Materials Science and Engineering, A125 (1990) 43-48
Ductile Fracture Mechanisms in A I S 1 4 3 4 0 Steel J. K. CUDDY and M. N. BASSIM
Metallurgical Sciences Laboratories, Department of Mechanical Engineering, University of Manitoba, Winnipeg, Manitoba R3T 2N2 (Canada) (Received August 4, 1989; in revised form October 17, 1989)
The ductile fracture process in metallic materials is initiated by inhomogeneous plastic deformation, which in some metals is manifested as a dislocation cell substructure. The role of welldeveloped dislocation cells in the fracture process of AISI 4340 steel was investigated using transmission (TEM) and scanning electron microscopy (SEM). It was found that the ferrite and pearlite constituents of the steel exhibit different dislocation substructures, with cells forming in ferrite grains and shear bands occurring in the ferrite lamellae of pearlite. The combined TEM and SEM results suggest that microcracks are initiated at cementite platelets in pearlite under the combined action of tensile loading and localized shear in the adjacent ferrite. Microcracks then propagate by the nucleation and coalescence of voids along the well-developed dislocation cell walls in the ferrite grains.
The fracture process in ductile metals is characterized microscopically by the nucleation and growth of voids which coalesce to produce the macroscopic fracture surface. Since this process is due to the preceding localization of plastic deformation at microstructural heterogeneities in materials, the importance of inhomogeneous arrangements of dislocations in the fracture process is unequivocal. The majority of experimental studies, including the work of Brock  on aluminium alloys, have correlated the sites of localized plastic deformation with second-phase particles and inclusions. A number of criteria have been proposed by several investigators [2-5] to describe the initiation and growth of voids at the particle-matrix interface. 0921-5093/90/$3.50
For materials containing second-phase particles there is an established relationship between these particles and mechanisms of ductile fracture. However, Wilsdorf  has pointed out that the presence of particles is not a prerequisite for void nucleation. In metals with a low particle content, for example, void nucleation has been observed at grain boundaries and grain boundary triple points . Bauer and Wilsdorf  correlated void initiation with vacancy clusters in AISI 304 stainless steel, based on the evidence of a measured void density 100 times larger than the average particle density. A series of investigations by Wilsdorf and coworkers [9-11] was the first to consider the role of well-developed dislocation structures in the micromechanisms of ductile fracture. They observed microcrack propagation by the formation and coalescence of voids at dislocation cell walls during in situ high voltage electron microscopy (HVEM) straining experiments on a number of pure metals. Wilsdorf  has suggested that the reduction of the strain energy due to the misorientation of the cell walls is the driving force for void nucleation (in materials with a low particle content). This is supported by recent studies on low carbon steels by Bassim et al.  and Yumen and Jingen , which correlated the minimum dislocation cell size observed by transmission electron microscopy (TEM) with the size distribution of voids on fracture surfaces as determined by scanning electron microscopy (SEM). These studies clearly show the importance of the preceding plastic deformation and, in particular, the occurrence of inhomogeneous dislocation arrangements in the fracture process of ductile metals. The present work examines the fracture process in a two-phase steel containing a microstructure of ferrite and pearlite. The objective is to further understand the relationship between © Elsevier Sequoia/Printed in The Netherlands
44 TABLE 1 Chemicalcompositionof 4340 steel Element Amount
F ( ~ , x i = local distance from gauge mark
SEM fracture surface
2000FX electron microscope operating at 200 kV to reveal the dislocation substructure. The mean size of dislocation cells in specimens deformed to different plastic strains was determined from T E M micrographs using linear intercept measurements.
