2022 Volume 62 Issue 10 Pages 2095-2106
Resistance to Sulfide Stress Cracking (SSC) caused by local hard zones of pipe inner surface has been required in low alloy linepipe steel. In this study, using two samples with different surface hardness, the detailed SSC initiation behavior was clarified by four-point bend (4PB) SSC tests in which immersion time and applied stress were changed in a sour environment containing 0.15 bar hydrogen sulfide (H2S) gas. SSC cracks occurred when the applied stress was higher than 90% actual yield strength (AYS) in higher surface hardness samples over 270 HV0.1. From the fracture surface observation of SSC crack sample, it was found that the mechanism gradually shifted from active path corrosion (APC) to hydrogen embrittlement (HE), and that the influence of APC mechanism remained partially in the process of SSC initiation at the tip of corrosion pit or groove. The polarization measurement in the 4PB SSC test showed that the anodic and cathodic reactions (especially cathodic reactions) were activated when the applied stress was 90% AYS or higher. The FEM coupled analysis simulating the stress and strain concentration at the bottom tip of the corrosion groove and the hydrogen diffusion and accumulation was carried out. The principal stress in the tensile direction showed the maximum value at 0.04–0.06 mm away from the tip of the corrosion groove, and the hydrogen accumulation became the maximum. It was analytically found that the SSC crack initiated and propagated with HE mechanism dominated type when the threshold value of about 0.82 ppm is exceeded.
Low alloy steel line pipe (UOE pipe) manufactured by using plate steel as a raw material is used for natural gas transportation pipeline.1) As the characteristics of the line pipe, corrosion resistance is required in addition to strength and toughness.2) In particular, in a wet sour environment containing corrosive gases such as hydrogen sulfide (H2S) gas and carbon dioxide (CO2) gas, sufficient resistance to hydrogen-induced cracking (HIC) and sulfide stress cracking (SSC) is required.3,4,5)
During pipeline operation, tensile stress is applied in the circumferential direction of the inner surface of the pipe, and in the sour environment, there is a fear of a fracture accident due to SSC. Since the SSC is mainly affected by three factors: material,6,7,8) sour environment,9,10) and applied stress,11,12) the material has a hardness limit as shown in the NACE MR 0175/ISO 15156-1 standard, and a hardness upper limit of 22 HRC (about 250 HV10) is specified for carbon steel and low alloy steel.13) However, as the drilling environment has become more severe in recent years, it has been found that this hardness restriction alone is not always sufficient in the boundary region where the H2S concentration is high.14,15) Therefore, in the previous report,16) an experimental investigation was carried out on the effect of local hard zone near the extreme surface layer on the surface corrosion behavior and SSC crack initiation and propagation. The previous report clarified the effect of the extreme surface layer hardness distribution in the range of about 1 mm from the inner surface of the low alloy line pipe (especially the micro hardness HV0.1 at the position of 0.25 mm of the surface layer in the pipe) and the H2S partial pressure condition on the SSC resistance. As a result, it was found that the critical hardness value of the extreme surface layer where SSC cracks occur is about 270 HV0.1 under the low pressure condition of 0.15 bar H2S partial pressure, and 250 HV0.1 under the condition of 1 bar or more H2S partial pressure. When the partial pressure of H2S was 0.15 bar (less than 1 bar) and 1 bar or more, the surface corrosion behavior and crack formation were different, and the SSC mechanism was different. At 0.15 bar, the formation of FeS,17) which is considered to have corrosion protection, was thin and heterogeneous. SSC is a combined type mechanism of Active Path Corrosion (APC) and Hydrogen Embrittlement (HE) with local corrosion pit formation. In the case of 1 bar or more, corrosion mainly occurred on the whole surface, and the amount of hydrogen penetration increased as the concentration of H2S increased,18,19) and SSC crack initiation was mainly caused by HE mechanism. Thus, it is suggested that the contribution of APC and HE mechanisms differs depending on the environment of H2S partial pressure.
In the previous study,19,20,21,22) SSC of low alloy steel is regarded as a kind of HE. On the other hand, other studies on low-concentration H2S environments23,24,25,26,27) suggest a strong contribution of APC as a precursor process. For example, Yamane et al.24) have shown in SSC tests under potentiostatic conditions that the transgranular microcracks (also called fisher) grow deeply in accordance with anodic polarization when Ni is added at 0.5 mass% or more in high strength materials of YS 780 MPa class. They have proposed an SSC mechanism in which HE occurs at the bottom tip of the transgranular microcrack, which is a localized corroded hole. The authors examined the SSC mechanism in the low H2S partial pressure condition of 0.15 bar for the general X65 grade (YS≧450 MPa) material as a low alloy steel line pipe, and clarified that the transgranular microcrack formation process is the anodic dissolution control based on the APC mechanism.28) In addition, it was clarified that the local dissolution was promoted by plastic deformation29) and higher material hardness,30) and it was estimated from compact tension (CT) test with pre-crack that the initiation and propagation process of SSC crack was caused by HE mechanism.16) However, since there exist multiple factors such as microscopic structural morphology and crystal orientation distribution of bainite structure, microscopic elasto-plastic deformation of tensile surface with increasing applied stress, and sour environmental conditions such as H2S partial pressure and pH, it is not clear how these factors relate to the initial corrosion pit and groove formation process and SSC crack initiation and propagation process. In particular, it is not clear whether the mechanistic transition from the iron dissolution corrosion reaction of the APC mechanism to the crack initiation/propagation mode of the HE mechanism actually occurs, and if so, what the criteria for the mechanistic transition are.
