Russian Physics Journal, Vol. 45, No. 3, 2002
ELECTRON DIFFRACTION ANALYSIS OF THE FRACTURE ZONE OF Fe–0.45C–17Mn–3Al STEEL SUBJECTED TO MULTICYCLIC FATIGUE TESTS N. A. Popova, O. V. Sosnin, L. N. Ignatenko, S. V. Konovalov, V. V. Kovalenko, Yu. F. Ivanov, V. E. Gromov, and N. A. Koneva
UDC 669.112.227.342 : 539.4.015
The methods of transmission electron microscopy are used to investigate the structure and phase content of Fe–0.45C–17Mn–3Al steel subjected to multicyclic fatigue tests to failure after intermediate electrostimulation. It is demonstrated that electrostimulation does not change the mechanism of steel failure. An increase in the operating life time of samples after electrostimulation in the intermediate test stage is connected with the hindered γ → ε martensitic transformation, the relaxation of stress concentrators, and the reduction of the volume fraction of the critical dislocation (network) substructure in which ε-martensite is mainly developed. INTRODUCTION
Many parts of machines, mechanisms, and structures operate under repeated and variable loading and frequently fail after the repeated action of a force with a small amplitude. Despite long-term history of investigations, the problem of fatigue failure of steels and alloys is important now [1–3]. From the modern viewpoint, the failure is the final stage of substructure evolution when the accommodation potentialities of the material have already been exhausted and the critical substructure has been formed [4–6]. For this reason, the problem of diagnostics of the critical deformation stage of the material subjected to fatigue tests seems to be very important. The ultrasonic method of diagnostics based on the dependence of the ultrasonic speed υ on the number of cycles of loading of the examined material N was suggested and justified in [7–10]. Sharp bending of the curve υ(N) indicates the formation of the critical substructure whose further development leads to the failure of the sample. The detection of the moment the critical substructure is formed is the necessary but insufficient condition of longer operating life time of parts of machines, mechanisms, and structures. No less important is the problem of failure origin in the critical substructure. Electrostimulation is one of the efficient methods of failure suppression to increase the operating life time of structural steels by 20–30% [11–13]. The process of electrostimulation includes transmission of high-current pulses with optimal frequency, amplitude, and time of action through the material in the critical stage of changing the dislocation structure and υ(N) [11]. It is obvious that knowledge of mechanisms of current pulse action on the dislocation substructure and phase content of materials [11, 12, 14–18] is required for purposeful application of the given method. In the present paper, this problem is solved for Fe–0.45C–17Mn–3Al steel based on analysis of evolution of the defect substructure and phase content of structural austenitic steel subjected to multicycle fatigue tests to failure. 1. MATERIAL AND EXPERIMENTAL PROCEDURE
We investigated Fe–0.45C–17Mn–3Al austenitic steel after hot rolling [19]. Samples for fatigue tests were cut with the long axis parallel to the direction of rolling. The samples whose shape and sizes were discussed in [11] were subjected to fatigue tests by cyclic symmetric cantilever bending [20] on a special machine at room temperature. The frequency of sample loading was 18 Hz. The number of cycles before failure of samples without electrostimulation was N1 = 10.2⋅104.
Tomsk State University of Architecture and Building; Siberian State Industrial University. Translated from Izvestiya Vysshikh Uchebnykh Zavedenii, Fizika, No. 3, pp. 100–108, March, 2001. 1064-8887/02/4503-0319$27.00 2002 Plenum Publishing Corporation
319
a
b
0.2 µm
c
0.2 µm
d
0.5 µm
e
0.5 µm
f Ⱥ
0.5 µm
0.5 µm
Fig. 1. Types of dislocation substructures in Fe–0.45C–17Mn–3Al steel in the initial state: dislocation chaos (a), network nondisoriented (b) and disoriented (with broken ends) substructures (c), and fragmented (d) and recrystallized grain substructures (e and f) with a recrystallized grain indicated by A. When the number of cycles of loading reached N2 = 7⋅104, mechanical tests of some samples were interrupted. Current pulses (f = 19 Hz) were transmitted through the chosen samples for τ = 45 s. The electrostimulated samples were then subjected to fatigue tests by the above-indicated scheme. For these samples, the number of cycles before failure was N3 = 17.9⋅104. The samples were heated up to ~300°C during electrostimulation. The temperature of the samples increased only insignificantly during tests. The examined samples were treated in reference reagents by the methods of chemical thinning and electropolishing [21]. The dislocation substructure of steel was analyzed by the methods of transmission electron microscopy of thin foils using an ÉM-125 transmission electron microscope. The average scalar dislocation density [22], the scalar density of the dislocation substructure of the given type, the volume fraction of dislocation substructures of different types [23], and the curvature-torsion amplitude of the crystal lattice were determined [24]. The structure of steel was analyzed for the cross sections perpendicular to the long axis of the samples located at distances of 500, 800, and 2300 µm from the fracture surface (corresponding to the zones of maximum loading for the intermediate sample) and also in the bulk of the material adjacent to the fracture surface.