d;i = local diameter of slice
Fig. 1. Method of obtaining TEM and SEM samples from fractured tensile specimensof 4340 steel. plastic deformation, in terms of dislocation cell formation, and void mechanisms of ductile fracture. The individual fracture behaviour of the fen'ite and pearlite constituents in the steel is also considered. 2. Experimental procedure
The material used in this investigation was a low alloy steel conforming in chemical composition to that of AIS14340 as given in Table 1. Samples from the as-received material were austenitized at 840 °C followed by furnace cooling to produce an annealed microstructure of proeutectoid ferrite grains and pearlite colonies. Cylindrical tensile specimens of 6.35 mm (0.25 in) diameter were then prepared and loaded to fracture at an initial strain rate of 3.3 x 10 -4 S -1. Following the mechanical testing, the fracture surfaces were removed and transverse slices were sectioned from the broken tensile halves as shown in Fig. I to obtain samples for microscopic examination by SEM and T E M respectively. The topographical features of the tensile fracture surfaces were studied using a JEOL JXA840 scanning microanalyser operating at 25 kV. The statistical distribution of void (or dimple) sizes was determined using a mean linear intercept method on SEM micrographs taken from random areas of the fracture surfaces. The transverse slices were thinned for TEM by mechanical grinding followed by electropolishing with 5% perchloric acid and 95% methanol at - 5 0 °C and 18 V until perforation. The thin foil specimens were then examined with a JEOL
3. Experimental results
The deformed substructure of ferrite and pearhte regions in the steel is typified by the TEM micrographs in Figs. 2(a) and 2(b). Figure 2(a) illustrates the well-defined dislocation cell structure formed in proeutectoid ferrite grains as a result of tensile deformation. The variation in dislocation cell size with plastic strain is given in Fig. 3. From this figure it can be seen that the cell size decreases with strains up to 0.40-0.50, whereupon the cell size remains about constant at 0.2-0.3 /zm with further straining to fracture. Such behaviour is predicted by the "mesh length" theory of work hardening of Kuhlmann-Wilsdorf [14, 15] in which the size of dislocation cells is an inverse function of the applied level of stress and strain during deformation. The substructure of pearlite in Fig. 2(b) is quite different from that observed in proeutectoid ferrite grains. Profuse dislocation activity in the interlamellar ferrite phase is generally constrained by the neighbouring cementite (Fe3C) platelets which act as impervious barriers to dislocation movement, and thus only an "incipient cell structure" is observed. In some regions, however, separation of the Fe3C platelets is observed in conjunction with dense bands of dislocations as indicated by arrows in Fig. 2(b). SEM photomicrographs of the tensile fracture surfaces are shown in Figs. 4 and 5. The fine distribution of voids or "dimples" in Fig. 4 shows that the annealed 4340 steel failed by a fibrous mode of fracture. The voids in this micrograph exhibit a fairly wide variation in size and shape. The elongated shape of some voids may indicate the occurrence of localized shear stresses in addition to tensile stresses during their formation . It is also noticeable that many of the larger voids present are actually composed of many smaller,
0.2 0.3 0.4 0.5
Fig. 3. Variation of dislocation cell size with plastic strain.
Fig. 4. SEM micrograph of fibrous fracture in the ferrite.
Fig. 2. TEM micrographs of dislocation cell structures (a) in ferrite and (b) in pearlite.
shallow voids. There is a general lack of evidence for inclusions or particles in conjunction with these voids. This is to be expected owing to the maintenance of low concentrations of sulphur, oxygen and nitrogen during production of 4340 steel, which minimizes the inclusion content. A
statistical analysis of the distribution of void sizes is given in Fig. 6. The histogram shows a fairly wide distribution, with a mean void size of about 0.8/am. The fracture surface topography in regions containing pearlite is characterized by microcracks running normal to the pearlite colonies as illustrated in Fig. 5(a). At the higher magnifications shown in Figs. 5(b) and 5(c), shallow microvoids of size less than 0.5 /~m are visible in the interlamellar regions between eementile platelets, which appear as bright, parallel ridges in the mierographs. This indicates that a void mechanism of fracture is also applicable to the pearlite colonies, although they display a fracture surface topography which is markedly different from that observed for ferrite grains.
~ 3(~' 200" 10t3. 0 0
Fig. 6. Void size distribution in fractured 4340 steel.
Fig. 5. Fracture surface of the pearlite showing (a) evidence of cleavage and microcracks and (b), (c) evidence of microvoids in interlammellar ferrite.
The results of this study indicate that ductile fracture of 4340 steel is controlled by void mechanisms in both ferrite grains and pearlite colonies. However, the fracture surface topog-
raphy of ferrite and pearlite reflects a difference in the processes of void nucleation. This is best understood by considering the individual deformation behaviour of the two constituents prior to fracture. In ferrite grains, plastic deformation leads to the formation of a dislocation cell substructure which decreases in size with continued tensile strain (as shown in Fig. 3). The final cell size of 0.2-0.3/~m attained before the onset of fracture coincides closely with the size distribution of voids on the fracture surfaces given in Fig. 6. This, coupled with the low inclusion content in the steel and the lack of evidence for inclusions associated with voids, would suggest that the walls of the well-developed dislocation cell substructure are the probable sites for void nucleation in ferrite grains. This conclusion is supported by the HVEM studies of Wilsdorf and coworkers [9-11] on pure metals, and by similar correlations between dislocation cell size and fracture surface void size distributions found in steels by other investigators [12, 13]. The results of the TEM part of this investigation show that plastic deformation of the pearlite constituent prior to fracture is microscopically characterized by the separation of cementile plates in conjunction with the accumulation of dislocations into shear bands in the adjacent ferrite, as seen in Fig. 2(b). The process by which this inhomogeneous plastic deformation leads to the formation of the voids and microcracks observed on the pearlite fracture surface by SEM can be explained by considering a model proposed earlier by Miller and Smith . They proposed a mechanism for the fracture of pearlite based on a detailed study of a series of steels
Crack propagation through the ferrite grains proceeds by void initiation and coalescence at the dislocation cell walls, leading to the ductile fracture of the material.
