Minggu, 02 Februari 2014

Relationship Between Yield Strength and Near-Neutral pH Stress Corrosion Cracking Resistance of Pipeline Steels-An Effect of Microstructure

In this paper the relationship between the near-neutral pH stress corrosion cracking (SCC) resistance and the yield strength of pipeline steels was investigated and an attempt was made to understand the microstructural effect on such a relationship. Pipeline steels ranging from X52 to X100 steels and the weldments of X70 and X65 were adopted as the test materials and various heat treatments were used to achieve different microstructures and strength levels. The results indicate that the near-neutral pH SCC resistance of pipelinesteel is reduced, generally, with an increase in the strength level, but the strength dependence of SCC resistance is heavily affected by the microstructures of the pipeline steels. The steels with a fine-grained, bainite-ferrite structure possess a much better combination of strength and SCC resistance than those with a ferrite + pearlite structure. However, the introduction of the welding process will significantly degrade SCC resistance in the steels containing a bainitic ferrite structure. This degradation effect is caused mainly by the decomposition of the bainitic ferrite structure into a separate microstructural entity. On the other hand, an increase in the pearlite content in the microstructure has a detrimental effect on the SCC resistance of pipelinesteels with a ferrite + pearlite structure. The experimental results indicate that the SCC resistance of thepipeline steels in the near-neutral pH environment can be approximately correlated to the polarization resistance with a linear relation. This relationship is used to evaluate the microstructure effect of weldments on the SCC resistance. The applicability of this method is discussed briefly.
INTRODUCTION
The near-neutral pH stress corrosion cracking (SCC) of pipeline steels occurs at the external surface ofpipelines in dilute groundwater with a pH value close to 6.5.1-3 The experimental evidence has indicated that hydrogen charging and cyclic loads promote the SCC process. The substantial dissolution is observed at trie walls of transgranular cracks.4-6 Although it is commonly agreed that both the hydrogen dissolved in steels and anodic dissolution present on the surface contribute to the mechanism of the SCC process,1,2,4-5,7 the adequate mechanism governing the occurrence of SCC is yet to be clearly understood. On a laboratory scale, the near-neutral pH SCC can only be reproduced under cyclic or dynamic loading conditions instead of static loading,8 and the SCC is more likely to initiate at the edge of specimens where the restriction of the plastic deformation is minimum.9 This finding acknowledges the important role of dynamically microplastic deformation in SCC development.2-3,10-11 Based on the fact that the stress needed to initiate plastic deformation at the surface layer of the pipe is lower than that needed for the bulk material,12 SCC can initiate at stress levels well below the yield strength of steels. It was expected1,3 that the cyclic softening character of pipeline steel would influence the SCC initiation, but a relationship between the cyclic stress-strain behavior and the SCC susceptibility of pipelines still has not been established.13