Crack initiation and propagation in SSC are considered to be cracks governed by the HE mechanism which is determined by three factors: “local sour environment such as pH at the bottom tip of corroded pit and groove”, “degree of stress and strain concentration at the bottom tip of corroded pit and groove”, and “material properties at the bottom tip”. The purpose of this study is to clarify experimentally and analytically that the process of SSC crack initiation from the bottom tip of corroded pit and groove is mainly controlled by HE mechanism in 0.15 bar H2S environment in low alloy steel line pipe. First, the details of SSC initiation behavior were clarified by four-point bend SSC test using the material in the vicinity of SSC initiation limit hardness 270 HV0.1 in the 0.15 bar H2S environment by changing immersion time and applied stress. And, the existence of the transition from APC mechanism to HE mechanism was evaluated by the crack fracture surface observation of SSC crack material. Next, the polarization measurement of the four-point bend SSC test was carried out, and the effect of material and applied stress on the anode and cathode activation behavior after 24 hours of the initial immersion period was examined. Furthermore, assuming that HE at the bottom tip of the corroded pit and groove is the dominant mechanism for SSC crack initiation and propagation, FEM coupled analysis was carried out to simulate stress concentration and hydrogen diffusion and accumulation at the bottom tip of the corroded pit and groove, and the mechanism transition behavior from APC to HE and the validity that crack initiation is dominated by HE mechanism were verified.
Low carbon low alloy steel X65 grade line pipes (YS≧450 MPa) with 20 or 30 mm thickness with bainitic microstructure were used as a test material. Coupons of 300 mm square were cut from pipes with different surface cooling rates under TMCP (Thermo-Mechanical Controlled Process) conditions during the production of plate steels, and subjected to coating simulation aging treatment at 250°C for 1 hour. The specimen for microstructure observation was mechanically polished to a mirror surface, and after 3% nital-etching, microstructure observation was performed. In addition, in the crystal orientation analysis (EBSD: Electron Backscatter Diffraction), finish polishing was performed using colloidal silica to a measurable extent. Figure 1 shows the typical microstructure of the surface parallel to the inner surface of the pipe at 0.2 mm position by SEM (Scanning Electron Microscope) observation. Based on the results of the previous report,16) samples having a hard lath bainite (LB) matrix at a surface cooling rate exceeding 200°C/sec (It was cut from the inside region of the pipe thickness of 30 mm. Hereinafter referred to as 15W.) and samples having a granular bainite (GB) matrix at a surface cooling rate of about 100°C/sec (It was cut from the inside region of the pipe thickness of 20 mm. Hereinafter referred to as 9W.) were used. The former 15W corresponds to a sample of 270 HV0.1 with SSC initiation limit hardness in a 0.15 bar H2S environment with low H2S partial pressure. Figure 2 shows the IPF (Inverse Pole Figure) orientation map and the KAM (Kernel Average Misorientation) map obtained by EBSD at 0.2 mm of the inner surface. In the IPF map, lath like fine microstructure is seen more in the sample 15W than in the sample 9W. It can also be seen that plastic strain remains over a wide area in the KAM map of the same field of view. The mean misorientation angle within the field of view measured from the KAM map was 0.75° and 0.54° for samples 15W and 9W, respectively. Figure 3(a) shows an example of Vickers hardness distribution in the thickness direction from the inner surface of the pipe. The Vickers hardness of 15W at the sample surface layer is about 60 HV0.1 higher than that of 9W, and the difference in hardness gradually decreases as the distance from the sample inner surface increases, and the Vickers hardness becomes almost the same at about 8 mm. From this, it can be seen that the hardness distribution of the samples 15W and 9W is different only in the inner surface layer of the pipe. Figures 3(b) and 3(c) show the pipe circumferential hardness distributions of HV0.1 at 0.25 mm and HV10 at 1.0 mm below the pipe inner surface, respectively. A pipe sample of 15W using a steel plate having a high cooling rate exceeding 200°C/sec showed a high hardness exceeding 270 HV0.1, and a tendency of large hardness variation (σ 12) was observed. On the other hand, in the case of the sample 9W at a cooling rate of 100°C/sec, the hardness was relatively low at an average of 250 HV0.1 or less, and the hardness variation was slightly small (σ 10). Table 1 shows the strength characteristic evaluation results of round bar tensile tests of 6 mm diameter taken around the position of 1/4 t of the inner surface and rectangular tensile tests of 5 mm and 1 mm thickness taken from the inner surface of each pipe sample. In the sample 15W having a high inner surface hardness, YS and TS of rectangular tensile test increased as the sample acquisition position was closer to the surface layer. Thus, in the case of 15W and 9W, there was a difference in local mechanical properties in the surface layer of the pipe. In order to clarify the effect of these differences in mechanical properties on SSC, specimens of 5 mm thickness were taken from the inner surface of the pipes after aging, and four-point bend SSC tests and four-point bend polarization measurements were carried out.