2. RESULTS OF INVESTIGATIONS AND THEIR DISCUSSION 2.1. Structure and phase content of initial steel
Examined steel in the initial state is a polycrystalline aggregate with the average transverse and longitudinal grain sizes D = 10.3 µm and L = 20.3 µm formed by the FCC iron-based solid solution. After rolling, grains were anisotropic. 320
a
b
0.5 µm
0.5 µm
c
d
e
g
h
0.5 µm
f
0.5 µm
Fig. 2. Initial state of Fe–0.45C–17Mn–3Al steel: bending extinction contours in the fragmented dislocation substructure (a and b) and cracks along the γ → ε1 interphase boundaries (c and d), precipitation of aluminum particles on dislocations in the γ-phase (e–f), dark-field image recorded in the ª¬ 201º¼ + ª¬ 220 º¼ reflections (g), and the diffraction pattern of region f (h). Arrows indicate the ª¬ 201º¼ γ
Al
γ
and ª¬ 220 º¼ reflections. Here d shows the electron diffraction pattern of region c. The arrow indicates the Al ª¬ 101º¼ reflection in which dark-field image d was recorded. ε
They were elongated along the rolling axis, with the scattering angle of the structural texture vector being equal to ~8°. Recall that the structural texture vector is a vector whose direction coincides with the long grain axis and the modulus is equal to the anisotropy factor [25]. The intragranular structure of steel in the initial state has dislocation substructures of different types classified below according to [23]. The dislocation chaos substructure (Fig. 1a), the nondisoriented (Fig. 1b) and disoriented network substructures with broken, geometrically necessary [26], and low-angle boundaries (Fig. 1c) (as a rule, this substructure type arises near the large-angle grain boundaries), and the fragmented dislocation substructure (Fig. 1d) were identified by the methods of transmission electron microscopy. Volume fractions of the above-indicated dislocation substructures were in the ratios 1 : 2 : 7 (nondisoriented and disoriented dislocation substructures were grouped into the common volume fraction). In addition, grains of micron sizes with a small number of dislocations were observed in steel (such a grain is indicated by A in Fig. 1). These grains are produced in the course of dynamic recrystallization. Their volume fraction in the grain ensemble of steel was ~0.25–0.30. One of the reasons for a wide variety of dislocation substructures formed in steel is its thermomechanical treatment during high-temperature rolling. Different recovery processes and dynamic recrystallization that proceeded in the material during hot rolling [27] yielded the standard set of substructures [28]. The grains of examined steel, irrespective of the types of dislocation substructures, have a large number of bending extinction contours. This is indicative of elasto-plastic bending of the material (Fig. 2a and b). In this case, the curvature321
torsion of the grain as a whole was observed in most cases, that is, the reason for the creation of internal long-range stress fields was to a greater degree the incompatibility of deformations of adjacent grains and their groups rather than the intragranular dislocation substructure. Such grains are driving forces for the dynamic recrystallization process. A particular case of the instantaheous pattern of dynamic recrystallization is shown in Fig. 1f. Fragmentation in the front of the migrating grain boundary with subsequent growth of low-angle shoulders is one of the mechanisms of grain growth during dynamic recrystallization [29]. In some cases, regions of the material containing ε-martensitic crystals are observed in steel (Fig. 2c–e). As a rule, these regions are located along the grain boundaries. This allows us to suggest several reasons for the γ → ε martensitic transformation. In addition to the well-known fact that the ε-phase arises in noticeable quantities in Fe–Mn alloys at temperatures less than 200°C when the Mn concentration is between 14 and 25 wt.