Effect of ECAP and heat treatment on mechanical properties, stress relaxation behavior and corrosion resistance of a 321-type austenitic steel with increased delta-ferrite content

Hot rolled commercial metastable austenitic steel 0.8C-18Cr-10Ni-0.1Ti (Russian industrial name 08X18H10T, analog 321L) with strongly elongated thin delta-ferrite particles in its microstructure was the object of investigations. The lengths of these delta-particles were up to 500 mkm, the thickness was 10 mkm. The formation of the strain-induced martensite as well as the grinding of the austenite and of the delta-ferrite grains take place during ECAP. During the annealing of the UFG steel, the formation of the delta-phase particles takes place. These particles affect the grain boundary migration and the strength of the steel. However, a reduction of the Hall-Petch coefficient as compared to the coarse-grained (CG) steel due to the fragmentation of the delta-ferrite particles was observed. The samples of the UFG steel were found to have 2-3 times higher stress relaxation resistance as compared to the CG steel (a higher macroelasticity stress and a lower stress relaxation magnitude). The differences in the stress relaxation resistance of the UFG and CG steels were investigated. ECAP was shown to result in an increase in the corrosion rate and in an increased tendency to the intergranular corrosion (IGC). The reduction of the corrosion resistance of the UFG steel was found to originate from the increase in the fraction of the strain-induced martensite during ECAP.

The problem of increasing the strength of the austenitic steels preserving their high resistance to the intergranular corrosion (IGC) is one of the key problems of the materials science [5][6][7][8][9][10][11]. This makes the traditional approach to increasing the strength consisting in annealing leading to the nucleation of the chromium carbide particles at the austenite grain boundaries inapplicable [12][13][14][15][16]. In this connection, engineers are developing novel methods of simultaneous increasing the strength and the corrosion resistance of the austenite steels.
Forming the ultrafine-grained (UFG) structure is one of the popular methods of improving the strength and the operational characteristics of stainless alloys [17][18][19][20][21][22]. At present, various methods of Severe Plastic Deformation (SPD) are applied to form the UFG structure -Equal Channel Angular Pressing (ECAP) [17][18][19][20][21], high pressure torsion [22][23][24], rotary swaging [25,26], extrusion [26,27] etc. In spite of certain success in the improvement of the hardness and strength of steels, it should be noted that the application of SPD leads to strain-induced decomposition of austenite often [17,19,21,[29][30][31]. It may affect the corrosion resistance of the UFG austenitic steels negatively. In this connection, the applied problem of choice of the optimal regimes of heatdeformation processing of austenitic steels, which allow increasing the strength of the ones without the reduction of the corrosion resistance is relevant.
A problem of providing a high stress relaxation resistance of the austenitic steels is even more complex. The problem of increasing the relaxation resistance is especially important in the development of machine-building hardware providing simultaneously high characteristics of fatigue, creep resistance, stress corrosion cracking resistance, etc. [32][33][34][35][36]. The high stress relaxation resistance determines the capability of the hardware to provide the necessary level of downforce during a long operation time [37,38]. The improvement of the stress relaxation resistance of the materials with simultaneous proving a high strength will allow increasing the downforce of the hardware and keeping it during a notably longer operation time. Plenty of experimental and theoretical works was devoted to the problem of investigation of the stress relaxation resistance mechanisms for the coarse-grained (CG) materials [39][40][41][42]. For CG materials, it is supposed usually that the higher the level of internal stresses, the lower the stress relaxation depth (the magnitude of the decrease in the stress in given time interval). Therefore, the strain strengthening is a traditional method of increasing the stress relaxation resistance. From this viewpoint, the fine-grained metals and alloys fabricated using the SPD methods are promising candidates for application as the base materials for the heavy-duty relaxation-proof hardware.
