavior and a gradual drop in load during fracture, following plastic deformation and stable crack growth (Fig. 3a-c) with a corresponding LEFM fracture toughness of 7.5 MPam1/2. Fracture toughness calculations for these samples were repeated using J-integral measurements. The JC and JR were found to be 488 N/m and 3203 N/ m, respectively employing Eq. (6). Converting JC and JR to K values using Eq. (3) gives an initiation Kc of 10.6 MPam1/2 and a propagation KR of 27 MPam1/2 for a drawing strain of 320% (Fig. 2) at the end of the test. The procedural details of such a calculation are given in Wurster et al. [16]. This increase from 7.5 MPam1/2 at initiation (from LEFM) to 27 MPam1/2 after propagation (using J-integral) accompanied with stable crack growth hints at possible R-curve effect for this drawing strain, which remains to be systematically quantified. It is important to J note that the requirement for B > 25 σ is not met in this case and hence
complied at micro- and nano- dimensions. CTOD measurements could not be carried out due to lack of sufficient imaging resolution in our SEM. Even if J-integral measurements are carried out, only a lower limit of fracture energy can be ascertained due to specimen size limitations. Fig. 2a-b shows the P-d curve of a 320% strain sample subjected to cyclic loading with ~10% increments in displacement after each cycle. The specimen attained its maximum load at the fourth loading cycle and recorded gradual increase in compliance after that, corresponding to stable crack growth (Fig. 2c-d). The stiffness change was calculated from the unloading slopes after every cycle to correlate it to the change in crack length using Eq. (5) [16]. The change in crack length (W-ai) was in turn used to estimate both the initiation JC and propagation JR according to Eq. (6).
W − ai =
Ji =
kL3 BE
2Apl (i) (Ki )2 (1−ν 2) + E B (W − a 0)
(5)
(6)
where i represents increments in unloading cycles, Apl represents the plastic area under the P-d curve.
ys
these cannot be considered standard J-integral measurements. Post-mortem analysis of the fracture surface shows interface dominated cleavage fracture in all cases (Fig. 3c & f). The 320% drawn specimen shows a mixed transgranular and intergranular fracture with an average fracture feature spacing of 200 nm, while the 420% and
3. Results Fig. 3a & d shows the P-d curves (from monotonic loading) and Fig. 3b-c & e-f the corresponding crack trajectory and fracture surfaces
Fig. 1. (a)-(b) FIB micro-machined micro-cantilevers on cold drawn wire of 420% drawing strain. (c) Insitu micro-cantilever bending of the cantilever with the crack plane aligned along the longitudinal drawing direction, using a conical indenter, inside the SEM. (d) Load displacement curves showing responses of cold drawn wires of 320% and 520% drawing strains. A linear elastic response is seen for the 520% drawing strain while considerable deviation from elastic behavior is seen for the 320% drawing strain before fracture.
166
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Fig. 2. (a) A 320% drawn wire specimen with L = 7.09 µm, W = 2.2 µm, B = 1.55 µm and a0 = 0.62 µm, loaded cyclically to determine the elastoplastic fracture toughness JC. (b) Corresponding load-displacement response, (c) stiffness plot as a function of the number of unloading cycle showing a progressive increase in compliance, and (d) incremental increase in crack growth as a function of unloading cycle, calculated using Eq. (5) from the unloading stiffness. The final crack length measured from SEM is 0.8 µm which comes close to the crack length of 0.75 µm calculated from compliance changes.
While the 200 °C annealed specimen shows linear elastic brittle fracture like the as-drawn specimen, the 500 °C annealed specimen shows significant plastic deformation and no visible crack propagation even when subjected to large strains. While cleavage fracture features are part of the 200 °C annealed sample (Fig. 4-II), dimpled ductile fracture
520% drawn specimen show increasing tendency for fracture along the inter-columnar boundary parallel to the drawing direction with an average fracture feature spacing of 107 nm and 100 nm respectively. Fig. 4 compares the P-d curves and fracture features of the 420% asdrawn specimen with those annealed at 200 °C and 500 °C (30 min).
