You are on page 1of 7

Proceedings of the 2012 9th International Pipeline Conference IPC2012 September 24-28, 2012, Calgary, Alberta, Canada

IPC2012-90685

THE EFFECT OF MICROSTRUCTURE ON TENSILE BEHAVIOUR OF X80 MICROALLOYED STEEL


Katherine Jonsson University of Alberta Edmonton, Alberta, Canada Douglas G. Ivey University of Alberta Edmonton, Alberta, Canada Hani Henein University of Alberta Edmonton, Alberta, Canada

Shahrooz Nafisi Evraz Inc., NA Regina, Saskatchewan, Canada Thomas Garcin University of British Columbia Vancouver, British Columbia, Canada ABSTRACT A high degree of work hardening is desirable for steels to be employed in strain-based pipeline designs. In an effort to enhance work hardening characteristics, this study was conducted to determine the effect of thermal treatment on microstructural development and the subsequent relationship between microstructure and tensile behaviour of high strength microalloyed line pipe steel. A series of thermal schedules was applied to X80 steel samples using a Gleeble thermomechanical simulator in order to generate a variety of microstructures. The microstructures were quantified by calculating the phase fraction of individual phases using scanning electron microscopy (SEM). A focused ion beam (FIB) instrument was used to prepare electron transparent samples of specific grains that were characterized using transmission electron microscopy (TEM). The X80 microstructures were composed mostly of bainitic and ferritic grains with isolated pockets of martensite and M-A islands due to local carbon segregation. The effect of thermal treatment on microstructural evolution was determined based on varying the interrupt temperature, re-heat temperature and hold time at elevated temperatures. The overall effect of microstructure on the mechanical properties was evaluated, with a particular focus on hardness values and the shape of the stress-strain curves. The effect of thermal history and microstructure development on the work hardening characteristics was also determined.

Laurie Collins Evraz Inc., NA Regina, Saskatchewan, Canada Warren Poole University of British Columbia Vancouver, British Columbia, Canada

INTRODUCTION Microalloyed steels are a class of high strength low alloy (HSLA) steels that contain small amounts (less than 0.10 wt % each) of alloying elements, such as Nb, V and Ti, which provide strengthening through grain refinement and precipitation strengthening [1]. Microalloyed steels are able to achieve an excellent balance between strength and toughness due to thermo-mechanical controlled processing (TMCP), which creates a fine-grained structure by restricting austenite recrystallization during rolling and simultaneously enhancing precipitation of fine Nb and Ti carbonitrides which hinder grain growth [2]. Furthermore, the high cooling rates achieved in TMCP also improve both strength and toughness by refining the final microstructure and generating high strength lowtemperature microconstituents [3]. Microalloyed steels are often used in pipeline applications due to their high strength, good low-temperature toughness, excellent corrosion resistance, high strain hardening and good weldability. More specifically, microalloyed steels are being considered for Arctic pipeline applications, which require steel products with high buckling resistance to withstand the large plastic strains induced during the freezing and thawing cycles in permafrost regions [4]. Buckling resistance is achieved with a high work hardening coefficient and a round engineering stress-strain curve. Microstructure plays a large role in buckling resistance and work hardening behaviour, although a definitive relationship