The present work was aimed at studying the effect of SPD and annealing on the relaxation resistance and the resistance to IGC of the Russian metastable austenitic steels 0.8%C-18%Cr-10%Ni-0.1%Ti (Russian industrial name 08X18H10T, Russian analog of steel 321L). This steel is used widely in nuclear mechanical and power engineering for making the machine building hardware operated in the condition of simultaneous impact of elevated temperatures, mechanical loads, and corrosion-aggressive ambients. In particular, a low strength and high stress relaxation rate in the austenite steels result in difficulties in operations of assembling and disassembling the products after long-term operation. An increased content of -ferrite is a special feature of the object of investigations. It is a defect of casting or of heat treatment of the cast workpieces but is present in the bulk austenite steel often.

Materials and methods
The Russian commercial metastable austenitic steel 08Х18Н10Т (composition: Fe-0.08wt.%С-17.9wt.%Cr-10.6wt.%Ni-0.5wt.%Si-0.1wt.%Ti) was the object of investigations. The formation of the UFG microstructure in the steel was performed by ECAP. The workpieces of 1414140 mm in sizes were cut out from hot-rolled rods of 20 mm in diameter. Prior to ECAP, the rods were annealed at 1050 о C for 30 min followed by quenching in water. ECAP was performed using Ficep HF400L press (Italy). The angle of crossing the working channel and the output one was /2. In the ECAP regime used, the workpiece rotated at the angle of  around its longitudinal axis during every cycle (regime "С", see [52]). The ECAP rate was 0.4 mm/s. The ECAP temperatures were 150 and 450 C, the number of pressing cycles (N) varied from one to four.
The investigations of the steel microstructure were carried out using Jeol  JSM-6490 and Tescan  Vega Scanning Electron Microscopes (SEMs) and Jeol  JEM-2100F Transmission Electron Microscope (TEM). X-ray diffraction (XRD) phase analysis of the stainless steels was carried out using Shimadzu  XRD-7000 X-ray diffractometer (CuK emission, recording in the Bragg-Brentano scheme in the range of angles 2 = 30-80 о with the scan rate 1 о /min). The crystal lattice parameters were determined and the mass fractions of the phases were calculated by Rietveld method.
The microhardness (Hv) of the steel was measured with Duramin  Struers 5 microhardness tester. The uncertainty of the microhardness measurements was ±50 MPa.
For the mechanical tests, flat double-blade shaped specimens were made by electric spark cutting. The sizes of the working part were 223 mm. The tension tests were carried out using Tinius Olsen  H25K-S machine with the strain rate 3.310 -3 s -1 (the tension rate was 10 -2 mm/s).
The tension tests were performed at the room temperature (RT) and in the temperature range 450-900 C. The specimens were heated up to the testing temperatures in 5 min. The specimens were kept at the testing temperatures for 10 min to establish the thermal equilibrium. In the course of the tests, the curves stress () -strain () were recorded. From these curves (), the magnitudes of the ultimate strength (b) and of the maximum relative elongation to failure () were determined.
The fractographic analysis of the fractures after the tension tests was carried out using Jeol  JSM-6490 SEM. The macrostructure of the specimens after the failure tests was investigated using Leica  IM DRM metallographic optical microscope. The investigations of microstructure and the microhardness measurements were performed in the fracture zones ("deformed area") and in the non-deformed areas near the capturers.
The stress relaxation tests were performed according the technique described in Appendix A to the paper [53]. For the tests, the rectangular specimens of 3×3 mm in cross-sections and of 6 mm in height were made. The specimens were loaded with the rate 0.13%/s during 0.3 s. Afterwards, the specimens were kept under a constant stress (i) during given stress relaxation time (tr = 60 s).
In the course of stress relaxation, a curve of the stress on the testing time i(t) was acquired.
Afterwards, the next loading step was performed. As a result of experiment, a dependence of the stress relaxation magnitude i on the magnitude of the summary load applied i() was obtained.
The dependence obtained was used also to determine the macroelasticity stress (0) and the yield strength (y).
The resistance of the steels to the intergranular corrosion (IGC) was investigated using R-8 potentiostat-galvanostat according to Russian National Standard GOST 9.914-91 by double loop austenite steel demonstrates an increased tendency to IGC.