Fig. 3. Fracture response of cold drawn wires of drawing strain (a-c) 320% and (d-f) 520% showing the load (P)-displacement (d) characteristics, crack trajectory and fracture surface respectively. A distinct difference in fracture behavior is seen with the lower drawing strains showing elastic-plastic fracture with a stable and tortuous crack growth and mixed fracture, and the higher drawing strains showing brittle fracture with a straight path along the inter-columnar boundary. (b) and (e) are side surface views showing the crack trajectory. (c) and (f) are inclined views along the beams showing the initial notch (ao) and fracture morphology in the longitudinal direction.
167
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Table 2 Summary of fracture results of all tested specimen with calculated plastic zone size and fracture energy and associated microstructural parameters. Sample
daverage (Li et al. [5,7])
KC
σy (Li et al. [5,7])
rp
G
Unit As drawn: ε = 320% As drawn: ε = 420% As drawn: ε = 520% Annealed 200 °C: ε = 420% Annealed 500 °C: ε = 420% Fe thin film
nm 18
MPam1/2 7.5 ± 0.85
GPa 3.8
nm 620
J/m2 243
15
5.2 ± 0.94
5.2
159
117
10
4.0 ± 0.27
6.2
66
69
19
5.1 ± 0.65
5.2
153
113
~130
N. A.
1.9
N.A.
N.A.
200
2.0 ± 0.2
1.7*
220
51
Fig. 5. Fracture toughness of all the tested wires as a function of their corresponding yield strength showing a clear trend of brittle fracture and low fracture toughness for wires of high yield strengths corresponding to SPD states. Data from other works [4] have been added for completion. The number in brackets (n) is the average grain size.
stable crack growth occurred for the drawing strain of 320%, pointing to a larger process zone and activation of additional toughening mechanisms at lower drawing strains. Consequently, the plastic zone size of 620 nm is a large fraction of the specimen width of ~ 2 µm for the 320% strained sample (Table 2), implying that the Kc is only a lower estimate for the same. For the heavily cold drawn samples, the plastic zone size is nearly 1/10th the beam width (Table 2) and can be assumed to satisfy plane strain. The average fracture toughness of the 200 °C annealed specimen was 5.1 MPam1/2. This is very close to the as-drawn condition, while that of the 500 °C annealed specimen could not be evaluated due to extensive plastic deformation. It is clear from Fig. 5 that the fracture toughness follows an inverse relationship with yield strength and grain size in general for a given alloy system. The fracture toughness of the nanocrystalline sputtered pure Fe thin film (with average grain size of 200 nm) was much lower at 2 MPam1/2, making it the most brittle among all the specimens tested in the present work (Table 1). The crack path was intergranular, while the P-d response was linear elastic (Fig. 6). This shows that even for a comparatively larger grain size, and with the crack being oriented in a direction perpendicular to the columnar grains, pure sputtered Fe films exhibit a lower fracture resistance compared to nanostructured heavily cold drawn ferrite, with C segregated at its sub-grain boundaries. Although possessing very different defect structures, these results hint at C playing an important role in improving grain boundary cohesion in steels as compared to pure Fe.
Fig. 4. (a) Fracture response of 420% drawn wire under cold drawn and annealed states showing linear-elastic response for the cold-drawn and low-temperature annealed states and showing elasto-plastic deformation for the high-temperature annealed state. The fracture surfaces reveal brittle inter-columnar boundary fracture for the former two states, while exhibiting dimpled, ductile fracture for the latter. (b) FIB polished and exposed surfaces of as-drawn and high temperature annealed samples under channeling contrast, showing evidence of grain coarsening.