% [30], there are two more reasons for its appearance. The first reason is the change in the elemental composition of steel when alloying elements go to the grain boundaries. The second reason is elastic stresses arising in regions of adjacent grains close to their boundaries due to their incompatible deformations. Practically always regions of the material containing ε-martensitic crystals were fractured by microcracking along the γ–ε interphase boundary in the process of preparing thin foils (Fig. 2c–e). This is unambiguously indicative of large elastic stresses acting here. As demonstrated in [31-34], ε-martensite is the undesirable phase from the viewpoint of plasticity, because it decreases significantly the operating life time of steel under few-cycle fatigue tests. One more phase component of steel in the initial state is precipitation of aluminum. Nanometer aluminum particles, as a rule, are localized on dislocations decorating them. A typical electron micrograph of aluminum precipitation is shown in Fig. 2f–h. Quantitative analysis of dislocation substructure of steel in the initial state demonstrated that the maximum density of bulk dislocations was observed in the network substructure, and the maximum curvature-torsion amplitude of the crystal lattice was observed in the fragmented substructure. Moreover, the elastic component of the curvature-torsion amplitude of the material is also maximum in the fragmented substructure. Therefore, the fragmented dislocation substructure of examined steel is the most stressed structural component of the initial material. A high level of elastic stress fields recorded in the fragmented substructure correlates with a maximum dislocation density. Regions with fragmented substructure are the most hardened ones. Analogous regions, as a rule, are places of forming dynamic recrystallization centers [27]. It is well known that at moderate temperatures the sequence of dislocation substructures in FCC steels is the following: chaos → network substructure → fragmented substructure [28, 35]. Usually microcracks arise on the subboundaries and the failure develops in the course of deformation of the fragmented substructure. At an elevated temperature, the following sequence of dislocation substructures takes place in steels of the given class: network – fragmented – subgrain dynamic recrystallization structure. Under subsequent deformation, grains grow up and the same cycle of changing the dislocation substructure is repeated in them. Substructures recorded in the study of the initial state arose naturally during high-temperature rolling.
2.2. Structure of the failure zone formed under cyclic tests
The number of cycles of loading before failure of steel under cyclic tests without intermediate action is N1 = 10.2⋅104. Investigations of the fracture surface of steel by the methods of transmission electron microscopy revealed all three above-indicated dislocation substructure types, namely, the dislocation chaos and network and fragmented substructures (Fig. 3). Volume fractions of substructure types were as follows: the dislocation chaos occupied ∼0.18 of the sample volume, the network substructure occupied ∼0.42 of the sample volume, and the fragmented substructure occupied ∼0.40 of the sample volume. The average grain had the width D = 7.4 µm and the length L = 13.8 µm. A comparison of these data with the results obtained in the study of the structure of the initial steel state shows that the evolution of the dislocation substructure of steel before failure proceeded as follows. The dislocation chaos structure formed as a result of thermomechanical action during rolling (in the initial state) disappeared completely in the process of cycling in a certain stage. However, a significant volume fraction of the material (Pv = 0.18) is occupied by the dislocation chaos structure at the instant of the fatigue failure of the sample. Obviously, this is the dislocation substructure formed under cyclic loading of steel. This substructure can be formed in dynamically recrystallized grains that were present in steel in the initial state and, as demonstrated below, arose in the process of cycling. The presence of the dynamic recrystallization 322
b
a
0.5 µm
0.5 µm
c
d
0.5 µm
0.5 µm
Fig. 3. Types of dislocation substructures formed after cyclic tests. The fracture surface has the dislocation chaos (a) and the network (b) and fragmented dislocation substructures (c). structure is confirmed by the metallographic experiments. After N = 7⋅104 loading cycles, a large number of grains did not have dislocation dimples after etching [36]. About 30% of grains have the dislocation structure, but only several percent bear evidence of martensitic transformations and twinning. Under subsequent loading (105 cycles), the martensitic structure occupied a greater volume fraction. The network dislocation substructure formed in steel during preliminary thermomechanical treatment by rolling developed stably during cycling. In the intermediate stage of loading