The Tafel curves ln(i) -E were measured in the same medium. From the Tafel curves, the corrosion current densities (icorr, mA/cm 2 ) and the corrosion potentials (Ecorr, mV) were obtained by standard method. Prior to the corrosion investigations, the surfaces of specimens of 51010 mm in size were subjected to mechanical grinding and polishing. From the results of measuring the icorr, the corrosion rate was calculated using the formula: Vcorr = 8.76icorrM/F, where  is the density of iron [g/cm 3 ], М -molar mass [g/mol], F = 96500 C -Faraday's constant. The verification tests of resistance against IGC were conducted according to GOST 6232-2003 by the boiling of the specimens in a solution of 25% H2SO4 + CuSO4. The character of the surface destruction after the corrosion tests was analyzed using Leica  IM DRM metallographic optical microscope.
To study the thermal stability of the structure and properties of the UFG steel, the specimens were annealed in air in the temperature range from 100 up to 900 C. The isothermic holding time was 60 min. Th uncertainty of maintaining the temperature was ± 10 C. The specimens were cooled down in water.

Microstructure investigations
As initial state, stainless steel had a uniform austenite microstructure (Figs. 1a-1d). The mean austenite grain sizes were ~20 m. The thin (up to 10 m in thickness) strips of the ferrite phase elongated along the deformation direction were observed in the microstructure of the CG steel (Figs. 1a-1d). The lengths of the -ferrite stripes were ~500 m. The lattice dislocations (Fig.   1f) as wee as few micron-and submicron-sized titanium carbide and carbonitride particles (Fig. 1e) were observed inside the austenite grains.
After the first ECAP cycle, the macrostructure of the steel workpieces comprised of alternating macro-bands of localized strain (Fig. 2). After N = 4 ECAP cycles, the specimens had a uniform macrostructure. Fig. 3a presents the XRD curves from the steel specimens in the initial state and after ECAP.
The results of the XRD phase analysis evidence the mean mass fraction of the -phase in the steel in the initial state to be ~1.5-3 %. The lattice parameter of the -phase in the steel Fe-Cr-Ni-Ti was 2.8869 Å, the one of the -phase was 3.5875 Å.
ECAP leads to an increase in the fraction of the -phase because of appearing the straininduced martensite. The scale and dynamics of the increasing of the strain-induced martensite with increasing the number of cycles depends on the ECAP temperature (Fib. 3b). After ECAP at 150 С, the mass fraction of the ()-phase was 5.9-7.7% and didn't change essentially with increasing number of ECAP cycles up to N = 4. The increasing of the SPD temperature up to 450C resulted in more intensive decomposition of the -phase (austenite) -the mass fraction of the ()-phase increased from 3.6% after N = 1 cycle up to 17.4% after N = 4 ECAP cycles. Note that the effect of strain-induced decomposition of the -phase during ECAP of the austenitic steel is known for a long enough time [19][20][21][22][23]. At the same time, it is worth noting that the effect of accelerated straininduced decomposition of austenite at higher SPD temperatures is an unexpected ones (see [24]). Along with the austenite strain-induced decomposition during ECAP, a grinding of the steel grain microstructure was observed. After N = 4 ECAP cycles at 150 and 450 C, an UFG microstructure with mean grain sizes of 0.3 and 0.5 m, respectively formed in steel (Fig. 4). For the specimens of steel obtained by ECAP at 150 о C, the crossing localization micro-bands, which lead to different orientation of the austenite grains were observed at the microscopic level (Fig. 4a, c, d). The microstructure of the specimens after ECAP at 450 о C was more uniform, no shear microbands manifested clearly were observed (Fig. 4d). The nanotwins are seen in some austenite grains ( Fig. 4d), which can be classified as the strain-induced martensite according to [17][18][19][20]. No nucleation of the chromium carbide particles was observed in the steel microstructure after ECAP.