4. Discussion
features were prominently visible for the 500 °C annealed sample (Fig. 4-III) (when fractured artificially after the loading sequence). This clearly shows the difference in fracture behavior due to microstructural changes occurring during the 500 °C annealing [10]. Fig. 5 shows the initiation fracture toughness calculated for all the specimens tested in the present work, superimposed with the results from equivalent strains of SPD specimens (as reported earlier by Hohenwarter et al. [4]), plotted as a function of their yield strength. It was seen that wires with 200% drawing strain or lower showed large scale ductility due to which their fracture behavior could not be quantified using existing micro-mechanical test methods and inapplicability of LEFM. Micro-scale fracture tests carried out for drawing strains of 320%, 420% and 520% revealed a significant drop in the initiation KC from 7.5 to 5.2 to 4 MPam1/2 respectively (Table 2 and Fig. 5). The corresponding yield strength data from [5] is also repeated in Table 2. Fracture was unstable and catastrophic for drawing strains ≥ 420% but
Our micro-cantilever fracture toughness evaluation of cold drawn pearlitic steel wire tested in the longitudinal drawing direction shows a clear drop in fracture toughness with increasing drawing strain, transitioning from stable, elastic-plastic, semi-brittle crack growth up to 320% to linear-elastic, brittle fracture above it (Fig. 3a & d and Fig. 4a). The fracture is mixed transgranular-intergranular in the case of the 320% sample while it proceeds along the inter-columnar boundary in the case of 420% and 520% drawing strain (Fig. 3). These results match very well with those reported by Hohenwarter et al. [9]. Compared to the as-drawn state, wires of drawing strain (420%) annealed at high temperatures (500 °C) exhibit substantial plastic deformation and no evidence of cracking (Fig. 4). Sputtered pure Fe thin film fail by brittle intergranular fracture (Fig. 6) and show a lower fracture toughness than the drawn wires. This is attributed to the absence of C in the pure unalloyed film in contrast to the cold drawn pearlite, where C is 168
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Fig. 6. (a) Snapshots of different stages of bending of sputtered Fe thin film micro-cantilever before final fracture. The crack plane is aligned perpendicular to the film plane in the crack arrestor orientation. (b) Corresponding load-displacement curve and (c) Fracture surface revealing the columnar nature of the film in the sputtering direction and subsequent brittle fracture. (d) Schematic showing orientation of the cantilever with respect to the Fe film/substrate combination. This cantilever orientation results in the crack-plane encountering multiple grains in its path.
cold drawn state is weak and can crack easily. This in turn can provide additional damage tolerance and toughening, when the wire is loaded in the transverse direction, as shown by Hohenwarter et al.[9]. A clear difference in the fracture behavior at the lower drawing strain of 320% was observed in comparison to the higher drawing strains of 420% and 520%. The load-displacement response showed a deviation from linear elastic deformation (Fig. 3a) and on loading cyclically, showed a gradual increase in compliance signaling slow crack growth (Fig. 2). In fact the J-integral measurements show that due to stable crack growth, the fracture toughness KC of the 320% as-drawn sample rises from ~10 to 27 MPam1/2 while the specimen records an increment in crack length from 0.63 to 0.76 µm (Fig. 2), in the specimen that was subjected to repeat load-unload cycles, signaling an Rcurve effect. Such stable crack growth was absent in the specimen subjected to higher drawing strains, they instead recording a near-catastrophic fracture at a single fracture load. Correspondingly, the microstructure at drawing strains of 320% and 420% as recorded independently by Li et al. [5] and Djaziri et al. [6] also showed significant differences. The wires which initially have distinct α-Fe and θ-Fe3C constituting phases with an interlamellar spacing of ~200 nm undergo extensive refinement to reach spacing of 20 nm up to a drawing strain of 320% [5]. Li et al., concluded that the extremely high dislocation density (> 1016/m2) accumulating at the ferrite/cementite interlamellar boundaries during severe cold drawing creates an avalanche effect sweeping the excess carbon from the decomposed cementite into the ferrite matrix [5]. Djaziri et al. showed that there is a sharp increase in tetragonality of the ferritic matrix as the drawing strain is increased from 320% to 420%, followed by a plateau value [6]. They correlated the C supersaturation in the ferritic matrix (due to cementite decomposition) and the high dislocation density to be responsible for the deformation induced martensite formation in heavily cold drawn pearlitic wires. With the help of atomistic simulations, they showed that such deformation induced martensite is stabilized at large drawing strains that the steels are subjected to during wire drawing owing to their high elastic limits.