it occupied 0.5 of the entire volume of the foil cut from the region with the maximum action force amplitude. This increase in the volume fraction of the material with the network substructure is connected first, with the corresponding transformation of the dislocation chaos structure. Second, it is connected with the development of the dislocation substructure in grains that already existed in the initial material and were virtually free from dislocations (initial dynamically recrystallized grains) in the course of cycling. A certain decrease in the volume fraction of the material occupied by the network dislocation substructure was found in the study of the fracture surface of steel. Obviously, this was caused by the reorganization of this substructure into the fragmented dislocation one. Therefore, cycling of Fe–0.45C–17Mn–3Al steel proceeded by the following scheme of transformation of the dislocation substructure:
Dislocation chaos network substructure ( ε-martensite) fragmented substructure ( ε-martensite). The fragmented dislocation substructure, being dominant after thermomechanical treatment of steel by rolling, gradually loses its position in the process of cycling. As indicated above, the fragmented dislocation substructure occupied ∼0.7 of the total volume of the material in the initial state. After intermediate cycling (N2 = 7⋅104), it occupied 0.5 of the volume of the material, and its volume fraction in the fractured material (N1 = 10.2⋅104) was equal to 0.4. In a number of papers (for example, see [4–6]) it was demonstrated that the fragmented substructure is the final stage in the development of the dislocation substructure before failure of the material under cold mechanical tests of different forms. For deformation on heating, the fragmented substructure is transformed into the subgrain substructure with the subsequent dynamic recrystallization process [27]. Naturally, this process is accompanied by a decrease in the volume fraction of the material occupied by the fragmented substructure. Therefore, cycling of Fe–0.45C–17Mn–3Al steel is accompanied by dynamic recrystallization. We note that these results are confirmed by the study of the grain structure of steel conducted by the methods of metallography of etched microsections [36]. The local character of dynamic recrystallization was confirmed by mesuring the average grain size after N1 = 10.2⋅104 cycles (D = 7.35 µm and L = 13.8 µm). The nonmonotonic character of 323
a
b
c
e
f
0.5 µm
d
0.5 µm
Fig. 4. Electron micrographs showing the fracture surface after cycling tests: deformation microtwins (a–c), ε-martensite (d and e), dark-field image recorded in the [002] reflection of twins (b), and darkfield image recorded in the ª¬122 º¼ reflection (e). Arrows indicate the [002] reflection of twin b and ε
the ª¬122 º¼ reflection of twin f. ε changes in the volume fraction of the dislocation substructures and other parameters is indicative of the local dynamic recrystallization that takes place during the substructural transformation. Analogous facts were observed by us in [25] when we investigated ferrite-pearlite steel. The amplitudes of stress fields formed in the fracture surface of steel averaged over the entire volume of the material and for each substructure exceed significantly (by factors of 1.5–3) the given characteristics of steel in the initial state. Moreover, in the intermediate stage of loading (when N2 = ∼0.7 Nfail), changes in the curvature-torsion amplitude of the crystal lattice of the material were insignificant. As already indicated above, this character of changes in the given parameter of the dislocation substructure is connected with the specific features of dislocation substructure reorganization. Obviously, this is indicative of the critical instant in the substructural evolution after which the material has irreversibly failed. Exactly for this reason, the value N2 = 7⋅104 was chosen for subsequent electrostimulation to increase the plasticity of the material. Cyclic deformation of Fe–0.45C–17Mn–3Al steel was accompanied by forming microtwins in the fracture surface (Fig. 4a–c). The microtwins were formed predominantly in the network dislocation substructure, that is, in the substructure with the highest scalar dislocation density and curvature-torsion amplitude of the crystal lattice. On the other hand, in the intermediate cycling stage (for N2 = 7⋅104), the strain-induced twins were not observed, though the scalar