No -phase particles were reveled in the UFG microstructure during the metallographic and SEM investigations that allows making a conclusion on a strong fragmentation of these ones during ECAP. The presence of separate point reflections in the electron diffraction pattern evidenced the presence of the high-angle grain boundaries in the UFG steel obtained by ECAP at 450 о C (Fig. 4f).
The electron diffraction patterns from the specimens of the UFG steel after ECAP at 150 о C were more blurry (Fig. 4b).
The investigations of the thermal stability of the UFG microstructure during annealing demonstrated the temperature of recrystallization in the UFG steel (N = 4, TECAP = 450 C) to be Т1 = 750 C. The recrystallization had a clearly expressed abnormal character accompanied by a formation of a multi-grained structure. After annealing at 750 C, 1 hour, the recrystallized metal regions with the mean grain sizes of 5-7 m were observed in the microstructure of the UFG steel.
The volume fraction of these regions was ~3% or less. At increased annealing temperatures, an increase in the volume fraction of the recrystallization metal as well as an increase in the mean grain sizes were observed -after annealing at 900 C, 1 hour, an equiaxial austenite microstructure with the mean grain sizes of 8-12 m formed in the steel (Fig. 5). Increasing the number of ECAP cycles up to N = 4 at ТECAP = 450 C didn't result in any change of the recrystallization temperature Т1 but was accompanied by a decrease in the mean recrystallized grain sizes (Fig. 5).
The XRD phase analysis demonstrated a decrease in the mass fraction of the ()-phase with increasing annealing temperature up to 600°C (1 h). After annealing at 800 о C, the mass fraction of the ()-phase was beyond the measurement uncertainty ±1 wt.% (the intensity of the XRD peaks from the ()-phase didn't exceed the noise level) regardless to the ECAP regimes.
The in situ TEM investigations demonstrated the nucleation of the light-colored nanoparticles in the UFG steel when heating up to 600 о C. The mean size and the volume fraction of the particles increased with increasing heating temperature. After heating up to 800 о C and holding for 0.5 hrs, the mean particle size was about 50 nm (Figs. 6, 7). Because of the presence of the ()-phase having an essential residual magnetization in the steel, we failed to analyze the composition of the nucleated second phase nanoparticles by EDS. We suggest these particles to be the -phase ones. The peaks from the -phase were absent in the XRD curves from the annealed specimens, probably, due to small sizes of the nucleated particles. Earlier, the possibility of nucleation of the -phase particles during annealing of UFG steel 08Х18Н10Т was reported in [18,19]. The analysis of the yield strength dependence on the mean grain sizes has shown that this dependence can be interpolated by a straight line in the y -d -1/2 axes with a good precision (Fig.   8b). This evidences the Hall-Petch relation to hold: where K is the grain boundary hardening coefficient (Hall-Petch coefficient) describing the contribution of the grain boundary structure in the strength of the metal. The mean value of the coefficient K determined from the dependence in Fig. 8b is K = 0.46 MPam 1/2 . So far, one can conclude the processing of the austenitic steel by ECAP to result in the increasing of its stress relaxation resistance -in the increasing of the macroelasticity stress o (see above) and in the decreasing of the stress relaxation magnitude i at increased loads.
The recrystallization annealing resulted in a decreasing of the stress relaxation resistance parameters of the UFG steels -as one can see in Fig. 12b, the increasing of the annealing temperature above 650-700 о C resulted in a displacement of the stress relaxation curves i() towards the smaller stresses. After annealing at 800-900 о C, the stress relaxation curves of the deformed steel specimens had usual tree-stage character corresponding to the stress relaxation curve i() of the coarse-grained steel specimens (Fig. 12a).
Tension testing at elevated temperatures Table 2 presents the dependencies of the ultimate strength and of the elongation to failure on the testing temperature for the coarse-gained steel specimens and for the UFG ones obtained in different ECAP temperatures. Fig. 13 presents the stress-strain curves () for the tension tests at elevated temperatures.