enriched at the grain boundaries. The following discussion attempts to discern the role of different microstructural and compositional components (grain size, dislocation density, C as a solid solution strengthener vs segregator) in influencing the fracture properties of severely plastically deformed pearlitic steel in the as-drawn and annealed states. Factors are addressed with respect to their influence on the fracture behavior alone while their role on strengthening has already been extensively discussed by Li et al. [2].
4.1. Cold drawn The microstructure of cold drawn wires retain the fiber texture and columnar nature even at very high drawing strains [2,5]. The grain size (lamellar thickness) along the wire cross-section decreases from 18 to 10 nm for drawing strains from 320% to 520% as previously reported [5,6]. Because of the large volume fraction of nano-scaled interfaces/ sub-grains, fracture properties of these wires are primarily interface dominated. Hohenwarter et al. carried out macro-scale fracture tests on bulk high pressure torsion samples [4]. They found that the orientation where the pre-notch was parallel to the deformation direction and hence to the ferrite-cementite lamellae, showed a sharp drop in fracture toughness with increasing drawing strain, exhibiting a near cleavage fracture at higher strains. This is much like the cold drawn wires in the present study. Hohenwarter et al. [4,9] also observed large anisotropy in fracture behavior in SPD samples where the crack plane was aligned perpendicular to the lamellae, which showed at least a 10 times higher fracture toughness and extensive crack tortuosity due to crack kinking along the interfaces. This, along with our results indicates that a low fracture toughness in the longitudinal direction is a direct result of the weak inter-lamellar grain boundaries in cold drawn pearlitic steel wire coupled with the elongated grain structure offering a smooth and easy path for crack propagation. Micro-cracking parallel to the drawing direction seems to accompany the growth of the primary crack in our experiments for drawing strains of 320% and 420% (Fig. 7). This is further evidence that the inter-columnar ferritic grain boundary in the 169
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
Fig. 7. (a) Micro-cracking along the inter-columnar boundaries, parallel to the drawing direction, in the 320% drawn wire, (b) schematic showing the crack propagation path in red, along the weak cohesion paths of the ferrite-cementite and ferrite-ferrite interfaces. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
drawing strain of 420%. This drawing strain represents a significant change in microstructure due to the nearly complete decomposition of cementite, and the onset of linear elastic brittle fracture along the intercolumnar boundaries. The critical annealing temperature at which the drop in tensile strength occurs is higher for lower drawing strains [18], which means the thermal stability of 420% drawing strain wires is expected to be intact up to 300 °C. Hence we choose two annealing temperatures: 200 and 500 °C for 30 min, at either ends of the spectrum, for our study. Fracture propagates along the inter-columnar boundaries in both the as-drawn sample and the sample annealed at 200 °C (Fig. 4). In contrast, crack propagation was not observed even after extensive plastic deformation in the sample annealed at 500 °C. The fracture toughness of samples annealed at 500 °C (average grain size of 130 nm) therefore could not be evaluated. In the former two cases, fracture along inter-columnar grain boundaries was continuous and clean, as though individual columns were peeled off, while in the latter, it was sluggish with several barriers preventing crack propagation. The yield strength of the samples after high temperature annealing drops considerably from 4.5 to 2 GPa following coarsening of the dislocation substructure, while the dislocation density itself is also lower (~1014/m2) after recovery [10]. The cantilevers were broken open post-mortem for fracture surface analysis. The fracture surface bears a clear signature of brittle cleavage fracture in the as-drawn and low temperature annealed sample, whereas void coalescence and dimpled ductile rupture is clearly visible on the high temperature annealed sample (Fig. 4). Microstructurally, there is no significant difference between the as-drawn and low temperature annealed wires. This is reflected