Pipe failure as a result of near-neutral pH SCC was detected in pipes with diameters ranging from 114 mm to 1,067 mm, wall thicknesses ranging from 3.2 mm to 9.4 mm, and strength grades ranging from 241 MPa (35 ksi) to 448 MPa (65 ksi).1 Both electric resistance-welded and double-submerged arc-welded pipes were involved in the SCC-related failure. Near-neutral pH SCC behavior is strongly affected by the metallurgical features ofpipeline steels, such as chemical composition, microstructure, mechanical properties, and surface conditions.1,9,14-16 In the SCC process, the cracks are likely to initiate at the locations where inclusions exist.1,6,17-19 CANMET (Ottawa, Ontario, Canada) examined the inclusion lengths in samples cut from ten pipes. Out of these samples, five were found with significant cracking while the last five showed no significant cracking. The conclusion was made that the pipes with significant cracking had larger inclusions than those with nonsignificant cracking.20
Beavers, et al.,13 analyzed 14 used pipelines on which SCC colonies had been detected. The grades of the pipes analyzed were API X52, API X60, API X65, and API X70.(1) They found that SCC was likely to be found at regions near the weld seams of pipes, and the micro-hardness in SCC zones on the pipes was slightly higher than that in the non-SCC zones. Depending on the welding processes, the microstructure of the weld metal and heat-affected zone (HAZ) may be quite different from that of the base metal. Unfortunately, the relationship among the microstructure, mechanical properties, and SCC resistance of pipeline steels is still not very clear.2,20
In engineering applications, strength is one of the most important material properties. A huge effort has been made to raise the strength of steels with improvement in metallurgical technologies. The new generation ofpipeline steels, still in the developmental stage, possesses fine-grained bainite or bainite + ferrite microstructures, and higher strength and high toughness in the as-hot-rolled condition.21 However, a thorough understanding of SCC performance of steels with such microstructure is far from being clarified.9,14-15,17-20 In the current investigation, the relationship between the near-neutral pH SCC resistance and the yield strength of pipeline steels, as well as the microstructural effect on such a relationship, is to be investigated.
EXPERIMENTAL PROCEDURES
Test Materials and Heat Treatments
The test materials were API X52, API X60, API X65, API X70, API X80, and API XlOO steels, and their compositions are listed in Table 1. Under hot-rolled conditions, the microstructure of the X52, X60, and X65 steels was a ferrite + pearlite structure. Owing to the adoption of techniques of micro-alloying and controlled rolling, the X70, X80, and XlOO steels possessed a fine-grained bainite + ferrite structure. The microstructures of test materials under as-rolled conditions are shown in Figure 1. One set of specimens was heat-treated to achieve different microstructures. The heat-treatment processes and microstructures obtained are listed in Table 1 and the typical microstructures are shown in Figure 2. To investigate the effects of welding, the weldments of API X65 and API X70 steels were also used in the experiments. The welding process for the X70 steel was submerged arc welding and the process for the X65 steel was electric resistance welding. The base metals (BM) of both steels were under as-rolled conditions. Figures 3 and 4 show the microstructures of weld metals and HAZ of X70 and X65 pipeline steels. The microstructure of the weld metal of X70 steel (X70W) is basically dendritic ferrite (Figure 3[a]), and that of X70 in HAZ (X70HAZ) is mainly the ferrite + pearlite (Figure 3[b]), respectively. For the welded X65 pipeline steel, the different microstructures are produced in the weld metal of the first fusion run at the inner side of the pipe (X65WM1) and the second run at the outer side of the pipe (X65WM2). The X65WM2 displays a dendritic ferrite structure (Figure 4[a]) but the dendritic structure disappeared in the X65WM1 (Figure 4[b]) because of heating by the second run of the welding gun. As shown in Figure 4(c), the Wiedmanstantten structure could be found in the HAZ of X65 steel (X65HAZ).
The changes in microstructure in the welded zones will affect the mechanical properties of materials. The micro-hardness distribution over the welded zones was used to demonstrate such an effect. The Vickers hardness of the samples (Hv) was measured with a hardness tester and an indenting load of 2 kg was adopted for the measurement. The hardness maps for the two welds investigated are depicted in Figure 5. Little difference in the hardness can be found in the weld metal (WM) of X70 steel that formed during the first and second runs of the welding gun. The high-to-low order for the welded pipe was WM > BM > HAZ. The high hardness in the dendritic ferrite structure of WM may relate to the more defective microstructure produced by the fast cooling during solidification, and the low hardness of HAZ is due to the decomposition of the bainitic structure. The hardness order for the welded X65 is WM2 > BM > WM1HAZ. The low hardness in HAZ and WM1 may be related to the heating process during welding, which acts as a normalizing treatment, resulting in a decrease in the crystal defects.
Slow Strain Rate Tensile Tests