The stress-strain curves () for the CG steel specimens had the form typical for highplasticity materials (Fig. 13a). The duration of the localized plastic strain stage was much smaller Note that the increasing of the testing temperature resulted in a nonmonotonous variation of the elongation to failure for the UFG steel that differs from the same dependencies for the CG steel.
The analysis of the data presented in Fig. 13b shows the elongation to failure for the CG steel to decrease monotonously from 125% down to 70% with increasing testing temperature from RT up to 750 C. The character of the dependence (Т) for the UFG steel was more complex -the elongation to failure decreased insufficiently with increasing testing temperature from RT up to 450 C. At further increasing of the temperature from 450 C up to 750-800 C, the elongation to failure of the UFG steel increased and was several times higher than the  of the CG steel. For the UFG steel specimens obtained by ECAP at ТECAP = 150 C, the elongation to failure at the testing temperature of 750 C reached 250 %. Further increasing of the testing temperature resulted in a decreasing of the elongating to failure for the UFG steel specimens again.
The fractographic analysis of the fractures (Fig. 14) demonstrated the areas of the fibrous zones and of the radial ones to increase and the areas of the cut zones -to decrease with increasing testing temperature. At the testing temperature of 600 C, the cur zone area didn't exceed 5-10% of the whole fracture area. For the UFG steels (TECAP = 450 C), the cut zones were absent that also evidences an increased plasticity of the UFG material as compared to the coarse-grained state.
In our opinion, the non-monotonous character of the dependence of the elongation to failure on the testing temperature for the UFG steel originates from the recrystallization processes in the UFG steel starting after annealing at ~650-700 C.
In particular, the microhardness testing results (Table 2)  the tension tests (Fig. 15). As one can see from Fig. 15 and from the data presented in Table 3, the testing at 800 о С resulted in the formation of a well uniform fine-grained structure. No essential grain growth was observed. The mean grain sizes in the deformed parts were slightly smaller than in the non-deformed ones.
Corrosion resistance Fig. 16a presents the Tafel curves ln(i)-E for the coarse-grained steel specimens and for the UFG ones. The results of the electrochemical testing are summarized in Table 3. The curves ln(i)-E had usual character. One can see the coarse-grained steel specimens to have smaller corrosion rates than the UFG ones. For the UFG steel specimens obtained by ECAP at Т = 450 о C, the values of mean corrosion current density icorr (of the mean corrosion rate Vcorr) were 10-15% greater than the same characteristics for the UFG steel specimens obtained by ECAP at 150 о C.  Table 3. It follows from the data presented in Table 3 that the ratios of the areas under the passivation curves (S1) and the reactivation ones (S2) (K = S1/S2) were small and appeared to be much less than the ultimate value Kmax = 0.11. This result evidences both coarse-grained steel and UFG one to be highly resistant against IGC. At the same time, the magnitudes of the coefficient K for the UFG steel specimens were 1.5-2.5 times higher than for the CG ones. The metallographic analysis of the surfaces has shown the large elongated -ferrite particles to be the places of accelerated corrosion destruction of the surfaces in the DLEPR testing (Fig. 17a). No IGC traces were observed on the surfaces of the UFG steel specimens (Fig. 17b).
The results of standard tests of the resistance against IGC according to GOST 6232-2003 confirmed high corrosion resistance of the UFG steels. As one can see in Fig. 18a, after testing during 24 hrs, the corroded elongated -ferrite particles were observed on the CG steel surfaces. In some areas of the surfaces, the IGC corrosion defects or the pitting corrosion no more than 10-15 m in depth are seen. On the surfaces of the steel specimens with the UFG structure formed as a result of 1 or 2 ECAP cycles, few corrosion pits were observed. On the surfaces of the UFG steel specimens (N = 3, 4), the corrosion defects were absent (Fig. 18b).
So far, the UFG steel specimens have high strength, stress relaxation resistance, and high resistance against the intergranular corrosion simultaneously. It allows an efficient application of the UFG steel for making the stress relaxation-proof machine-building hardware utilized in the conditions of enhanced loads and corrosion-aggressive media.