in their very similar KC values and nearly identical fracture surfaces. It is also probable that the deformation induced tetragonality [6] is retained during low temperature annealing since the matrix remains supersaturated with C, as shown by APT measurements [10]. This will add to its brittle nature. In case of the high temperature annealed sample, re-formed cementite particles act as points of stress concentration for void nucleation but increased re-segregation of C to the grain boundaries enhances boundary cohesion, promoting ductile fracture. The sub-grain coarsening helps to promote dislocation activity in the grain interior, leading to extensive plasticity before fracture. This results in the low yield strength and a large plastic zone around the crack tip, rendering both plane-strain LEFM conditions as well as J-integral calculations invalid in micro-scale specimen. Fig. 5 summarizes the results of the above two sections, showing that fracture toughness follows an inverse relationship with yield strength in the case of cold drawn pearlitic steel wires. This implies that increasing nanocrystallinity leads to lower fracture toughness and brittle fracture, as is observed here. Samples with grain size above 200 nm show ductile fracture and fracture toughness above 50 MPam1/ 2 while those in the grain size regime of 20 nm and below exhibit brittle
It must be noted that there is still a small volume fraction of crystalline cementite in these heavily drawn wires, but they are non-stoichiometric and predominantly located at the ferritic grain boundaries [2]. In addition, the heavily cold drawn wires were reported to have a high vacancy concentration of 10−5−10−4 (compared to equilibrium vacancy concentration of 10−20 in as-patented Fe-C alloys), which agglomerates into clusters inside α-Fe [17]. These could operate as crack initiation sites. While the microstructure transitions from lamellar pearlite to nanostructured α-Fe with sub-grain boundaries, it inherits the basic grain orientation of the drawing process, albeit with an increasing fraction of low angle grain boundaries (LAGBs) formed by rearrangement of high density of dislocations at the former ferrite/cementite interfaces. A combination of a highly work hardened matrix with a nanocrystalline grain size therefore not only leads to high strength but severe loss in ductility and low fracture toughness, making these highly cold drawn steels increasingly prone to cleavage fracture and an expected change in fracture micro-mechanism from stable crack growth to catastrophic cracking. 4.2. Annealed Li et al. carried out microstructural investigations into the strength and ductility variations in heavily cold drawn wires (drawn to 650%) after several annealing treatments [7]. They observed that the tensile strength of these wires remained constant for low temperature annealing (< 200 °C) and showed a consistent and steady drop at higher (> 200 °C) annealing temperatures. Surprisingly, ductility does not monotonically increase as expected, and instead shows a rise up to the annealing temperature of 350 °C and then drops significantly above it. Li et al. [7,18] argue that the low temperature microstructural stability was a consequence of two opposing effects: slight reduction in dislocation density by dislocation annihilation and rearrangement, compensated by relaxation of internal stresses at the sub-grain boundaries, making them stronger obstacles for dislocation penetration. Carbon resegregation at the grain boundaries begins to occur at temperatures above 250 °C, resulting in formation of stoichiometric cementite again. C is a grain boundary stabilizer, known to improve grain boundary cohesion in α-Fe by reducing its energy [12]. Both together lead to achieving high temperature thermal stability to some extent. While the loss in strength was attributed to grain coarsening and spherodized carbide formation at the grain boundaries and triple junctions, the loss in ductility did not have satisfactory explanation. The loss in both strength and ductility will eventually dictate whether the fracture is brittle or ductile. We wanted to determine if annealing can bring about an enhancement in damage tolerance while not suffering a significant loss in strength, by examining the impact of low and high temperature annealing on the fracture behavior of the cold drawn steel at the transition 170
Materials Science & Engineering A 707 (2017) 164–171
B.N. Jaya et al.