dislocation density in steel in this state was higher than that at the instant of failure. Therefore, one of the reasons for twinning in examined steel under cyclic loading is the presence of strong long-range stress fields that reach their maximum values exactly in the network dislocation substructure. As indicated above, a small number of ε-martensitic crystals located along the grain boundaries is observed in steel already in the initial state. In the intermediate stage of cycling (for N2 = 7⋅104), their number remains virtually unchanged, but the volume fraction of ε-martensitic crystals in the fracture surface increases significantly (2–3 times). Crystals of εmartensite in the fracture surface are mostly formed in the network dislocation substructure (Fig. 4d–f). They are much rarely observed in the chaotic substructure. No ε-martensite was detected in fragments. Such behavior of the material indicates the critical character of the substructure state in the vicinity of N2 = 7⋅104. Crystals of the martensite phase cause microcracking of the material when thin foils are fabricated. Microcracks in the initial state, as indicated above, are formed along the interphase boundaries that separate γ- and ε-phases. The microcrack density in the foil prepared from the bulk of the material adjacent to the fracture surface noticeably (by a factor 324
a
b ɋ B 0.5 µm
0.5 µm
d
c
0.5 µm
e
0.5 µm
g
f
0.5 µm
Fig. 5. Electron micrograph of Fe–0.45C–17Mn–3Al electrostimulated loaded steel: the structure of recrystallized grains are indicated by B and C (a and b), the region with long dislocations (c), the fracture surface (d–g), two intersected systems of ε-martensitic plates (d–f) along which microcracks are formed, the dark-field image in the ª¬1 11º¼ reflection (e) indicated by the arrow in diffraction ε
pattern e, and nucleation centers of microtwins in the network structure (g). of 8.5) exceed that in the initial state. These microcracks, as in the initial state, are formed along the boundaries separating γ- and ε-phases. Thus, our investigations of Fe–0.45C–17Mn–3Al steel subjected to cyclic fatigue tests performed by the methods of transmission electron microscopy demonstrated that one of the fracture mechanisms is the γ → ε transformation with subsequent formation of the nonequilibrium interphase boundaries separating microcracks that transform into straight cracks. In their turn, long-range stress fields localized in the network dislocation substructure cause the γ → ε martensitic transformation. Sources of long-range stress fields are primarily the incompatibility of deformations of adjacent grains and their groups and also γ- and ε-phases. 2.3. Structural and phase transformations after electrostimulation of steel subjected to fatigue tests
One of the most efficient methods of increasing the cyclic fatigue limit of steel is the transmission of electric current pulses (electrostimulation) through the material in the intermediate test stage [11]. Electrostimulation engenders several types of relaxation processes in the material. The first process is the reorganization of the grain structure due to the nucleation and growth of grains caused by the development of the local dynamic recrystallization. As a rule, new grains are formed near the boundaries of already existing grains or in their joints (Fig. 5a and b) [25, 27, 37]. Two processes proceed simultaneously in the material, namely, the fragmentation and recrystallization. Figure 5b shows a growing grain (denoted by C) which substitutes a fragmented substructure. Recrystallization removes internal stresses that give the main 325
contribution to the driving force of this process. One more contribution to the recrystallization process comes from the formation of the low-energy boundaries (regions B and C in Fig. 5a and b). The recrystallization is caused by a local decrease in the energy in the regions of boundary migration between dislocationless and cold-hardened grains. This decrease in the energy is caused by the following reasons: 1) disappearance of dislocation substructure and sharp decrease in the dislocation density, 2) changes in the type and defect structure of the lower-energy grain boundaries, 3) decrease in the amplitude of local and long-range stress fields, and 4) reorientation of anisotropic elastic grains in the direction of less Young’s modulus. The last three factors form a new texture. All these reasons are connected with micro- and