ISSN 00360295, Russian Metallurgy (Metally), Vol. 2013, No. 4, pp. 313–320. © Pleiades Publishing, Ltd., 2013. Original Russian Text © V.G. Khanzhin, S.A. Nikulin, V.A. Belov, V.Yu. Turilina, A.B. Rozhnov, 2012, published in Deformatsiya i Razrushenie Materialov, 2012, No. 1, pp. 40–47.
DIAGNOSTICS AND MECHANICAL TEST TECHNIQUES
Hydrogen Embrittlement of Steels: II. Effect of Strength V. G. Khanzhin, S. A. Nikulin*, V. A. Belov, V. Yu. Turilina, and A. B. Rozhnov National University of Science and Technology MISiS, Leninskii pr. 4, Moscow, 119049 Russia *email:
[email protected] Received September 26, 2011
Abstract—The hydrogen embrittlement resistances of 35KhGM steel in states of different strengths are com pared. The effect of strength on the fracture mechanism and kinetics in the steel during hydrogen embrittle ment is demonstrated using a joint analysis of acoustic emission signals and stress–strain diagram during ten sion, metallography, and fractography. DOI: 10.1134/S0036029513040058
1. INTRODUCTION The sensitivity of steels to hydrogen embrittlement depends on many, often interrelated factors (chemical composition, inclusion segregation, structure, strength, internal stresses) that determine the ability of a material to absorb hydrogen and on the mechanism and kinetics of its delayed fracture [1–7]. To control the resistance of a metal to hydrogen embrittlement efficiently, one needs data on the effect of these factors on the fracture mechanisms and kinetics. Water electrolysis in the medium of dissolved indus trial gases can be a source of atomic hydrogen in a struc tural steel during its operation. If dissolved atomic hydrogen cannot leave a metal, it can accumulate in structural traps in the molecular (gaseous) form, creat ing high mechanical stresses in the volume surrounding the traps due to its pressure. The formation, growth, and coalescence of hydrogencontaining pores represent the characteristic hydrogen embrittlement mechanism in low and mediumstrength steels with a low plastic deformation resistance. The traps of atomic hydrogen in steel can be inclusions, secondphase particles, and the regions of high tensile stresses at interphase bound aries, where a high hydrogen concentration is ensured by uphill diffusion. Hydrogencontaining voids in highstrength steels with a high strain resistance often initiate sharp micro cracks, which results in intra or intergranular brittle fracture. Hydrogeninduced grainboundary fracture can be enhanced in the presence of fineparticle segre gation along grain boundaries [6, 7]. The purpose of this work is to study the fracture kinet ics and mechanisms in structural steel in states having different strengths during hydrogen embrittlement.
0.90 Mn, 0.85–1.25 Cr, 0.15–0.35 Mo, <0.030 P, and <0.030 S. Cylindrical samples 5 mm in diameter with a gage length of 100 mm were prepared from forged rods. Two states of the steel with different strength lev els (states 1 and 2) were achieved by tempering at dif ferent temperatures of samples preliminarily quenched from 850°C in oil: state 1, upon tempering at 550°C for 1 h; state 2, tempering at 400°C for 1 h. In both cases, the samples were cooled in oil. The mechanical properties of the samples are given in the table. The steel structure in state 1 corresponded to tem pered troostite and the structure in state 2 corre sponded to troostite–martensite. In state 1, coarse fer rite plates repeated the orientation of the former mar tensite crystals; in state 2, the steel had a finer structure. The austenite grain size was 12–14 μm in both cases (Fig. 1). Electronmicroscopic investiga tion (JEM 2000 FXII microscope) showed that the difference in the strengths and ductilities in states 1 and 2 was mainly determined by the strength of the matrix rather than the precipitation of carbide and carbonitride particles: the densities of their distribu tion in the steels of states 1 and 2 were almost the same, (2.8 ± 1.1) × 108 and (2.7 ± 0.5) × 108 mm–2, respec tively. Steel samples in states 1 and 2 were hydrogen satu rated according to the technique developed in [8]: it includes loading by tension to a given load (σ = Mechanical properties of 35KhGM steel during tension
2. EXPERIMENTAL We studied structural heat treatable steel 35KhGM containing (wt %) 0.32–0.39 C, 0.15–0.35 Si, 0.55– 313
State
HV
σ0.2, MPa
σu, MPa
εu, %
ε, %
ψ, %
1 2
350 ± 1.5 500 ± 2.0
940 1360
1050 1460
8.3 4.2
11.5 7.8
63 54
Note: εu is the uniform strain, ε is the relative elongation, and ψ is the relative reduction of area.