I bt t
> Fig. 7. Fracture mechanism in pearlite (after ref. 17): (a) cracking of a cementile plate; (b) shear zone developing in ferrite causing cracking of adjacent plates; (c), (d) void formation and coalescence.
containing pearlite. In their model, as depicted in Figs. 7(a)-7(d), separation of a cementite plate occurs under the combined influence of applied tensile stress and localized shear in the ferrite lamellae (Fig. 7(a)). Deformation becomes concentrated in a shear band (Fig. 7(b)), causing the separation of adjacent Fe3C plates and leading to void growth and coalescence (Figs. 7(c) and 7(d)). The final ductile fracture surface then consists of shallow dimples between the broken cementite plates. This is strikingly similar to the fracture topography of a pearlite colony in the steel of the present study, as shown in Fig. 5(c). It would thus be expected that void formation and microcracking are initiated in pearlite colonies which are most closely aligned with the tensile axis of the specimen. These results form a picture of the ductile fracture process in 4340 steel. Microcracks, nucleated at Fe3C platelets in pearlite owing to the applied tensile stress and localized shear in the adjacent ferfite, propagate through the colonies and encounter softer ferrite grains.
5. Conclusions The relationship between the ductile fracture process and the microstructural aspects of plastic deformation preceding fracture was studied in commercial AIS14340 steel. Transmission electron microscopy reveals that tensile deformation leads to the formation of inhomogeneous dislocation arrangements; namely, dislocation cells in ferrite grains and shear bands in the ferrite lamellae of pearlite. The results of scanning electron microscopy suggest that microcracks are initiated at cementite platelets in pearlite under the combined influence of the applied tensile load and the localized shear stress. The distribution of void sizes measured from SEM rnicrographs indicates that microcracks propagate through the ferrite grains by the nucleation and coalescence of voids along well-developed dislocation cell walls. This investigation has shown the individual nature of the plastic deformation and fracture processes in the ferrite and pearlite constituents of a two-phase steel. In addition, the necessity for the combined use of TEM and SEM techniques for understanding the role of plastic deformation in the initiation of ductile fracture has been demonstrated. Acknowledgments The authors acknowledge the support of this investigation by the Natural Sciences and Engineering Research Council through a scholarship (JKC) and a grant (MNB). References 1 D. Brock, Eng. Fract. Mech., 5 (1973) 55. 2 S.H. Goods and L. M. Brown, Acta Metall., 27 (1979) 1. 3 A. S. Argon, J. Im and R. Safoglu, Metall. Trans. A, 6 (1975) 825. 4 J. R. Fisher and J. Gurland, Met. Sci., 15 (1981) 193. 5 J. Gurland and J. Plateau, Trans. Am. Soc, Met., 56 (1963) 351. 6 H. G. F. Wilsdorf, Mater. Sci. Eng., 59 (1983) 1. 7 J. N. Greenwood, D. R. Miller and J. W. Suiter, Acta Metall., 2 (1954) 250. 8 R.W. Bauer and H. G. F. Wilsdorf, Scr. Metall., 7(1973) 1213.
48 9 R. N. Gardner, T. C. Pollock and H. G. E Wilsdorf, Mater Sci. Eng., 29 (1977) 1969. 10 R.N. Gardner and H. G. F. Wilsdorf, Metall. Trans. A, 11 (1980) 653. 11 T. C. Pollock and H. G. F. Wilsdorf, Mater. Sci. Eng., 61 (1983) 7. 12 M. N. Bassim, R. J. Klassen, M. R. Bayoumi and H. G. F. Wilsdorf, Mater. Sci. Eng., 92 (1987) 107.
13 L. Yumen and Z. Jingen, Mater. Sci. Eng., 84 (1986) 137. 14 D. Kuhlmann-Wilsdorf, in J. E Hirth and J. Weertman (eds.), Work Hardening, Gordon and Breach, New York, 1968,p. 97. 15 D. Kuhlmann-Wilsdorf, Mater. Sci. Eng., 86 (1987) 55. 16 C.D. Beachem, Metall. Trans. A, 6 (1975) 377. 17 L. E. Miller and G. C. Smith, J. Iron Steel Inst., 208 (1970) 998.