mesolevels. One more reason – stress concentrators that increase the elastic energy – arises at the mesolevel. They are dislocation pileups, joint disclinations, steps on the grain boundaries, etc. Closing of microconcentrators is also the moving force of recrystallization. The second process is the annihilation of dislocations causing fixation of the chaotic dislocation substructure in the electrostimulated material that was absent in steel subjected to cycling before electrostimulation. In some cases dislocation chaos was formed by long parallel dislocation segments (Fig. 5c). We note that the scalar dislocation density in the chaotic dislocation substructure formed after electrostimulation is much less than in the structure arising during hot rolling of steel (6.5⋅109 cm–2 and 3.8⋅109 cm–2, respectively). The third process – the partial reorganization of the dislocation substructure – includes a decrease in the volume fraction of the network substructure (down to ∼0.4 of the steel volume) and an increase in volume fractions of chaotic (∼0.05) and fragmented substructures (∼0.55). We note that in both cases, the reorganization of the dislocation substructure is accompanied by a decrease in the internal energy accumulated by the material during deformation. The fourth process is the formation of a large number of microtwins. As a rule, microtwins are observed in the network dislocation substructure. It is most likely that their formation is caused by thermal stresses arising in steel after electrostimulation, because twins often nucleate in regions of stress concentration. Quantitative analysis of the dislocation substructure of electrostimulated steel performed for the region of the material under maximum loading showed that under the action of electric current first, the scalar dislocation density decreases. Second, the curvature-torsion amplitude of the crystal lattice decreases significantly (by a factor of 1.8). Third, the grain structure changes: the structural texture vector increases significantly against the background of slow grain growth. This naturally increases the plasticity of the material decreasing its anisotropy. Fourth, the volume fraction of the network substructure decreases. This change occurs in two directions – volume fractions of chaotic and fragmented substructures increase. These processes are local in character. It is most likely that they proceed in regions of relaxation of stress concentrators or their complete dissipation. They clearly indicate physical reasons for the plasticization of the material after electric stimulation.
2.4. Phase content and defect substructure of the fracture surface formed after cyclic tests of steel electrostimulated in the intermediate stage
We have already demonstrated above that if we interrupt the mechanical tests of steel after some (in our experiments after N2 = 7⋅104) cycles of loading and conduct electrostimulation, the fatigue operational life time of the material increases significantly (by a factor of ∼1.65), and the number of cycles before failure increases by a factor of 1.8 (compare N2 = 10.2⋅104 with N3 = 17.9⋅104 cycles of loading). In this section, we present the results of investigations of the dislocation substructure and phase state of the fracture surface of Fe–0.45C–17Mn–3Al steel formed in the sample electrostimulated in the intermediate stage of cyclic tests. As in the case of steel failure after continuous cyclic tests, in the examined case three dislocation substructures were formed in the fracture surface of steel, namely, chaotic, network, and fragmented ones. The main type of the dislocation substructure in the fracture surface of the electrostimulated sample was network substructure whose volume fraction increased fast with the number of cycles of loading. Volume fractions of the two remaining substructure types (the dislocation chaos and fragments) monotonically decreased despite a certain increase in their volume fractions after electrostimulation (Table 1). The fact that the network dislocation substructure, as a rule, precedes the failure has been mentioned above. Exactly in this dislocation substructure ε-martensite was observed with subsequent origin of microcracks on interphase and intraphase boundaries. This is indicative of the main role of the network substructure in failure of ductile materials [6]. 326
TABLE 1. Characteristics of the Structure and Phase Content of Fe–0.45C–17Mn–3Al Steel Subjected to Multicyclic Fatigue Tests to Failure after Electrostimulation (ES) State of the sample Without ES With ES