The configuration of specimens used in the slow strain rate tensile (SSRT) test is illustrated in Figure 6(a). To investigate the effect of microstructure in the welded zones on the SCC resistance, the specimens to be machined for the SSRT tests were arranged one-by-one parallel to the welding seam, as shown in Figure 6(b). Only the weldment of X70 steel was used to conduct the SSRT tests. The welded X65 steel was not used because its hardness distribution over the welded zones was nonuniform, which caused extra difficulty in machining a specimen with a relatively uniform microstructure in gauge length.
The tests were conducted with a materials test system (MTS) test machine under the strain rate of 10^sup -7^ s^sup -1^. Following the approach of Parkins, et al.,2-3 the ratio of the reduction in area (RA) in solution to that in air, RA^sup SCC^/RA^sub Air^, of tensile specimens is defined as the SCC resistance of material, where RA^sub SCC^ and RA^sub Air^ are the values of reduction in area measured by the SSRT in the corrosive environment and in air, respectively. The test solution is a NS4 solution saturated with nitrogen + 5% carbon dioxide (CO2). The composition of the NS4 solution in g/L was 0.122 potassium chloride (KCl), 0.483 sodium bicarbonate (NaHCO^sub 3^), 0.137 calcium chloride (CaCl), and 0.131 magnesium sulfate (MgSO^sub 4^.7H2O). The SSRT tests commenced after bubbling 95% N^sub 2^/5% CO2 into the test solution for more than 1 h to create an anaerobic environment and to create a pH value around 6.7. The anaerobic condition and pH value were held until the test finished.
Polarization Resistance Measurements
Prior to each corrosion test the epoxy-mounted samples were ground with 400-grit and 600-grit silicon carbide (SiC) papers successively, rinsed with deionized water, and degreased with acetone (CH^sub 3^COCH^sub 3^). The polarization resistance of the materials was measured in a conventional three-electrode electrochemical cell, using a corrosion measurement system. A long platinum wire was used as the counter electrode, and a saturated calomel electrode (SCE) was used as the reference electrode. The test solution was the NS4 solution saturated with 95% N^sub 2^/5% CO2. In the polarization resistance (R^sub p^) measurement, the potential range was from -10 mV to 10 mV relative to the open-circuit potential (E^sub corr^), and the potential scanning rate was 0.1 mV/s. The polarization resistance is the slope of the potential vs. current curve near open-circuit potential.
RELATIONSHIP BETWEEN STRESS CORROSION CRACKING RESISTANCE AND YIELD STRENGTH

The microstructure of X70W (weld metal) formed directly from the liquid metal is mainly the dendritic ferrite. The microstructure of X70 in the HAZ has been transferred from bainitic ferrite into a ferrite + pearlite structure. The effect of welding on the SCC resistance of pipeline steel is well illustrated by the tensile curves of welded X70 steel obtained by the SSRT shown in Figure 7. The relative degradation in maximum strain causing specimen failure (ε^sub f^), which is characterized by ε^sub f(SCC)^/ε^sub f(Air)^, indicates clearly that both the WM and HAZ are more sensitive to the near-neutral pH SCC than the BM is. Among them the WM possesses the worst resistance to SCC. Note that the stress state in the SSRT tests was quite different from the service condition and a large plastic deformation was involved, and even a strain rate as low as 10^sup -7^ s^sup -1^ was adopted. The experimental evidence20 indicated that the crack growth rate ofpipeline steel was almost independent of the strain rate when it was under the loading rate in the order of 10^sup -7^ s^sup -1^. As indicated by the scanning electron microscopy (SEM) fractograph, the fracture surface of WM possesses cleavage-like characteristics mixed with shallow dimples (Figure 8[a]) and many secondary cracks (Figure 8[b]). On the fracture surface of BM or HAZ, except for the transgranular SCC appearance in some areas (Figure 8[d]), many cavities produced by the micro-plastic deformation can be observed (Figure 8[c]). The fracture appearance in Figure 8 is somewhat different from that of failed pipes, resulting from the near-neutral pH SCC.20 The plastic deformation in the SSRT test might obscure the roles of metallurgical factors in the SCC mechanism. To understand the near-neutral pH SCC behavior of the pipelineswell, it is more reasonable to evaluate the SCC resistance of pipeline steels under low-frequency cyclic loading conditions.1-3,8-9 However, this kind of experiment is very time-consuming. Recent experimental research on the SCC behavior of welded X70 steel was conducted in our laboratory with low-frequency cyclic loads (8 cycles/day).22 The results indicated that the changes in microstructure introduced by the welding process displayed a significant impact on the near-neutral pH SCC behavior of X70 steel, especially on the resistance to crack initiation. The high-to-low order of the crack initiation lifetime of the welded X70 steel was BM > HAZ > WM.22 These results agree with those reported in the present study, as shown in Figures 7 and 9. It indicates that the test data of SSRT can still be used to evaluate the microstructural effect on the SCC resistance ofpipelines, although some uncertainty exists. The low SCC resistance of weld metal may relate to its dendritic ferrite structure, which is detrimental to the ductility of steel.