SEM microstructure of each cooling rate sample.
IPF map and KAM map from EBSD measurement in each cooling rate sample.
(a) Hardness distribution in thickness direction (Sample 15W and 9W). (b) Surface hardness distribution of Sapmle 15W. (c) Surface hardness distribution of Sample 9W.
| Sample No. | Wall Thickness (mm) | 6 mmΦ round bar Tensile test (1/4 t) | 5 mmt rectangular Tensile test | 1 mmt rectangular Tensile test | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| YS (MPa) | TS (MPa) | uEl (%) | YS (MPa) | TS (MPa) | uEl (%) | YS (MPa) | TS (MPa) | uEl (%) | ||
| 9W | 20 | 592 | 628 | 5.2 | 626 | 653 | 4.2 | 642 | 660 | 4.6 |
| 15W | 30 | 593 | 618 | 5.3 | 669 | 682 | 4.4 | 721 | 739 | 2.3 |
YS: 0.5% Yield Strength (underload), TS: Tensile Strength uEl: Uniform Elongation
In order to evaluate the SSC resistance, especially the SSC crack initiation process from the bottom tip of the corroded pit and groove, the four-point bend SSC test was carried out under the low H2S partial pressure of 0.15 bar H2S in accordance with NACE TM0316 standard.31) The four-point bend test pieces of 5 mm thickness, 15 mm width, and 115 mm length from the inner surface layer of the pipe were subjected to surface 0.2 mm mechanical grinding and collected with the surface finish of #240. In the #240 surface finish, it was confirmed that there was a processed layer with a depth of several μm, which may affect the formation of corrosion pits in the initial stage of a short immersion time. However, in a long immersion time of 720 hours in which corrosion was advanced, relatively uniform and clear corrosion pits were observed,28) and the same #240 finish was used this time. The SSC test conditions are shown in Table 2. NACE TM0177 Buffered Solution32) was used and aimed at a partial pressure of H2S of 0.15 bar, a partial pressure of CO2 of 0.85 bar, and a starting pH of 3.1. Figure 4 shows a schematic diagram of the jig for the four-point bend SSC test. The inner surface side of the pipe was set to the tensile side of four-point bending, and the stress was applied aiming at no load, 324 MPa (72% Specification Minimum YS (SMYS)), 533 MPa (90% Actual YS (AYS)), 622 MPa (105% AYS). The stress value was controlled by a strain gauge (1 mm between marks) attached to the compression surface side. The stress load was set based on the results of the round bar tensile test in Table 1. The test time was set at 3 levels of 7, 168, and 720 hours. The appearance of the tensile side surface after the test was observed, and the central part of the test piece was cut and polished to evaluate the number density of corrosion pits and the existence of SSC cracks. Further, referring to reference,8) it was judged that the one which could be clearly discriminated visually was the one with SSC crack. The hardness (HV0.1) at a position 0.25 mm below the surface layer was measured at a pitch of 1 mm using the same cut specimen, and was evaluated at the maximum value of 31 points. Weight loss after the test was measured to assess the amount of corrosion after the test. The post-test corrosion pit and groove depth was evaluated by measuring the maximum depth Rz value of the laser microscope observation.
| Test Standard | Test solution (NACE TM0177) | pH (start/final) | Partial pressure (bar) | Immersion time (h) | |
|---|---|---|---|---|---|
| H2S | CO2 | ||||
| NACE TM0316-2016 | Buffered solution (5.0 wt%NaCl + 5.0 wt%CH3COOH + 0.40 wt%CH3COONa + H2O) | 3.1/4.0 | 0.15 | 0.85 | 7 168 720 |
| Sample No. | Applied stress (MPa) | |||
|---|---|---|---|---|
| 0 | 72%SMYS | 90%AYS | 105%AYS | |
| 9W | 0 | 324 | 533 | 622 |
| 15W | 0 | 324 | 533 | 622 |
Schematic illustration of four-point bend loading jig. (Online version in color.)