fracture and a Kc below 10 MPam1/2. Interfaces in pure nanocrystalline metals are prone to embrittlement, their high energies offering easier paths for crack propagation. Coarse-grained bcc-Fe has a fracture toughness above 200 MPam1/2 [19] but that of nanocrystalline SPD-Fe with an average 200 nm grain size is around 14 MPam1/2 [19]. The fracture toughness measurements on pure nc-Fe thin films (sputtered to 2.2 µm thickness) with an average in-plane grain size of 200 nm obtained within this study (Fig. 6) showed linear-elastic response with a KC as low as 2 MPam1/2 and unstable brittle fracture. This is the lowest fracture toughness reported for a nanocrystalline Fe based system, which is otherwise believed to be moderately ductile. Taking all these results into account it must be concluded that the composition and morphology of interfaces play an important role in the fracture behavior of nanocrystalline Fe based alloys. The individual contributions to fracture could not be separately quantified and require further microstructural characterization around the crack tip.
Acknowledgements B. N. J would like to thank Prof R. Pippan (Erich Schmid Institute of Materials Science, Austrian Academy of Sciences, Leoben, Austria.), Dr. A. Hohenwarter and Dr. B. Volkert (Department of Materials Physics, Montanuniversität Leoben, Austria) for useful discussions. References [1] M. Elices, Influence of residual stresses in the performance of cold-drawn pearlitic wires, J. Mater. Sci. 39 (12) (2004) 3889–3899. [2] Y.J. Li, P. Choi, C. Borchers, S. Westerkamp, S. Goto, D. Raabe, R. Kirchheim, Atomic-scale mechanisms of deformation-induced cementite decomposition in pearlite, Acta Mater. 59 (10) (2011) 3965–3977. [3] T.A. Zubkova, I.L. Yakovleva, L.E. Kar’kina, I.A. Veretennikova, Study of the hardness and elastic modulus of cementite in the structure of granular pearlite by the nano-indentation method, Metal. Sci. Heat. Treat. 56 (5) (2014) 330–335. [4] A. Hohenwarter, A. Taylor, R. Stock, R. Pippan, Effect of large shear deformations on the fracture behavior of a fully pearlitic steel, Metall. Mater. Trans. A 42 (6) (2011) 1609–1618. [5] Y. Li, D. Raabe, M. Herbig, P.-P. Choi, S. Goto, A. Kostka, H. Yarita, C. Borchers, R. Kirchheim, Segregation stabilizes nanocrystalline bulk steel with near theoretical strength, Phys. Rev. Lett. 113 (10) (2014) 106104. [6] S. Djaziri, Y. Li, G.A. Nematollahi, B. Grabowski, S. Goto, C. Kirchlechner, A. Kostka, S. Doyle, J. Neugebauer, D. Raabe, G. Dehm, Deformation-induced martensite: a new paradigm for exceptional steels, Adv. Mater. 28 (35) (2016) 7753–7757. [7] Y.J. Li, P. Choi, S. Goto, C. Borchers, D. Raabe, R. Kirchheim, Evolution of strength and microstructure during annealing of heavily cold-drawn 6.3 GPa hypereutectoid pearlitic steel wire, Acta Mater. 60 (9) (2012) 4005–4016. [8] B.N. Jaya, V. Jayaram, Fracture testing at small-length scales: from Plasticity in Si to brittleness in Pt, JOM 68 (1) (2016) 94–108. [9] A. Hohenwarter, B. Völker, M.W. Kapp, Y. Li, S. Goto, D. Raabe, R. Pippan, Ultrastrong and damage tolerant metallic bulk materials: a lesson from nanostructured pearlitic steel wires, Sci. Rep. 6 (2016) 33228. [10] Y.J. Li, A. Kostka, P. Choi, S. Goto, D. Ponge, R. Kirchheim, D. Raabe, Mechanisms of subgrain coarsening and its effect on the mechanical properties of carbon-supersaturated nanocrystalline hypereutectoid steel, Acta Mater. 84 (2015) 110–123. [11] Y. Ivanisenko, W. Lojkowski, R.Z. Valiev, H.J. Fecht, The mechanism of formation of nanostructure and dissolution of cementite in a pearlitic steel during high pressure torsion, Acta Mater. 