The correlation between the RA ratio and the yield strength of pipeline steels is shown in Figure 9. Among the materials involved in this figure, X70, X80, and XlOO have fine-grained, bainitic ferrite structures and X52, X60, and X65 have ferrite + pearlite structures under as-rolled conditions. Various microstructures were achieved in the X70 steel under different heat-treatment conditions. The microstructure under annealed and normalized conditions (X70A and X70N) is coarse-grained ferrite + pearlite,23-24 while under the water-quenched condition (X70WQ) and water-quenched + high-temperature temper (X70WQ+HT), they are mainly coarse-grained bainite and coarse-grained, polygonal ferrite + fine carbide precipitation,23"24 respectively, as shown in Figure 2. The data in Figure 9 clearly show that the SCC resistance of pipeline steels decreases, generally, with increasing yield strength provided that the steel shows similar microstructure. However, the correlation between the SCC resistance and yield strength is microstructure-dependent. In view of the combination of strength and SCC resistance, steel with a fine-grained bainite + ferrite structure is better than one with a microstructure of ferrite + pearlite. Although the annealed steel displays the highest SCC resistance, its strength is too low for practical application. The microstructural effect observed in Figure 9 agrees well with the SCC initiation lifetimes of X70 and X65 steels measured under the low-frequency cyclic loading conditions.6,9,17 Although the strength levels of these two steels are quite close, the SCC initiation lifetime of X70 was found to be several times that measured from X65 steel, owing mainly to the difference in the microstructure.9 The same conclusion was also reported by a committee of The Iron and Steel Institute of Japan25 and it was based on the test data of the threshold cyclic stress to initiate cracks from notches, which were collected from a range of pipelines, from X52 to X80 grades under thermomechanical-controlled processing and heat-treatment conditions. As demonstrated in Figure 1, under the as-rolled conditions, the X65 has a typical ferrite + pearlite structure but the X70 has a bainitic ferrite structure. However, if the microstructure of X70 steel transfers into a ferrite + pearlite structure, like that occurring in the HAZ, the SCC resistance will degrade remarkably, although the yield strength of the material decreases simultaneously. In the view of practical engineering, the welded seam can reduce the local stress level in weld metal, as does the possibility for the occurrence of near-neutral pH SCC, since the crack initiation lifetime increases with the decreasing level of local stress. However, the local stress level at the HAZ is much higher than that at the other parts in pipelines, owing to the stress concentration at the weld toe. If exposed to the corrosive environment, the HAZ is more likely to initiate SCC. So, it is of practical importance to improve the SCC resistance of HAZ by optimizing the welding processes.

The data in Figure 9 also indicate that the bainitic structure obtained in X70 steel from the water-quench is less resistant to SCC, compared to material under the as-rolled conditions, although water-quenching treatment raises the yield strength. In fact, the formation of pearlite in X70 steel is suppressed by water-quenching and only a small amount of fine carbides precipitate in the microstructure during the temper treatment. As shown in Figure 9, the temper after the water-quenching can improve the SCC resistance. The same phenomenon was also observed in X80 steel.16 It suggests that the high susceptibility of steel under the water-quenched condition might result from the high micro-stresses produced by the excess carbon trapped interstitially.26 X70WQ+HT has a coarse-grained polygonal ferrite structure (Figure 4) but SCC resistance is close to, or a little better than, that of the same steel with fine-grained bainitic ferrite, implying that the observed poor SCC resistance of steels with a ferrite + pearlite structure may relate to the pearlite in the microstructure.
EFFECT OF MICROSTRUCTURE