In order to evaluate the effect of material and applied stress on the initial corrosion and hydrogen embrittlement behavior in the four-point bend test, polarization tests were carried out based on NACE TM0177 Method A.32) The sample 15W was used after aging with a high cooling rate exceeding 200°C/sec. A four-point bend specimen of 5 mm thickness from the inner surface layer of the pipe was machined with a surface of 0.2 mm and collected with a surface finish of #240. The specimen surface was coated so that only a certain area was exposed (here, 15 mm square in the center of the tensile surface). The test conditions are shown in Table 3. The stress was applied aiming at no load, 72% SMYS (324 MPa), 90% AYS (533 MPa), and 105% AYS (622 MPa). NACE TM0177 Buffered Solution32) was used and aimed at a partial pressure of H2S of 0.15 bar, a partial pressure of CO2 of 0.85 bar, and a starting pH of 3.1. Polarization was measured for −250 mV (vs. OCP) → +500 mV (vs. OCP) from the OCP (Open Circuit Potential) at a sweep speed of 20 mV/min based on the OCP after 24 hours immersion after 24 hours from the saturation of the test gas. A saturated KClAg/AgCl electrode was used as the reference electrode, and a platinum electrode was used as the counter electrode.
| Test Standard | Test solution (NACE TM0177) | pH (start/final) | Partial pressure (bar) | Immersion time (h) | |
|---|---|---|---|---|---|
| H2S | CO2 | ||||
| NACE TM0316 | Buffered solution (5.0 wt%NaCl + 5.0 wt%CH3COOH + 0.40 wt%CH3COONa + H2O) | 3.1/4.0 | 0.15 | 0.85 | 24 |
| Sample No. | Applied stress (MPa) |
|---|---|
| 15W | 0, 324 (72%SMYS), 533 (90%AYS), 622 (105%AYS) |
In the four-point bend SSC test, corroded pits and grooves are formed after a certain period of time according to the material and applied stress. As a result of stress and strain concentration at the bottom tip, hydrogen diffuses and accumulates, and hydrogen embrittlement cracks seem to occur. In order to understand this behavior, stress distribution analysis and hydrogen diffusion analysis were carried out using ABAQUS 6.12-1. Figure 5 shows a 1/4 finite element model of a notched four-point bend specimen simulating the formation of corroded pit and groove. The test specimen was 5 mm thickness, 15 mm width, and 115 mm length in the same shape as in the four-point bend SSC test, the notch tip radius was set to 0.06 mm, the notch width was set to 0.2 mm, and the notch depth was set to 2 levels of 0.1 mm and 0.2 mm so as to simulate the corrosion pit depth for 168 hours and 720 hours immersion time. Stress-strain curves obtained from 5 mm thickness rectangular tensile test of the sample 15W were applied at 90% AYS (533 MPa, deflection in Y direction 1.14 mm) and 105% AYS (622 MPa, deflection in Y direction 1.33 mm) stresses.
FEM model of four-point bend SSC test with notch simulated corrosion pit. (Online version in color.)
In the calculation of hydrogen diffusion and accumulation behavior, steady hydrogen content (csteady = 0.59 ppm), diffusion coefficient (D = 4.0 × 10−4 mm2/s), and solubility (s = 0.02 ppm) obtained from hydrogen permeation test results under pH 3.0 and 0.1 bar H2S conditions of X65 sour material of the same strength grade were used.33)
The diffusion problem is defined by Eq. (1) of the law of mass conservation for diffusible matter.34)
| (1) |
V represents an arbitrary volume region, S represents the surface of the volume region, n represents the outward normal of S, J represents the concentration flux in the diffusion stage, and nJ represents the concentration flux in the direction away from S.
Diffusion is represented by Fick’s law, and Abaqus’s Fick’s law is provided as a special form of the general chemical potential. Hydrogen diffusion is assumed to be caused by gradients in the general chemical potential, including both concentration gradients and hydrostatic stress gradients.34,35,36,37) In this case, it is defined as shown in Eq. (2).
| (2) |
D is the diffusion coefficient, s is the solubility, c is the mass concentration of the diffusion material, and kp is the pressure coefficient representing diffusion by the gradient of the equivalent stress p. It has been analyzed that the coefficient of the stress gradient term is about 10 times higher than that of the concentration gradient term,35) and a similar coefficient was used here.