314
KHANZHIN et al.
500 nm (b)
(а)
500 nm
Fig. 1. Microstructure of 35KhGM steel in states (a) 1 and (b) 2 (transmission electron microscopy).
0.8σ0.2) and local galvanic hydrogenation for 4–168 h at a given load (in an electrochemical cell). The hydro gencharged samples were unloaded to σ = 0.2σ0.2 and tensioned to failure at a rate of 6 × 10–4 s–1 and a tem perature of 20°C on an Instron LX150 tensiletesting machine, and acoustic emission (AE) signals were simultaneously recorded. It should be noted that uniaxial tension is the most rigid scheme among those ensuring uniform plastic deformation. We tested 45 samples in each state. The stress–strain diagram in the coordinates stress σ = P/F0 (P is the load, F0 is the initial crosssectional area of the sample) versus strain ε = ΔL/L0 up to the maximum load was converted into the true coordi nates stress S = σ(1 + ε) versus strain e = ln(1 + ε). Stress–strain diagrams S(e) were approximated by the Ludwig–Hollomon equation S = ken [5, 6], and strainhardening exponent n and the rootmean square deviation of its average value nav, 2
S nav = ± Σ { ( n i – n av ) /m ( m – 1 ) }, where m is the number of samples, were determined by the least squares method (using linear regression lnS = lnk + nlne). The strain in a neck under the fracture surface was determined from the grain elongation. 35–50 grains in each of the 10 passes along the horizontal and vertical were measured in each cross section in longitudinal polished sections (after etching in a saturated solution of picric acid in water) using an Axiovert 40 MAT opti cal microscope at a magnification of ×1000 in 10– 12 fields of view 0.8 × 0.8 mm in size. The grain elon gation along the sample axis was estimated as [9] 2/3
2/3
2/3
ε d = ( d 1 – d 2 )/d 1 , where d1 and d2 are the average chord lengths along and across a deformed sample.
The macrogeometry elements of fracture surfaces after failure (bottom crack diameter and crack open ing f were determined from the difference between the depths at the center of fracture and in the shear region; longitudinal crack “ridges” were determined) were measured along the vertical accurate to 2 μm on the dial of a PMT3M microscope upon focusing accurate to 10 μm along the horizontal using an UIM21 toolmaker’s microscope. Fracture surfaces were studied with a Hitachi S800 scanning electron microscope. The hydrogen absorbabilities of the states of steel were also compared. The hydrogen solubility in the metal was estimated from the gas volume released from a cylindrical sample 10 mm in diameter and 20 mm in height after hydrogenation in an electro chemical cell for 15 h (Fig. 2a). For this purpose, a thin cover glass was pressed against the polished end surface of a hydrogenated sample through a glycerol layer (Fig. 2b). The sample was placed on the micro scope stage of an Axioskop40 Pol (Carl Zeiss) optical microscope, and the realtime video filming of hydro gen release from the sample into glycerol was per formed with an AxioCamIC web camera at a magnifi cation of ×50 (Figs. 2b, 2c). “Flat” macrobubbles occupied the space between the sample surface and the cover glass (Fig. 2b). The released hydrogen vol ume was determined from the sum of the bubble vol umes V H2 = Σ(Sih), where Si is the bubble area and h ≈ 100 μm is the glycerol layer thickness, in an automatic manner when images were analyzed on a computer with the software package developed in [10]. Using the technique of estimating a local hydrogen distribution in a steel structure [11], we observed the distribution of hydrogen microbubbles in the steel at a magnification of ×(500–1000) at a bubble diameter resolution of 1– 2 μm (Fig. 2e).