Because of the limited size and complicated microstructure of welding zones, sometimes it is difficult to prepare the tensile specimens with identical microstructures in their gauge length, especially when the HAZ is involved. A fast screen test is designed to evaluate the effects of metallurgical factors on the near-neutral pH SCC resistance of pipeline steels, based on the test results reported by Bulger and Luo.2324 These data are depicted in Figures 10 and 11, where relationships of the yield strength vs. the Vickers hardness and the RA ratio (RA^sup SCC^/R^sub Air^) vs. the polarization resistance (R^sub p^) ofpipeline steels in anaerobic NS4 solution are displayed. The linear relationship between hardness and yield strength of steels has been well recognized.27 A recent study showed that, around the open-circuit potential, both the ingress of hydrogen and anodic dissolution promoted the development of near-neutral pH SCC, and a synergism of hydrogen- and anodic dissolution-promoted plasticity might be the SCC mechanism.28 Since the reaction of anodic dissolution is always balanced by the cathodic reduction of hydrogen in the near-neutral pH SCC, any factor enhancing the anodic dissolution may increase the likelihood of hydrogen ingress into the steel. Because the anodic dissolution rate at the open-circuit potential is reversed proportional to the polarization resistance, the linear relationship between the RA ratio and polarization resistance shows that anodic current plays a role in the near-neutral pH SCC around the open-circuit potential, suggesting that the ingress of hydrogen, the anodic dissolution, and the synergy between the two may control the near-neutral pH SCC. There is some experimental evidence that the SCC susceptibility of pipeline steels in near-neutral pH soil environments increases with the corrosiveness of environments, such as an increased CO2 level and a decrease in pH.2-3,29 In light of the model of corrosion-enhanced plasticity, anodic dissolution will promote the SCC but the crack growth rate cannot be formulated with Faraday's law. Based on the correlations shown in Figures 10 and 11, the relationship of hardness vs. polarization resistance is used in the present investigation to evaluate the microstructural effects on the SCC resistance of test materials.
Correlation Between Polarization Resistance and Hardness of Pipeline Steels Under As-Received and Welded Conditions

Figure 12 shows the correlation between the polarization resistance and hardness under as-received conditions. The results indicate that, in general, the polarization resistance decreases with an increase in hardness for pipeline steels with similar microstructure. If two steels have the same yield strength, steels with a bainitic structure are more resistant to near-neutral pH SCC than those with a ferrite + pearlite structure. The weld metal of X70 possesses poor SCC resistance. These results agree with the results obtained from the SSRT experiments mentioned above, indicating that the correlation of R^sub p^ vs. Hv can be approximately used as a tool for the preliminary evaluation of near-neutral pH SCC, especially for the materials with limited size, such as the HAZ of welded pipelines.
The SCC resistance of the pipeline steels is sensitive to welding. Such an effect is caused virtually by the changes in microstructure. According to the hardness measurements, for the welded X65 steel, the weld metal produced in the second fusion run (X65WM2) was harder than that produced in the first fusion run (X65WM1). Because the heating process of the second fusion run may play a role as a normalization treatment, it reduces the hardness but improves the corrosionresistance. This result implies that the SCC resistance of weld metal can be improved by a postwelding heat treatment. However, for the welded X70 steel, the morphology of the microstructure and the hardness of the weld metal are found to be almost the same in the zones of different fusion runs, suggesting a relatively higher stability of the dendrite in the welded metal. Figure 12 shows that the relationship of R^sub p^ vs. Hv in the weld metal follows a pattern different from that in the base metals because of the significant difference in microstructure. If the hardness is held unchanged, the values of R^sub p^ in the weld metals are lower than that of the base metals. This simply indicates that the pipeline steels with dendritic ferrite structures are less resistant to near-neutral pH SCC than those under as-rolled conditions.
Although the pipeline steels with a bainitic ferrite structure possess a higher SCC resistance than the steels with a ferrite + pearlite structure if both structures have the same yield strength, as shown in Figures 9 and 12, its SCC resistance is more sensitive to welding. Based on the fact that bainitic ferrite is a metastable phase, it is likely to decompose during the welding process, as it occurs in the HAZ of X70 steel. Such a change in microstructure will cause the degradation in SCC resistance, as demonstrated by the low polarization resistance and low hardness of the HAZ. Figure 12 shows that the polarization resistance of X70HAZ is close to the trend line of R^sub p^ vs. Hv for the steels with the microstructure of ferrite + pearlite. This result agrees again with the test data of SSRT tests in Figure 9.
In the HAZ of X65 steel close to the weld metal, a Wiedmanstatten ferrite structure is observed, as shown in Figure 4(b), and it is quite different from the microstructure of the base metal. The test result in Figure 12 indicates that the polarization resistance of the HAZ is lower than that of BM, and the data point is close to the trend line of R^sub p^ vs. Hv for the weld metal. In accordance with the discussion above, the SCC resistance of the HAZ can be improved by eliminating the Wiedmanstatten ferrite structure.
Correlation Between Polarization Resistance and Hardness of Heat-Treated Pipeline Steels
In line with Figures 9 and 12, under as-rolled conditions, the pipeline steels with a bainitic ferrite structure have a better strength-SCC resistance combination than those with a ferrite + pearlite structure. This statement is confirmed by the experimental results shown in Figure 13, which were obtained for X70 steel under different heat-treatment conditions. As shown in Table 2 and Figure 2, where the ferrite + pearlite structure is obtained with annealed or normalized treatment, the bainitic ferrite is achieved with water-quench and water-quench + 250°C temper treatments, and the polygonal ferrite is obtained with oil-quench, oil-quench + temper, and water-quench + 650°C temper treatments. The polarization resistance of X70 steel with polygonal ferrite is relatively higher than that with bainitic ferrite but the former displays lower hardness.
The effect of pearlite content on the polarization resistance-hardness relationship is shown in Figure 14. The microstructure for X70 under quenched + 650°C tempered, oil-quenched, and oil-quenched + tempered conditions is mainly the polygonal ferrite with a small amount of fine carbide, which is uniformly distributed. For this reason, the steel with polygonal ferrite displays the highest SCC resistance. Under the annealed or normalized conditions, X70 steel possesses SCC resistance higher than X52, X60, and X65. Because X70 has a carbon content lower than the X52 and X60 steels,6,18 the pearlite content in the heat-treated X70 steel is less than that in X52, X60, and X65. This result indicates that an increase in the amount of pearlite in the microstructure will reduce the SCC resistance of pipeline steels. It agrees with the experimental observations reported by Chu, et al.,15 and Beavers, et al.13 Chu, et al.,15 reported recently that small cracks were more likely to initiate and to propagate in the pearlite colonies or along the boundary of pearlite/ferrite. One possible reason is that the quenched + tempered treatment produces a uniform microstructure while pearlite colonies in the ferrite + pearlite structure are likely to promote nonuniform micro-plastic deformation,30 since pearlite possesses higher strength than ferrite. Beavers, et al.,13 pointed out that the micro-hardness in SCC zones on the pipes was slightly higher than that in non-SCC zones. For steel with the ferrite + pearlite structure, SCC zones are likely to contain more pearlite because the pearlite is harder than ferrite. A recent study of ours indicated that both the metallurgical and environmental factors that enhanced the local plastic deformation would accelerate the development of near-neutral pH SCC of pipeline steels.9,28