Photographs of the appearance of samples 9W and 15W in the range of 40 mm between the marks on the tensile surface after a 720 hours immersion of four-point bend SSC test are shown in Figs. 6(a) and 6(b). In both cases, the direction from left to right in the photograph corresponds to the tensile direction. The formation of corrosion pits was hardly observed in the unloaded and applied stress of 324 MPa. On the other hand, when the applied stress was 533 MPa (90% AYS) or more, many corrosion pits extending in the direction perpendicular to the tensile direction were formed as shown in Fig. 6(c) as an example of the roughness measurement profile result by the laser microscope. In addition, as indicated by the red arrow in Fig. 6(b), it was visually confirmed that SSC cracks were randomly generated in the inside of the specimen and at the width end in the case of the applied stress of 533 MPa or more in the sample 15W. Figures 7(a) and 7(b) show the relationship between the immersion time and the number density of localized corrosion pits under applied stresses of 533 MPa (90% AYS) and 622 MPa (105% AYS) at 15W and 9W, respectively. In the measurement of the number density, a local corrosion pit having a depth of 10 μm or more was used. The number density of localized corrosion was larger in the sample 9W of soft GB main microstructure than in the sample 15W of hard LB main microstructure regardless of the immersion time. The number density tends to be highest at the immersion time of 168 hours, and settles to a constant value at the immersion time of 720 hours. This is because as the immersion time increased, the corrosion progressed so as to erode in the depth direction and in the circumference, and as the fine localized corrosion pits disappeared, the localized corrosion pits grew and coalesced and became coarse. The relationship between the immersion time and the maximum corrosion pit depth measured by laser microscope is shown in Fig. 8 for each applied stress condition. The corrosion pit depth increased with increasing immersion time. In particular, the depth of the corrosion pits in the sample 9W tended to be larger than that in the sample 15W under all conditions. On the other hand, in the sample of 15W, SSC was occurred in the sample subjected to the applied stress of 533 MPa (90% AYS) and 622 MPa (105% AYS) after immersion for more than 168 hours. The maximum values of hardness HV0.1 within 0.25 mm from the surface layer after immersion for 720 hours were 279 and 249, respectively, at the applied stress of 533 MPa (90% AYS) for samples 15W and 9W, and the hardness threshold value for SSC crack initiation was the same as that in the previous report.16) Figure 9 shows the relationship between the applied stress and the local corrosion number density, where the index obtained by dividing the applied stress by YS of the rectangular tensile test of 1 mm thickness in Table 1 (as %YS of the macroscopic surface layer side). In the case of the sample 9W, which is a soft GB-based microstructure, when the applied stress is 533 or 622 MPa, compared with the sample 15W, which is a hard LB-based microstructure, the surface tensile stress is relatively high, and when the surface tensile stress is applied, there are many microscopic plastic strain concentrated places caused by the distribution of the microhardness, structure or crystal orientation as shown in Figs. 2 and 4,38,39) and it is considered that local corrosion was easy. On the other hand, in the sample 15W having a hard LB main microstructure, the frequency of occurrence of local corrosion is extremely low, especially when the applied stress is 533 MPa, and the frequency of occurrence of local corrosion increases when the applied stress is increased to 622 MPa, but the frequency of occurrence of local corrosion is less than that of the sample 9W. Besides the microscopic plastic strain concentration, it is considered that the localized corrosion is caused by the high dislocation density of the hard LB microstructure matrix.
Effect of applied stress on surface corrosion and SSC behavior after 720 h immersion.
Effect of immersion time and applied stress (90%AYS and 105%AYS) on the number density of corrosion pits.
Effect of immersion time and applied stress on maximum corrosion pit depth after four-point bend SSC test. (Online version in color.)
Relationship between the actual applied stress level on the tension side and the number density of corrosion pits. (Online version in color.)
In the hard sample of 15W, under the applied stress of 622 MPa, SSC was occurred in the sample with the immersion time of 168 hours and 720 hours, and the crack depth was 1.4 mm and 4.7 mm, respectively. In addition, under the applied stress of 533 MPa, SSC was occurred only in the sample with the immersion time of 720 hours, and the crack depth was 0.57 mm. The SSC crack depth increased with increasing stress and immersion time. The increase in the crack depth with increasing immersion time suggests that the SSC crack propagation process is time-dependent. On the other hand, although corrosion pits were remarkably formed in the soft sample 9W, no SSC occurred. Figure 10(a) shows the optical microstructure of the pit formation surface layer side and the crack of the sample in which SSC was occurred after the test of the hard sample of 15W. In the pits on the surface layer side, the right and left microstructures sandwiching the pits did not correspond due to corrosion melting, while in the case of cracks observed in the region deeper than 200 μm from the surface layer, the right and left microstructures corresponded from the initial stage of formation to the crack tip, and the cracks developed in the grain, not in the grain boundary. In addition, many linear cracks peculiar to pseudo-cleavage cracking were observed, and some cracks branched and fractured, suggesting that the crack initiation and propagation mechanism is a hydrogen embrittlement mechanism. From the above results, it can be seen that after the corroded pit and groove is formed at a depth of about 200 μm from the surface layer, the corrosion pit and groove transitions to an SSC crack from the bottom tip of the corroded pit and groove and develops into the inside of the sample. On the other hand, as Fig. 10(b) shows the optical microstructure of the corrosion pit tip of the soft sample 9W in which no SSC crack was generated after the test, it branched and developed in a plurality of directions at the pit tip and the corrosion melting region expanded, and no transition to an SSC crack as observed at 15W was observed.
Detailed observation of surface region (Sample 15W and 9W, 90 AYS-720 h). (Online version in color.)