RUSSIAN METALLURGY (METALLY)
Vol. 2013
No. 4
HYDROGEN EMBRITTLEMENT OF STEELS: II. EFFECT OF STRENGTH (a)
(b)
315
(c)
Microscope Cover glass Solution
h
“+” Sample
“–”
Glycerol
Sample (d)
(e)
10 μm
10 μm
Fig. 2. Cells for hydrogenation of samples and the observation of hydrogen release in glycerol: (a) scheme of the cell for the hydro genation of a cylindrical sample; (b), (c) scheme and appearance of the cell for observation with an optical microscope; and (d), (e) release of hydrogen macro and microbubbles from a hydrogenated steel sample into glycerol.
3. RESULTS AND DISCUSSION
formation of ductile internal cracks when the metal plasticity margin is exhausted [12].
3.1. HydrogenFree Fracture Mechanism and Kinetics The ductile fracture of steel during tension is known to be preceded by the following two unstable states: the onset of necking and the onset of unstable crack growth in the neck [12]. The times of these events indicate the values of uniform (εu) and concen trated (ψconc) deformation. For a material to undergo stable flow during deformation under operating true stress S and strain e, the hardening intensity should be dS/de > S. The loss of stable flow and deformation localization in the neck take place under the condition dS/de = S usually without a break in the material integrity due to “geometrical” softening as a result of sample diameter fluctuations; in this case, we have eu = n [5, 6]. In practice, however, the relation eu = n is usually not reached, and the loss of stable flow (necking) occurs earlier, at eu < n, e.g., because of the RUSSIAN METALLURGY (METALLY)
Vol. 2013
Figures 3a and 3b show the stress–strain diagrams of the steels samples in states 1 and 2 without hydrogen and AE diagrams synchronized with them using time marks [8]. The maximum load corresponds to the end of uniform flow and deformation localization in a neck for the samples of all states during tension. It should be noted that, when changing the harden ing of martensite in tempering, heat treatment changed the hardness, strength, and plasticity of the steel at comparable strainhardening exponents n (Fig. 4): n = 0.045 for the steel in state 1 and n = 0.036 in state 2 ( S nav = ±0.002). The stable flow threshold (eu = n) was reached only during tension of the steel samples in state 1 (strain εu = 8.3% corresponds to this state; see Fig. 3a). For the steel in state 2, deformation local ization in a neck was detected at eu < n. At the maxi
No. 4
316
KHANZHIN et al. σ, MPa 1500
Vu, dB
(a)
1000
60
500
0 σ, MPa 1500
0.08
1000
0
40
0.08
500
30
20
20
10
10
0.12 ε
0 σ, MPa 1500
0.04
0.08
0.12 ε Vu, dB
(d)
1000
60
50
50
40
40
30
0.04
60
40
60
500
1000
50
Vu, dB
(c)
Vu, dB (b)
50
30
0.04
σ, MPa 1500
500
30
20
20
10
10
0.12 ε
0
0.01
0.02 ε
Fig. 3. Stress–strain diagrams and AE diagrams during the tension of steel samples in states 1 and 2 before and after hydrogena tion: (a), (b) steel in states 1 and 2 without hydrogenation; (c), (d) steel in states 1 and 2 after hydrogenation.
mum load (εu = 5.6%), strain eu is only 67% of strain hardening exponent n (Fig. 4). The causes of the differences in the mechanisms of loss of stable flow in the steel of states 1 and 2 were revealed using AE measurements. During the tension of the steel samples of all states, the transition from elastic to plastic deformation is accompanied by a usual (due to deformation nonuniformity [6]) increase in the AE amplitude (to Vu = 30–50 dB) and a decrease (to 5–10 dB) after deformation at ε = 2–3%. From two to four strong AE pulses with an amplitude Vu = 15–25 dB are recorded in the steel samples of highstrength state 2 in the uniform deformation seg ment (Fig. 3b). The samples were cut along their axis after recording the last AE pulse. From two to four rounded microcracks (pores) 0.05–0.1 mm long ori ented across the sample axis were observed on unetched longitudinal polished sections at a magnifi cation up to ×200 (Fig. 5a).