Chaos 18 3
Pv, % Networks 42 77
–10
Fragments 40 20
ρ⋅10 , –2 cm 0.8 1.72
–1
χ, cm 915 953
–4
ρε⋅10 , –1 cm 1.16 0.21
–4
ρtw⋅10 , –1 cm 0.17 0.43
–4
ρcr⋅10 , –1 cm 0.17 0.05
Analysis of the results presented in Table 1 demonstrates that irrespective of the dislocation substructure type, the curvature-torsion amplitude increases with the number of cycles of loading and reaches its maximum when the material fails. Intermediate electrostimulation of steel also leads to a noticeable decrease in this parameter of the steel structure. Comparison of the curvature-torsion amplitude of the crystal lattice and the density of bending contours after failure of samples without and with electrostimulation demonstrates that the relaxation of the most powerful stress concentrators is observed after electrostimulation. As a result, the density of concentrators remains virtually unchanged and may even increase (at the expense of small concentrators), but the average curvature-torsion amplitude decreases. Therefore, processes of nucleation and especially growth of cracks are shifted toward larger number of cycles of loading. As indicated above, electrostimulation of steel causes the γ → ε martensitic transformation. As a rule, one system of martensitic crystals is formed in individual regions. Further loading triggers related systems of the γ → ε martensitic transformations in volumes of steel already containing martensitic crystals rather than produces new volumes containing martensitic crystals. Undoubtedly, this process is relaxation and reduces the probability of failure. Microcracks nucleate on the boundaries of ε-martensitic crystals. They nucleate in the region of intersection of different crystallographic systems of deformation transformation. One more specific reason for plastificization of the material after electrostimulation was detected. Comparison of dislocation substructures of two samples without and with electrostimulation after failure indicates changes in the electron structure of the solid solution and its concentration. Along with microtwins, their nucleation dislocations, namely, rectilinear dislocations are clearly seen (Fig. 5g). Electrostimulation decreases the content of the ε-phase and hence increases the work of microcrack formation, because cracks are located predominantly on the γ-ε interphase boundaries. Along with dislocation shear, a decreased contribution of the γ → ε transformation to the deformation is compensated by twinning. Indeed, the scalar dislocation density and the density of microtwins after electrostimulation and subsequent failure reach their maxima (see Table 1). Changes in the electronic structure of steel are caused by changes in the solid solution concentration. This is connected with the fracture of aluminum microparticles by moving dislocations, twins, and ε-martensitic plates. Along with stress fields, the electronic structure increases the plasticity of the material. Thus, the concentration of tricharged aluminum ions in solid solution increases after electrostimulation. This leads first, to the anomalous mass transfer of carbon and aluminum ions that interact actively with dislocations causing their motion. Second, this leads to the formation of carbon and aluminum segregations on dislocations. This intensifies the motion of dislocations under the action of electric field due to the presence of ion charges. Third, nonuniform local heating must take place during electrostimulation due to inhomogeneous electrical resistivity of solid solution caused by inhomogeneous atomic and defect substructures. All this increases the stability of steel and the number of cycles before failure by fatigue. REFERENCES
1. 2. 3. 4. 5.
V. S. Ivanova and V. F. Terent’ev, Nature of Metal Fatigue [in Russian], Metallurgiya, Moscow (1975). S. A. Golovin and A. A. Pushkar, Microplasticity and Fatigue of Metals [in Russian], Metallurgiya, Moscow (1980). V. T. Troshchenko, Deformation and Fracture of Metals after Multicyclic Loading [in Russian], Naukova Dumka, Kiev (1981). V. A. Likhachev, Models in Mechanics of Continuous Media [in Russian], Publishing House of the Siberian Branch of the USSR Academy of Sciences, Novosibirsk (1983). V. V. Rybin, Large Plastic Deformations and Fracture of Metals [in Russian], Metallurgiya, Moscow (1986). 327
6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.