In the hard sample of 15W, SSC cracks were generated by immersion in a short period of time because the applied stress increased from 533 MPa (90% AYS) to 622 MPa (105% AYS), and the amount of crack propagation greatly increased. It is presumed that this is because hydrogen embrittlement was accelerated due to the increase of stress and strain caused by stress concentration at the bottom tip of the corroded pit and groove. That is, in the sample 15W having the LB main microstructure with higher surface hardness than the sample 9W, the stress concentration at the bottom tip of the corroded pit and groove increased, the tensile stress acting locally in the opening direction increased. It is considered that the hydrogen embrittlement crack occurred because hydrogen diffused and accumulated easily and the limit amount of hydrogen required for cracking was exceeded. On the other hand, it is considered that, in the sample 9W having a soft GB-based microstructure, the stress concentration was small because the applied stress was small even if the localized corrosion pits were formed due to the advanced corrosion dissolution of the microscopic plastic field, and the limit amount of hydrogen required for cracking at the bottom tip of the corrosion pits and grooves was not exceeded, and the hydrogen embrittlement crack did not occur. There are many unclear points about the fracture mechanism of hydrogen embrittlement, especially the mechanism of hydrogen embrittlement when the partial pressure of H2S in the wet sour environment is changed. As for the role of hydrogen, there are lattice embrittlement theory and plastic deformation acceleration theory (HELP, HESIV, etc.), but recently the latter HESIV theory is particularly influential, and it is understood that hydrogen promotes the formation of atomic vacancy damage (nanovoid) accompanying plastic deformation.40,41,42,43,45,46) The fracture mechanism of SSC crack initiation samples observed in this study is described in detail in the next section.
3.1.2. SEM Observation of Fracture Surface of SSC CrackFigure 11(a) shows the cross-sectional observation results of the SSC crack generated under the condition of the sample of 15W under the applied stress of 622 MPa (105% AYS) and the immersion time of 720 hours. Fracture surfaces were observed at 0.1 mm, 0.2 mm, 0.5 mm, and 1.0 mm from the surface layer, as shown in Figs. 11(b), 11(c), 11(d), and 11(e), respectively. Up to about 0.2 mm in the surface layer, there were many concave and convex regions partially corroded and dissolved in a network shape in the dimple fracture surface subjected to plastic deformation, which is a characteristic of ductile fracture surface. On the other hand, at the positions of 0.5 mm and 1.0 mm in the surface layer, which advanced beyond 0.2 mm in the surface layer, no area with corrosion-dissolved concave and convex depressions was found, and a dimple fracture surface subjected to plastic deformation and a flat quasi-cleavage fracture surface were mixed. It is considered that this change of fracture surface morphology corresponds to the fact that the formation of corrosion pit and groove in the surface layer is controlled by the APC mechanism, while the formation of SSC crack is controlled by the HE mechanism, and corresponds to the fact that the main force of the reaction from the APC mechanism to the HE mechanism was shifted in the process of the generation of SSC crack from the bottom tip of corrosion pit and groove. However, as shown in Figs. 11(f), 11(g), and 11(h), traces of corrosion products containing a large amount of S were observed by EDX analysis in the relatively flat area of the fracture surface at both the surface layer 0.5 mm and 1.0 mm, indicating that the influence of APC remained partially after the occurrence of the SSC crack. The relatively low H2S concentration of 0.15 bar is considered to influence the thin and inhomogeneous FeS film formation16) and the small amount of penetrating hydrogen.
(a) Cross-sectional observation of SSC crack, (b), (c), (d), (e) Fracture surface observation by SEM, (f), (g), (h) EDX analysis. (Online version in color.)
It is considered that the APC mechanism is promoted by plastic deformation, and that the exposure of the newly formed surface and the increase of dislocation density accompanying slip deformation accelerate the anodic reaction of Fe dissolution.28,44) On the other hand, as mentioned in the previous section, the HE mechanism is understood as hydrogen promotes the formation of atomic vacancy damage (nanovoid) accompanying plastic deformation.40,41,42,43,45,46) It is considered that the local plastic instability (formation and connection of nanovoids) is promoted by hydrogen accumulation with the stress field expansion by the local stress concentration including plastic deformation.45) The fact that the SSC crack initiated only in the sample 15W with the hard LB microstructure suggests that hydrogen diffused and accumulated in the region where the stress increased due to the stress and strain concentration at the bottom tip of the corroded pit and groove, and when the amount of hydrogen exceeded a certain value, the local plastic instability condition was satisfied in the presence of hydrogen. As a result, nanovoids were formed and coalesced, leading to the initiation of HE-dominated cracks. The experimental verification on the nanovoids formation of HE mechanism in the H2S sour environment will be further advanced in future. Next, in order to examine the relationship between the stress concentration and the hydrogen accumulation, the stress-strain state and the hydrogen accumulation at the notch bottom simulating the corrosion pit and groove are evaluated by the FEM calculation when the depth of the corrosion pit and groove is assumed to be 0.1 mm (168 hour immersion time) and 0.2 mm (720 hours immersion time).