Thus, several cracks (pores) become open in high strength state 2 at the stage of uniform deformation before the onset of local narrowing (at eu < n). Their visible size in the axial section accounts for 1–2% of the sample diameter. After the formation of axial cracks, a small increase in the strain (Δε = 0.4–0.7%; see Fig. 3b) was sufficient for a neck to localize and begin to form near a group of pores. Crack formation and coalescence generates an AE pulse at a load peak (Fig. 3b). In the state 2 steel, the neck—consequence of fracture, i.e., crack (pore) opening—localizes deformation earlier (at eu < n) than the unstable flow threshold is reached. AE pulses exceeding the noise level were not detected in state 1 at the stage of uniform deformation (Fig. 3a). After the load decreased in the necking zone (at ψconc = 30–40%), no microcracks were detected on four longitudinal and transverse polished sections. Here, the neck develops without cracks until the axial
RUSSIAN METALLURGY (METALLY)
Vol. 2013
No. 4
HYDROGEN EMBRITTLEMENT OF STEELS: II. EFFECT OF STRENGTH
(two sample diameters) in the naturally narrowing neck in the state 1 steel samples against 5–6 mm (one diameter) for the neck nucleating from an internal crack in the state 2 steel. When the neck forms from an internal crack in the state 2 steel, the grain strain along the axis is εd = 0.54 ± 0.01. In the state 1 steel (where a bottom crack appears shortly before failure), the grain elongation under fracture is larger (εd = 0.71 ± 0.02); that is, the crack cut the formed neck.
eu 0.06 1 2 0.04
3 4
0.02
0
0.04
0.02
0.06 n
Fig. 4. Relation between true uniform strain eu and strain hardening exponent n during the tension of steel samples in states 1 and 2 before and after hydrogenation: (1), (3) states 1 and 2 without hydrogen; (2), (4) states 1 and 2 with hydrogen (hydrogenation for 10 h).
stress sufficient for the “bottom” to be broken away is reached, and this is accompanied by the first AE pulse with an amplitude Vu = 20 dB. The concentrated elon gation to failure in this case is ΔLc = 1.0–1.4 mm (see Fig. 3a). Thus, the appearance of a neck in the hydrogen free samples is related to the following two causes depending on the steel strength: geometrical softening at eu = n in state 1 and the formation of internal cracks before the maximum load (at eu < n, eu = 0.67n (Fig. 4)) in the highstrength state 2 steel. Different origins of the neck determine its different profiles and different grain strains under fracture. The total necking zone length before failure is 7–10 mm
(а)
317
100 μm
In all cases, the fracture of sample upon tension has a starlike shape (Fig. 5b). A bottom crack 0.7–0.8 mm in diameter is observed in the central (axial) part of the fracture surface. Then, from 8 to 20 ridges (up to 0.5 mm in height) diverge along radii and enter into a shear cone (Fig. 5b). The star ridges result from the opening of longitudinal grainboundary cracks in the neck by tangential tensile stresses [12]. The level of tangential stresses is determined by the neck profile (radius of curvature and its minimum radius [5]). The structural causes of the cracks are as follows: plasticity anisotropy (because of grain elongation) and grain boundary embrittlement by segregation of phase parti cles [13]. The crack formation is accompanied by a series of strong AE pulses with an amplitude Vu = 20– 50 dB as the load decreases (Figs. 3a, 3b): 5–10 s before fracture in the state 1 steel (concentrated elon gation ΔLc = 0.3–0.6 mm) and 10–15 s before fracture in the state 2 steel (ΔLc = 0.7–0.9 mm). Thus, the fracture of all hydrogenfree states dur ing tension begins with the formation of a bottom crack along the sample axis. In the lowplasticity steel (state 2), a bottom crack nucleates in a cylindri cal sample without a neck. The bottom crack in the more plastic samples of state 1 appears in a formed neck. In all states, the fracture resistance in a neck is anisotropic because of the elongation of grains with carbide and carbonitride particles located on them.