Copyright 2012 by ASME

between microstructure and work hardening characteristics remains undetermined. One study performed by JFE Steel argues that a dual-phase structure of bainite and M-A (martensite-austenite constituent) results in high strain hardenability [4], whereas another JFE report argues that a ferritic-bainitic dual-phase steel is ideal to prevent buckling [3]. The study presented in this paper was performed to investigate the relationship between thermal history and microstructure development and, subsequently, the effect of microstructure on the work hardening behaviour of an X80 grade of microalloyed steel. X80 steel samples were subjected to an interrupted thermal treatment (ITT) process, which involved cooling the samples to an interrupt temperature and then immediately re-heating to an elevated temperature to alter the final microstructure. The ITT trials were performed with the intent to create a primary structure upon initial cooling to the interrupt temperature, then refining the remaining microstructure through an on-line heat treatment process followed by slow cooling to room temperature. A dual-phase microstructure of bainite and ferrite was expected which would exhibit the best combination of strength, toughness and strain hardening. The introduction of a finely dispersed high carbon phase should increase the ultimate tensile strength (UTS) while maintaining the yield strength (YS), thereby decreasing the YS/UTS ratio. EXPERIMENTAL PROCEDURE A general microstructure development map was developed and is shown in Figure 1. The hold time at the interrupt temperature and the hold time at the re-heat temperature were varied while keeping the other variables (primary cooling rate, interrupt temperature, re-heat temperature and secondary cooling rate) constant in order to directly compare the effect of hold times at the two temperatures. The first set of samples was cooled to an interrupt temperature of 585 C at a consistent primary cooling rate of 10 C/s, held for different durations at the interrupt temperature and then quenched to room temperature (corresponding to points 1, 2 and 3 in Figure 1). The second set of samples was cooled to the interrupt temperature, immediately re-heated to 750 C and held for different lengths of time before being quenched to room temperature (indicated by points 4, 5 and 6). The full ITT process was applied to samples shown by curves 7, 8 and 9, which are identical to the thermal history for samples 4, 5 and 6 but with a controlled cooling rate of 3 C/s from the reheat temperature to room temperature. The reference sample (REF) was continuously cooled at 10 C/s, which was chosen to mimic the primary cooling rate used for the ITT samples. The complete thermal history for each sample is summarized in Table 1, with all samples austenitized at 945 C for 5 minutes before further thermal treatment. The X80 steel samples contained 0.06 wt% C, 1.7 wt% Mn, 0.3 wt% Mo, 0.3 wt% Si, and < 0.2 cumulative wt% of Cr, V, Nb and Ti and were

obtained from plate of as-rolled steel that was subjected to conventional TMCP conditions used in regular mill operations. The samples were machined into 203.2 mm by 50.8 mm strip specimens with a uniform thickness of 1.5 mm prior to thermal cycling in a Gleeble 3500 machine. After thermal testing, three sub-size tensile specimens were machined from each strip specimen according to the dimensions specified in ASTM 8M-04 [5]. A small section from the center of each strip sample was cut using a diamond saw and mounted in Bakelite powder with a hot press. The center section of the strip was chosen for metallographic analysis to guarantee that the region to be investigated microscopically experienced the intended thermal treatment and would therefore contain a representative microstructure. The mounted specimens were ground using 120, 320, 400, 600 and 800 grit silicon carbide papers and then polished using 1 m diamond paste followed by 0.05 m alumina slurry. The samples were etched using 2 % Nital for approximately 5 s to expose the microstructural features. Secondary electron (SE) images were obtained using a Philips XL30 scanning electron microscope (SEM) operated at 10 kV. A Hitachi NB5000 focused ion beam instrument (dual beam FIB/SEM) was used to fabricate site-specific transmission electron microscopy (TEM) samples, which were investigated using a JEOL JEM 2010 TEM operated at 200 kV. Specific grains of interest were located in the microstructure using the dual beam FIB/SEM and a small section (approximately 10 m long by 2 m wide by 10 m deep) was removed from the sample using the FIB Ga ion beam. The regions of interest that were plucked were then thinned to a thickness of approximately 100 nm to ensure electron transparency. Microhardness measurements were obtained from a Mitutoya MVK-H1 hardness testing machine with a 1 kg load. Tensile tests were performed on sub-size tensile specimens according to ASTM 8M-04 [5] on an INSTRON 8800 equipped with a 100 kN load cell and a servohydraulic actuator.