It is also shown in Figure 14 that X52 and X60 steels, under the as-rolled conditions, have higher hardness values than the steels under the annealed or normalized conditions. Their polarization resistance is almost the same, although all specimens possess a ferrite + pearlite structure. This result indicates that if the steels have a ferrite + pearlite structure, the substructure formed during the hot-rolling process can improve the strength of pipeline steels without the loss of SCC resistance. However, it does not seem to be true for the steels with bainitic ferrite structures. In Figure 15, the test data obtained from the X70, X80, and XlOO steels under the quenched + low-temperature tempered conditions are compared with those obtained under as-rolled conditions, where all specimens possess the bainitic ferrite structure. The polarization resistance displays a general decreasing tendency as the hardness increases, but no remarkable difference is found in the specimens under asrolled and heat-treated conditions.

Obviously, the conclusions obtained from this study are only preliminary. A further study should be done on the correlation between the SCC resistance of pipeline steels and the microstructure parameters. To achieve this, the investigations should define the microstructure parameters, establish the relationship between the SCC resistance and the microstructure parameters, and find the microstructure parameters controlling the SCC resistance.
CONCLUSIONS
* The SCC resistance of pipeline steels with microstructures of the same type decreases with increasing yield strength, but the relationship between the yield strength and the SCC resistance is microstructure-dependent.
Pipeline steels with a microstructure of fine-grained bainite + ferrite have a better combination of strength and SCC resistance than those with ferrite + pearlite structures.

* Welding reduces the SCC resistance of pipeline steels and this effect is more pronounced in the steels with bainitic ferrite structures.

Sumber : Lu, B TLuo, J L. "Relationship Between Yield Strength and Near-Neutral pH Stress Corrosion Cracking Resistance of Pipeline Steels-An Effect of Microstructure". 27 Januari 2014. http://search.proquest.com/docview/223124029?accountid=31562

Tidak ada komentar:

Posting Komentar