51 (18) (2003) 5555–5570. [12] R. Wu, A.J. Freeman, G.B. Olson, Effects of carbon on Fe-grain-boundary cohesion: first-principles determination, Phys. Rev. B 53 (11) (1996) 7504–7509. [13] B.N. Jaya, C. Kirchlechner, G. Dehm, Can microscale fracture tests provide reliable fracture toughness values? A case study in silicon, J. Mater. Res. 30 (05) (2015) 686–698. [14] K. Matoy, H. Schönherr, T. Detzel, T. Schöberl, R. Pippan, C. Motz, G. Dehm, A comparative micro-cantilever study of the mechanical behavior of silicon based passivation films, Thin Solid Films 518 (1) (2009) 247–256. [15] S. Brinckmann, C. Kirchlechner, G. Dehm, Stress intensity factor dependence on anisotropy and geometry during micro-fracture experiments, Scr. Mater. 127 (2017) 76–78. [16] S. Wurster, C. Motz, R. Pippan, Characterization of the fracture toughness of microsized tungsten single crystal notched specimens, Philos. Mag. 92 (14) (2012) 1803–1825. [17] Y.Z. Chen, G. Csiszár, J. Cizek, C. Borchers, T. Ungár, S. Goto, R. Kirchheim, On the formation of vacancies in α-ferrite of a heavily cold-drawn pearlitic steel wire, Scr. Mater. 64 (5) (2011) 390–393. [18] J. Godet, F.A. El Nabi, S. Brochard, L. Pizzagalli, Surface effects on the mechanical behavior of silicon nanowires: consequence on the brittle to ductile transition at low scale and low temperature, Phys. Status Solidi (a) 212 (8) (2015) 1643–1648. [19] A. Hohenwarter, R. Pippan, Fracture and fracture toughness of nanopolycrystalline metals produced by severe plastic deformation, Philos. Trans. R. Soc. A: Math. Phys. Eng. Sci. 373 (2038) (2015).
5. Conclusion High strength in conjunction with sufficient damage tolerance is a necessary condition for advanced applications of cold drawn pearlitic steel wires as suspension bridge cables, rail tracks and escalators. The evaluation of micro-mechanism of fracture and its quantification in the weaker drawing direction is an essential first step in determining the damage tolerance of these steels in their operating conditions. In this work, we show that cold drawn pearlitic steels become increasingly brittle in the longitudinal direction with increasing drawing strain. This transition in fracture behavior is correlated to the structural and microstructural changes that the steel undergoes between drawing strains of 320–420%. The continuous evolution of carbon from its crystalline cementite phase into its elemental form during wire drawing and its resegregation into spheroidal cementite during subsequent heat treatment is reflected in its changing fracture behavior. The low fracture resistance exhibited by these steels at higher drawing strains can be attributed to the increasing degree of nanocrystallinity in combination with highly C supersaturated, increasingly tetragonal matrix. The fracture resistance of these steels is lowest in the longitudinal orientation since the crack propagates along the weaker inter-columnar boundary. A weak interface in the drawing direction acts as a precursor to enhanced damage tolerance in the transverse direction via initiation of toughening mechanisms like crack deflection, kinking and microcracking in such highly anisotropic materials. Low temperature annealing does not change either the microstructure or the fracture behavior of drawn wires compared to its cold drawn state. In contrast, high temperature annealing results in significant improvement in damage tolerance, accompanied by large-scale plasticity, although associated with a drastic reduction in strength. These results have technological implications in the optimal design of ultra-high tensile strength cold drawn wires with sufficient fracture resistance, necessary for sustainable design and safe service operations.
171