Investigation of thermal stability
First, one should pay attention to the fact of nucleation of the second phase particles during the annealing of the UFG austenitic stainless steel. It is interesting to note that the nucleation of the second phase particles was observed not in all grains. In our opinion, the nucleation of the particles goes preferentially inside the grains of -phase, the lattice constant of which is much less than the one of the -phase. It leads to a formation of a strongly supersaturated solid solution (of chromium) in the -phase grains and, as a consequence, to its nucleation at further heating up. This assumption allows suggesting the nucleation of ferromagnetic -phase particles Fe-Cr to take place in the course of heating up. We suppose the nucleation of chromium carbide Cr23C6 particles to be hardly probable in this case since steel contains titanium, which reacts chemically with carbon and forms titanium carbide TiC [10, 18,20,22]. The possible nucleation of the -phase particles during annealing of metastable UFG austenitic steel Fe-Cr-Ni-Ti was reported in [18,19].
The analysis of the grain growth process revealed the grain growth activation energy (QR) determined from the slope of the dependence ln(d n -d0 n ) -Tm/T to be 6.0-8. Note also that at n = 2, the recrystallization activation energy QR takes non-physical values (3-4.3 kTm ~ 45-63 kJ/mol), which appear to be smaller than the activation energy of diffusion in the iron melt (see [52]). In our opinion, it evidences indirectly the nucleating nanoparticles to affect the grain boundary migration in the deformed austenite steel essentially.

Mechanical properties
The yield strength in the austenitic steel can be calculated using Hall-Petch equation (1) where the magnitude of the macroelasticity stress in the first approximation can be calculated as the sum of the following contributions: where PN is the stress of resistance of the crystal lattice (the Peierls-Nabarro stress),  с = ∑ A C According to [55,56], the contribution of the crystal lattice of the doped austenite for steel and the heat-resistance nickel alloys is PN = 60-70 MPa. The contribution of the second phase particles can be neglected in the first approximation since the nucleated particles were large enough Note also that the intensities of increasing of the macroelasticity stress о and of the yield strength y with increasing number of ECAP cycles (N) were different (Fig. 9a). Analysis of the data presented in Fig. 9a shows the magnitude of gb = y -о = Kd -1/2 in the initial state to be 175 MPa and to increase up to gb = 645-655 MPa with increasing N up to 3-4 (ТECAP = 450 C). Note that at the same time, the magnitude of the grain boundary hardening coefficient K calculated according to the formula K = (y -о)d 1/2 (see (1) In our opinion, the decreasing of the coefficient K in ECAP is related to the fragmentation of strongly elongated -ferrite particles (up to 10 m in thickness and up to 500 m long). The harder -ferrite particles crossing the austenite grains often (Fig. 1) can impede the propagation of the strain in the austenite grains as well as the "transfer" of the strain from one austenite grain to another. In our opinion, strong fragmentation of the harder particles during ECAP helps eliminating the additional type of the "barrier" obstacles and promotes the strain at the micro-and macrolevels.
In our opinion, the fragmentation of the large elongated -ferrite particles is one of the possible origins of the presence of the uniform strain flow stage in the stress-strain tension curves at room temperature (Fig. 10). As it has been shown above, the UFG steel has a higher stress relaxation resistance -the stress relaxation magnitude i in the UFG steel were much smaller at the same stress applied (Fig.   12a). Let us analyze the stress relaxation mechanisms in UFG steel underlying its improved stress relaxation resistance.
In general, the accommodative reconstruction of the defect structure (first of all -of the dislocation one) is well known to be the primary stress relaxation mechanism. In the coarse-grained materials at RT, the lattice dislocation glide in the field of the point defects distributed uniformly is such a mechanism most often. The dependence of the strain rate ε̇ on the stress  in this case can be described by the following formula where ε̇ is the pre-exponential factor, F is the activation energy of dislocation glide depending on the obstacles type, k is the Boltzmann constant, Т is the testing temperature, and  * is the nonthermal flow stress, which can be taken to be equal to the ultimate strength [55].