(c)
(b)
1 mm
Fig. 5. Microcracks in the neck of state 2 steel samples and their fracture surfaces: (a) microcracks (pores), longitudinal polished section; (b) starlike fracture surfaces, hydrogenfree sample; and (c) flat cup fracture surface, hydrogenated sample. RUSSIAN METALLURGY (METALLY)
Vol. 2013
No. 4
318
KHANZHIN et al.
(b)
(а)
100 μm
Fig. 6. Cracks on the surface of state 1 steel samples after hydrogenation for 50 h under load and fracture during tension: (a) cracks on the sample surface and (b) crack on a longitudinal polished section of the sample.
3.2. Fracture Mechanism and Kinetics during Hydrogenation Surface cracks nucleated in a 10mmthick hydro genation zone during local galvanic hydrogenation at a constant load (σ = 0.8σ0.2). The sample subjected to delayedfracture tests failed in this zone or a neck formed there during tension after hydrogenation. During delayedfracture tests of samples hydroge nated under load, the first AE signals generated by
(а)
(b)
10 μm
10 μm
Fig. 7. Fracture surfaces of state 2 steel samples for delayed fracture after 10h hydrogenation: (a) region of crack nucleation and slow growth and (b) breakaway region.
microcracks with an amplitude Vu = 5–10 dB were recorded 8–10 min after the beginning of hydrogena tion of the state 2 steel samples. The incubation time of delayed fracture of the state 1 steel samples was sig nificantly longer: AE signal above the noise level were detected only 8–10 h after the beginning of tests. In 50 h, cracks (0.8–1.1 mm wide and 0.06–0.08 mm deep) are clearly visible on the sample surface in the hydrogenation zone and on polished sections prepared from this zone (Fig. 6). The growth of one of these cracks leads to the fracture of the sample in 10–15 h during delayed fracture of the steel in state 2. An ellip tical surface crack grew via cleavage along the bound aries of the former austenite grains at a velocity of 6 × 10–6 mm/s (Fig. 7a). As usually, crack growth during hydrogen embrittlement is accompanied by AE pulses [14]. When the crack achieved its critical length (~0.4–0.7 mm), the sample underwent cleavage fail ure and a ductile–brittle fracture surface of the type “flat cup”⎯grainboundary facets with ductile bridges forms (Fig. 7b). The flat cup bottom is bounded along the perimeter by a narrow (0.5– 0.7 mm thick) metal shear cone, which formed when the plastic flow front reached the free sample surface (Fig. 5c). The surface crack growth rate in the state 1 steel is lower by an order of magnitude (5.2 × 10–7 mm/s). Delayed fracture under a load during continuous hydrogenation did not result in fracture of the steel samples in this lowstrength state 1 during 168h tests. We tested samples by tension to failure to estimate the ductility of the steel at various hydrogenation times and, correspondingly, different degrees of hydrogena tion. Figures 3c and 3d show the typical stress–strain diagrams of the hydrogenated samples and the related AE diagrams. When the hydrogenated steel samples in states 1 and 2 were tensioned, AE signals from surface cracks were recorded in the elastic region.