328
N. A. Koneva, L. I. Trishkina, and É. V. Kozlov, in: Modern Problems in the Physics and Mechanics of Materials [in Russian], NII NM SPGU, Saint Petersburg (1997), pp. 322–332. V. V. Murav’ev, L. B. Zuev, and K. L. Komarov, Speed of Sound and the Structure of Steels and Alloys [in Russian], Nauka, Novosibirsk (1996). L. B. Zuev, O. V. Sosnin, D. Z. Chirakadze, et al., Prikl. Mekh. Teor. Fiz., 39, No. 4, 180–184 (1998). L. B. Zuev, O. V. Sosnin, V. E. Gromov, et al., Metallofiz. Noveish. Tekhnol., 19, No. 8, 80–82 (1997). V. E. Gromov, E. V. Semakin, V. Ya. Tsellermaer, et al., Izv. Ross. Akad. Nauk, Ser. Fizich., No. 5, 1019–1023 (1997). O. V. Sosnin, V. E. Gromov, É. V. Kozlov, et al., Electrostimulated Fatigue after Few Cycles [in Russian], Nedra Communications Ltd., Moscow (2000). V. E. Gromov, L. B. Zuev, É. V. Kozlov, et al., Electrostimulated Plasticity of Metals and Alloys [in Russian], Nedra, Moscow (1996). A. F. Sprecher, S. L. Mannan, and H. Conrad, Acta Metall., 3, No. 7, 1145–1162 (1986). V. E. Gromov, Yu. F. Ivanov, G. V. Pushkareva, et al., Izv. Vyssh. Uchebn. Zaved., Fiz., No. 12, 31–35 (1990). V. E. Gromov, Yu. F. Ivanov, S. A. Bashkirova, et al., Izv. Vyssh. Uchebn. Zaved., Fiz., No. 9, 84–90 (1991). V. E. Gromov, É. V. Kozlov, V. E. Panin, et al., Metallofizika, 13, No. 11, 9–13 (1991). V. E. Gromov, Yu. F. Ivanov, V. Ya. Tsellermaer, et al., Metallofizika, 13, No. 4, 100–106 (1991). V. E. Gromov, É. V. Kozlov, V. Ya. Tsellermaer, et al., Probl. Mashinostr. Nadezhn. Mashin, No. 2, 46–51 (1994). V. G. Sorokin, ed., Specifications of Steels and Alloys [in Russian], Mashinostroenie, Moscow (1989). N. A. Makhutov and A. N. Romanov, eds., Mechanics of Fracture after Few Cycles of Loading [in Russian], Nauka, Moscow (1986). A. M. Glauert, ed., Practical Methods in Electron Microscopy [in Russian], Mashinostroenie, Moscow (1980). P. B. Hirsch, A. Howie, R. B. Nicholson, et al., Electron Microscopy of Thin Crystals [Russian translation], Mir, Moscow (1968). N. A. Koneva, É. V. Kozlov, and L. I. Trishkina, Metallofizika, No. 10, 49–58 (1991). N. A. Koneva, É. V. Kozlov, L. I. Trishkina, et al., in: Proc. Int. Conf. on New Methods in the Physics and Mechanics of a Deformable Solid [in Russian], Tomsk (1990), pp. 83–93. V. E. Gromov, É. V. Kozlov, V. I. Bazaikin, et al., Physics and Mechanics of Drawing and Bulk Extrusion [in Russian], Nedra, Moscow (1997). D. A. Hughes and N. Hansen, Metall. Trans., 24A, No. 9, 2021–2037 (1993). M. L. Bernshtein, V. A. Zaimovskii, and L. M. Kapustina, Thermomechanical Treatment of Steel [in Russian], Metallurgiya, Moscow (1983). N. A. Koneva and É. V. Kozlov, in: Structural Levels of Plastic Deformation and Fracture [in Russian], Nauka, Novosibirsk (1990), pp. 123–186. S. P. Efimenko, É. V. Kozlov, L. A. Teplyakova, et al., Metally, No. 5, 30–36 (1995). I. N. Bogachev and V. F. Egolaev, Structure and Properties of Iron and Manganese Alloys [in Russian], Metallurgiya, Moscow (1973). Yu. F. Ivanov, D. V. Lychagin, V. E. Gromov, et al., Fizich. Mezomekh., No. 3, 103–108 (2000). V. E. Gromov, O. V. Sosnin, V. V. Kovalenko, et al., in: New Industrial Technologies and Materials [in Russian], Sibirskie Ogni, Novosibirsk (2000), pp. 153–161. V. V. Kovalenko, O. V. Sosnin, Yu. F. Ivanov, et al., Fiz. Khim. Obrab. Mater., No. 6, 74–80 (2000). V. V. Kovalenko, Yu. F. Ivanov, O. V. Sosnin, et al., Izv. Vyssh. Uchebn. Zaved., Chern. Metall., No. 12, 57–59 (2000). N. A. Koneva and É. V. Kozlov, Izv. Vyssh. Uchebn. Zaved., Fiz., No. 2, 89–106 (1990). O. V. Sosnin, S. V. Konovalov, Yu. F. Ivanov, et al., Izv. Vyssh. Uchebn. Zaved., Chern. Metall., No. 6, 56–59 (2001). S. S. Gorelik, Recrystallization of Metals and Alloys [in Russian], Metallurgiya, Moscow (1978).