In the first approximation, the strain rate in the stress relaxation tests can be accepted to be proportional to the stress relaxation one: ε̇= σ̇/E, where E is the elastic modulus. The stress relaxation rate can be calculated as σ̇=  t ⁄ . Since the stress relaxation time tr = 60 s and Е = 217 GPa were the same for all specimens, the magnitude of the activation energy F/kT can be determined from the slope of the dependence ln() -1-/b (Fig. 19a).
As one can see in Fig. 19a, the dependence ln() -1-/y for the CG steel has a two-stage character. The activation energy of dislocation glide in the microplastic strain range is F1 ~4.8 kT (~0.62 Gb 3 ) that matches well to the data published in the literature (~ 0.5 Gb 3 for the steels AISI 304 and AISI 316 [57]). It allows concluding the gliding of lattice dislocations in the long-range stress field from other lattice dislocations to be main stress relaxation mechanism within the microplastic strain stage. At increased stresses, the activation energy of overcoming the obstacles tends to F2 ~ 0.9 kT (~0.12 Gb 3 ). According to the classification of [57], the obstacles with F < 0.2 Gb 3 are the classified as the "weak" ones for the dislocation motion. In the case of the coarsegrained steel deformed in the macroscopic strain range, obviously, the austenite grain boundaries can be such obstacles.
In the case of the UFG steel, the stage with the increased F1 ~4.9-6. . It is interesting to note that the magnitude of the activation energy F1 for the annealed UFG steel decreases monotonously from 5.6 kT (0.70 Gb 3 ) at Т = 800 о C up to 9.2 kT (1.16 Gb 3 ) at Т = 900 о С (Fig. 19b). In our opinion, this result is related to the nucleation of the second phase particles in the course of heating up (Fig. 7).
So far, the formation of the long-range internal stress fields from the nonequilibrium grain boundaries, which prevent the free motion of the lattice dislocations (prevent the accommodative reconstruction of the defect structure) is the origin of the increased stress relaxation resistance of the UFG steel. In our opinion, the increasing of the volume fraction of strain-induced martensite and, hence, the formation of the two-phase + microstructure is the main origin of increased corrosion rate and of reduction of the resistance against to IGC in the UFG steels. The strain-induced martensite particles having a different chemical composition (unlike austenite) have a higher corrosion (dissolving) rate. Therefore, the increasing of the volume fraction of strain-induced martensite will lead to increasing of the uniform corrosion rate according to the ordinary rule: Vcorr = fV + fV where V and V are the dissolving rates for the and -phases, respectively.
The formation of the two-phase microstructure leads to the appearing of the microgalvanic couples austenite -martensite in the material. These ones are the places of accelerated corrosion destruction during the electrochemical tests for IGC. So far, the increasing of the volume fraction of strain-induced martensite provides the conditions for the increase in the uniform corrosion rate and in the intergranular corrosion one.
The second factor promoting the reduction of the resistance of the stainless steel against IGC after ECAP can be the redistribution of the doping elements (chromium and nickel) during SPD. In [58], the grain boundaries in the nanocrystalline austenitic steel Fe-12%Cr-30%Ni with the grain size ~60 nm were shown to be enriched with nickel after SPD but to have a reduced chromium concentration. The width of the near-boundary zone enriched with nickel was predicted theoretically to increase with increasing temperature [58]. The strain-induced segregation of the Ni atoms at the austenite grain boundaries was utilized to explain the formation of the ferromagnetic clusters at the grain boundaries in the Fe-12%Cr-30%Ni and Fe-12%Cr-40%Ni steels during SPD [59]. Such a deformation-stimulated bundle of the solid solution would promote an accelerated electrochemical corrosion near the grain boundaries in the UFG steel Fe-18%Cr-10%Ni-0.1%Ti.   -120  250  290  105  240  185  290  120   800 220  75  250  110  150  200  200  150  152  220  205  160   900  ------98 190 ----