3.2. Polarization Characterization of Four-point Bend SSC Test and FEM Calculation of Hydrogen Diffusion and AccumulationFigure 12 shows the results of the effect of applied stress on anode and cathode reaction activation in the initial stage of sample 15W. In the sample 15W, the reaction was activated in both anode and cathode when there was stress compared with no stress. The higher the applied stress, the more remarkable it was. The activation of the anodic reaction with increasing applied stress is considered to be due to the exposure of the active newly formed surface. The activation of the cathodic reaction with the increase in the applied stress is considered to be due to the effect that the surface irregularity increased and the reaction area increased by the increase of the microscopic plastic deformation of the tensile surface. The effect of hydrogen overvoltage drop is also considered. Consequently, the corrosion current density also showed a high value. The increase in corrosion current density with the increase in stress corresponds well to the increase in corrosion loss due to stress application in the four-point bend SSC test.
Polarization curve in four-point bend SSC test (Sample 15W). (Online version in color.)
The polarization measurement result indicates that when a stress of 90% AYS or more is applied, the cathodic reaction is activated with the concentration of microscopic stress and strain, and hydrogen generation is promoted. This fact suggests that the SSC initiated under the stress condition of 90% AYS or more in the four-point bend SSC test of the sample 15W is the phenomenon driven by hydrogen embrittlement. It corresponds well with the SEM observation result (Fig. 11(e)) which recognized the localization of plastic deformation and pseudo-cleavage fracture surface which are features of hydrogen embrittlement.
In the four-point bend SSC test, corroded pits and grooves are formed after a certain period of time according to the material and applied stress. As a result of stress and strain concentration at the tip, hydrogen diffuses and accumulates, and hydrogen embrittlement cracks seem to occur. In order to understand this behavior, stress distribution analysis and hydrogen diffusion analysis were carried out using ABAQUS 6.12-1.
Figure 13(a) shows the distribution of hydrogen concentration in the notch depth direction after 24 hours obtained by FEM. Calculation of the time variation of hydrogen concentration under the conditions of 0.1 mm and 0.2 mm notch depth, 533 MPa (90% AYS) and 622 MPa (105% AYS) applied stress of sample 15W showed that hydrogen diffusion was saturated at 24 hours. In the case where the notch depth is 0.1 mm and 0.2 mm, the maximum value of hydrogen content is shown at the positions of depth 0.14 mm and 0.26 mm which are slightly inside of the notch bottom, respectively. The deeper the notch and the higher the applied stress, the higher the hydrogen accumulation. Figures 13(b) and 13(c) show the relationship between the distance from the notch bottom and the principal stress in the tensile direction and the equivalent plastic strain when the notch depth is 0.1 mm and 0.2 mm, respectively. The equivalent plastic strain reached a maximum at the notch bottom, and the principal stress reached a maximum at 0.14 mm and 0.26 mm. These correspond to the places where the hydrogen content is maximum. In comparison with the experimental results in Section 3.1, in a sample of 15W immersed for 720 hours, localized plastic deformation just below the bottom of a corroded pit and groove of about 0.2 mm corresponds to the occurrence of hydrogen embrittlement cracks (Figs. 10(a), 11). It is inferred that the occurrence of the hydrogen embrittlement crack starts from the position of the maximum equivalent plastic strain just below the notch bottom, and the increase of hydrogen accumulation in the principal stress intensity region within the range about 0.06 mm away from the notch bottom leads to the generation and growth of the secondary crack, leading to the crack propagation.
(a) Effect of notch depth and applied stress on hydrogen concentration and (b), (c) Distribution of principal stress and equivalent plastic strain at the notch tip calculated by FEM. (Online version in color.)
Figure 14 shows the relationship between the maximum principal stress in the tensile direction and the maximum hydrogen accumulation. From the experimental results in Section 3.1, SSC cracks were generated only in the sample of 15W of the hard LB main microstructure. It was found analytically that SSC cracks were generated and developed in the HE mechanism controlled type when the threshold values of about 0.82 ppm of hydrogen were exceeded in the case of a 168 hours immersion simulated notch depth of 0.1 mm and 622 MPa of applied stress, about 0.88 ppm of hydrogen was exceeded in the case of a 720 hours immersion simulated notch depth of 0.2 mm and 533 MPa of applied stress.
Effect of maximum principal stress on maximum hydrogen concentration. (Online version in color.)