RUSSIAN METALLURGY (METALLY)
Vol. 2013
No. 4
HYDROGEN EMBRITTLEMENT OF STEELS: II. EFFECT OF STRENGTH
The fracture kinetics of the state 1 steel samples after 50 h hydrogenation does not differ from that of the hydrogenfree samples, and the fracture geometry is the same, i.e., starlike shape (Figs. 3a, 3c). A stable flow is lost at eu = n (see Fig. 4). The concentrated elongation from the time of breakaway of the bottom upon a decrease in the load to failure is ΔLc = 1.8– 1.9 mm. The opening of the bottom to the cutting of the cone for five samples is f = 0.29 ± 0.04 mm, which accounts for 13% of the concentrated elongation of the sample. This value is close to f/ΔLc = 0.12 (f = 0.17 ± 0.01 mm, ΔLc = 1.3–1.4 mm) during tension of the hydrogenfree samples. The nucleation of a ductile bottom crack in a formed neck and its slow opening (with a rough bottom relief induced by pore coales cence) simultaneously with the neck shrinkage does not change the neck shape in the hydrogenated sam ples: the total neck zone length before fracture is 7.1 ± 0.2 mm (7.3 ± 0.2 mm without hydrogen). As the hydrogenation time increases, the bottom relief induced by pore coalescence becomes rougher. A deep surface crack changes the radial symmetric distribu tion of star ridges. For the state 1 steel samples, the uniform strain and the relative reduction of area in a neck at failure decrease noticeably (by 1.5%) only after 168h hydro genation under loading: ψ = 46–50% against ψ = 60– 63% without hydrogen. Upon longterm hydrogena tion, some samples subjected to tension have sepa rated regions in the fracture surfaces, where coarse plane axial cracks separate a sample into 2–3 sectors. For the state 2 steel samples, a short crack nucle ation time during hydrogenation under load and brit tle crack growth along grain boundaries sharply decrease the steel ductility margin for state 2 during tension. After 5h hydrogenation, uniform true strain eu accounts for only 9% of the possible strain at the achieved strainhardening exponent n (Fig. 4), and the state 2 steel samples after 8 to 15h hydrogenation undergo cleavage failure during tension in the elastic region (Fig. 3d). AE signals with an amplitude of 10– 55 dB accompany the growth of a surface crack from the beginning of tension to a critical length of 0.2– 0.3 mm. After short (~0.5%) concentrated deforma tion in a neck, the samples undergo “chevron” cleav age failure with a low grain strain (εd = 0.027 ± 0.004) under flat cup fracture. The residual relative reduction of area in a sharp neck (1.5–2.3 mm long) is ~7%. The bottom crack opening at fracture does not exceed f = 0.01–0.02 mm, which is 2–4% of the concentrated elongation (ΔLc = 0.5 mm). As follows from the measurement of the released hydrogen volume, the hydrogen solubility in the state 2 steel is higher than in state 1 by a factor of 40 under the same hydrogenation conditions at the same measure ment time. However, dissolved hydrogen can only weakly affect the fracture resistance of the metal. For example, the results of mechanical tests of the samples RUSSIAN METALLURGY (METALLY)
Vol. 2013
319
prepared from a steel cylinder for hydrogen storage after eight years of its operation at a hydrogen pressure of 32 MPa (steel structure consisted of tempered sor bite) showed that the relative elongation and the rela tive reduction of area did not change despite the fact that the hydrogen concentration in the steel was 20 cm3/100 g [6]. 4. CONCLUSIONS Fracture during uniaxial tension of steel 35KhGM steel samples locally hydrogenated under a constant load begins in the same manner irrespective of their strength: surface cracks nucleate in the hydrogenation zone. A neck also forms in this zone. In the steel with a lower strength (state 1), fracture begins with the breakaway of the “bottom” in a long neck, i.e., either in the zone of the maximum tensile stresses along the sample axis or at the top of the deep est surface crack after longterm (more than 100h) hydrogenation. A bottom crack during concentrated neck elongation opens via pore coalescence and cre ates a ductile relief. The longer the time of preliminary hydrogen of a sample, the coarser the relief. Concen trated deformation during the tension of the hydro gencontaining samples of ductile state 1 creates the grain elongation and the tangential stresses required for the opening of longitudinal cracks in a neck. As the hydrogen concentration in steel increases, the fracture of samples changes from a symmetric star to asymmet ric star and, then, to brittle separation into layers. In the hydrogencontaining steel of higher strength (state 2), fracture during tension is fully controlled by surface hydrogeninduced cracks. The chain of events during tension along the length of the crack induced by hydrogenation under load that is smaller than the critical length is as follows: the growth of surface grainboundary cracks during elastic deformation of a sample, short uniform deformation, loss of stable flow and short concentrated deformation, and fracture of the sample from a critical crack with a flat bottom hav ing a mixed chevron fracture surface. The sharp embrittlement of the state 2 steel upon hydrogenation is caused by a decrease in both ductil ity components, namely, the crack nucleation and propagation energies. In contrast, the microplastic deformation of the state 1 steel blunts a crack, decreases the stress concentration at its tip, and hin ders brittle fracture. ACKNOWLEDGMENTS This work was supported by the analytic goal pro gram Development of the Scientific Potential of Higher School (2010–2011) (project no. 2.1.2/5662) and the program of organization and development of National University of Science and Technology MISiS.