Based on the above results, it is estimated that in the four-point bend SSC test, corroded pits and grooves are formed after a certain period of time depending on the material and applied stress, and as a result of stress and strain concentration at the tip, hydrogen diffuses and accumulates, and hydrogen embrittlement crack initiated. Figure 15 schematically shows the difference in behavior of corrosion pit and groove formation, SSC crack initiation and propagation between a sample 15W with hard LB microstructure and a sample 9W with soft GB microstructure. As shown in Fig. 9, in the sample 15W of hard LB, the tensile test result of the surface layer 1 mm thickness shows macroscopic elastic region, but local stress and strain concentration occurs at the bottom tip of corroded pit and groove. As shown in Fig. 13(c), plastic deformation of about 3% occurs at the surface, and SSC crack driven by HE mechanism is initiated. As a result of hydrogen diffusion and accumulation in the principal stress intensity region within the range about 0.06 mm away from the bottom tip of corroded pit and groove, secondary crack generation and growth are induced, and SSC crack propagated. On the other hand, in the sample 9W having a soft GB main microstructure, the plastic strain at the bottom tip of the corroded pit and groove similarly increases and the corrosion progresses. However, the plastic strain at the bottom tip of the corroded pit and groove and the hydrogen accumulation in the stress expansion region are not sufficient. Therefore, the SSC crack controlled by the HE mechanism does not occur. Figure 16 shows the relationship between the true stress true strain curve and the work hardening rate of the rectangular tensile test of 1 mm for samples 15W and 9W. The sample 15W with hard LB satisfies the plastic instability condition σ≧dσ/dε earlier than the sample 9W with soft GB, and ductile fracture occurs at a lower strain when stress is increased. Although these are the test data in the air, it is considered that the local plastic destabilization was further promoted under the sour environment with hydrogen penetration in the steel,16,46) and that SSC crack driven by HE mechanism was more likely to occur in the sample 15W even at the same plastic strain amount compared with the sample 9W. The crack initiation at the bottom tip of the corroded pit and groove and the coupling process of hydrogen accumulation to secondary crack about 0.06 mm away from it need to be studied in more detail. The balance between the increase of plastic strain at the bottom tip of the corroded pit and groove and the increase of hydrogen accumulation with stress concentration will determine whether or not nano-void coalescence, coupling and final crack initiation by local plastic destabilization will occur. In the sample of 15W with high hardness on the inner surface side of the pipe, under the environment in which hydrogen penetrated into the steel, local plastic destabilization was promoted and nano-void was easy to form, resulting in crack connection and crack propagation. As described in Section 3.1.1, for example, in the case of 720 hours immersion under the applied stress of 533 MPa, it is considered that this led to large crack propagation at a depth of 0.57 mm.
Comparison between sample 15W and 9W regarding the corrosion groove formation and SSC crack initiation mechanism.
Comparison of true stress strain curve and work hardening rate.
In this study, the effect of various factors (Material surface hardness, applied stress, immersion time) on SSC crack initiation in a low alloy steel line pipe in a low H2S partial pressure environment of 0.15 bar was investigated by means of four-point bend SSC test. The main conclusions are as follows.
(1) SSC susceptibility was affected by hardness (material surface hardness and microstructure), applied stress, and immersion time, and SSC occurred in samples with high surface hardness. Specifically, SSC cracks initiated when the hardness was 270 HV0.1 at 0.25 mm or more and the stress was 90% AYS or more, whereas SSC cracks did not initiate when the hardness was 270 HV0.1 or less or the stress was 72% SMYS.
(2) From the crack morphology of the cross section and the observation results of the fracture surface, in the sample 15W mainly composed of hard LB, after formation of corroded pit and groove due to iron dissolution of APC mechanism, the main mechanism shifted to HE mechanism at the bottom tip of corroded pit and groove, and SSC crack initiated. However, it was suggested that there was some contribution of APC after the crack initiation. On the other hand, in the sample 9W mainly composed of soft GB, only the iron dissolution by the APC mechanism was observed, and the crack initiation driven by the HE mechanism did not occur.
(3) As a result of the polarization characteristic evaluation of four-point bend SSC test, in the case of LB main microstructure of high hardness and applied stress over 533 MPa (90% AYS), the initial anodic and cathodic reactions were greatly activated, especially the cathodic reaction. It was shown that not only APC iron dissolution but also HE hydrogen embrittlement was greatly promoted.
(4) As a result of FEM coupled analysis simulating stress concentration and hydrogen diffusion and accumulation at the bottom tip of corroded pit and groove, it was found analytically that SSC crack initiated and propagated in the HE mechanism dominant type, when the threshold value of hydrogen content of about 0.82 ppm was exceeded, in which the equivalent plastic strain quantity showed the maximum value at the bottom tip of corroded pit and groove, and the principal stress showed the maximum value at a place about 0.04–0.06 mm away from the tip, and hydrogen accumulation quantity became the maximum.
(5) It is considered that, in the sample 15W mainly composed of the hard LB microstructure, at the bottom tip of corrosion pit and groove formed by the APC mechanism, the mechanism shifted to the HE mechanism mainly with the increase of local stress and strain concentration and plastic destabilization was promoted, resulting in the SSC crack initiation.