No. 4
320
KHANZHIN et al.
REFERENCES 1. L. S. Moroz and B. B. Chechulin, Hydrogen Embrittle ment of Metals (Metallurgiya, Moscow, 1967). 2. S. M. Beloglazov, Hydrogenation of Steel during Electro chemical Processes (LGU, Leningrad, 1975). 3. W. W. Gerberich and Y. T. Chen, “Hydrogencontrolled cracking: an approach to threshold stress intensity,” Metall. Trans. A13 (2), 305–311 (1975). 4. J. P. Hirth and H. H. Johnson, “Hydrogen problems in energy related technology,” Corrosion 32 (1), 3–25 (1976). 5. R. V. Hertsberg, Deformation and Fracture Mechanics of Engineering Materials (Metallurgiya, Moscow, 1989). 6. M. A. Shtremel’, Strength of Alloys. Part II. Deformation (MISiS, Moscow, 1997). 7. G. N. Kasatkin, Hydrogen in Structural Steels (Intermet Inzhiniring, Moscow, 2003). 8. V. G. Khanzhin, S. A. Nikulin, V. A. Belov, et al., “Hydrogen Embrittlement of Steels: I. Analysis of the Process Kinetics Using Acoustic Emission Measure ments,” Deform. Razr. Mater., No. 8, 44–48 (2011). 9. S. A. Saltykov, Stereometric Metallography (Metal lurgiya, Moscow, 1976).
10. V. S. Radchenko, A. Yu. Romashev, E. A. Gronicheva, et al., “Estimation of the diffusion mobility of hydrogen in steels of various strengths from a LabVIEW and IMAQ Vision analysis of digital images taken with an optical microscope,” in Proceedings of IX International Conference on Educational, Scientific, and Engineering Applications in the LabVIEW Medium and National Instruments Technologies (RUDN, Moscow, 2010), pp. 446–448. 11. G. A. Filippov and V. I. Sarrak, “Local hydrogen distri bution and internal stresses in the structure of quenched steel,” Fiz. Met. Metalloved. 49 (1), 121–125 (1980). 12. S. A. Nikulin, M. A. Shtremel’, and V. G. Khanzhin, “Ductile fracture of highmanganese steel during ten sion,” Izv. Akad. Nauk SSSR, Ser. Met., No. 1, 145– 151 (1990). 13. M. A. Shtremel’, “Grainboundary fracture and frac ture toughness of steel,” Metalloved. Term. Obrab. Met., No. 11, 23–28 (1988). 14. H. N. G. Wadley and C. B. Scruby, “A study of defor mation and fracture processes in a lowalloy steel by acoustic emission transient analysis,” Acta Metall. 27, 613–625 (1979).
Translated by K. Shakhlevich
RUSSIAN METALLURGY (METALLY)
Vol. 2013
No. 4