FIGURE 1: MICROSTRUCTURE DEVELOPMENT MAP FOR ITT SAMPLES

Copyright 2012 by ASME

TABLE 1: THERMAL HISTORY OF ITT SAMPLES

Sample REF 1 2 3 4 5 6 7 8 9

Primary Cooling Rate (C/s) 10 10 10 10 10 10 10 10 10 10

Interrupt Temperature (C) 585 585 585 585 585 585 585 585 585

Hold Time (min) 0 5 10 0 0 0 0 0 0

Re-heat Temperature (C) 750 750 750 750 750

Hold Time (min) 0 5 10 0 5

Secondary Cooling Rate (C/s) quench to RT quench to RT quench to RT quench to RT quench to RT quench to RT 3 3

RESULTS AND DISCUSSION The microstructure from the reference sample is shown in Figure 2 and the ITT trial micrographs are shown in Figure 3. The reference sample mostly consists of darker regions (ferrite) with a small number of lighter regions (martensite and bainite) with some isolated islands of M-A. The ITT samples contained mixed fractions of ferrite, martensite, bainite and M-A in various proportions. A quantitative study on the microstructures was performed in order to determine the phase fractions of the various constituents. Before this step could be completed with confidence, however, the nature of certain indistinguishable grains that were observed in SEM micrographs was confirmed with TEM.

FIGURE 2: SEM SE IMAGE OF THE REFERENCE SAMPLE

Two TEM samples were prepared the first sample was obtained from sample 8 in Table 1 whereas the second sample was selected from sample 1. An SEM SE image of the region of interest from sample 8 is shown in Figure 4, with the black box indicating the specific region that was removed from the sample using the FIB technique. TEM results from the lefthand portion of the region indicated in Figure 4 are shown in Figure 5. Figure 5a shows a low magnification TEM

750 10 3 micrograph with the black box indicating the area imaged in Figure 5b, which was obtained at a higher magnification. Figures 5c and 5d were both obtained from the region identified by the black box in Figure 5b and represent high magnification images taken at different tilt angles. The microtwins observed in Figure 5c and 5d go in and out of contrast during tilting as a result of changing diffraction conditions. Figure 5e shows the diffraction pattern obtained from the region indicated in Figure 5d. The pattern was indexed to bcc ferrite with a zone axis close to a [513] orientation, but due to the low carbon content of this steel the c/a ratio for martensite is very close to 1 so this pattern can be considered to be martensite. The presence of the microtwins shown in Figure 5c and 5d is evidence that the region shown in Figure 4 is indeed martensite. The second region of interest that was examined in the TEM was from sample 1 described in Table 1. The specific region of interest plucked from the sample using the FIB process is indicated by the black box in Figure 6 with the corresponding TEM micrographs shown in Figure 7. Figure 7a is a low magnification image, whereas Figure 7b is a higher magnification image of the region identified by the black box in Figure 7a. Figures 7c, 7d and 7e are diffraction patterns from the regions labeled 1, 2 and 3, respectively, in Figure 7b. The three diffraction patterns are similar and were indexed to ferrite with a zone axis close to [113]. The small differences in spot intensities indicate that regions (laths) 1, 2 and 3 are slightly misoriented relative to one another. The lath-like structures, coupled with their similar orientations, are indicative of a bainitic structure. Based on the TEM results, which confirmed the nature of the unknown regions of interest observed in the microstructure of different samples, quantitative microstructural analysis was conducted on representative micrographs from each sample. The micrographs were analyzed using an image processing program (FIJI) which yielded the phase fractions of individual microconstituents. The results were averaged and are tabulated in Table 2. Four constituents were identified:

Copyright 2012 by ASME

FIGURE 3: SE SEM IMAGES OF THE THERMALLY TREATED SAMPLES; THE NUMBERS REFER TO SAMPLE ID IN TABLE 1

ferrite, bainite, martensite and M-A; however, due to difficulties discerning the individual features in the SEM micrographs, Table 2 includes the combined percentage of bainite and martensite.

FIGURE 4: SEM SE IMAGE OF REGION 1 (SAMPLE 8) FIGURE 5: TEM BRIGHT FIELD (BF) MICROGRAPHS AND DIFFRACTION PATTERN FOR REGION 1 FROM FIGURE 4

Copyright 2012 by ASME

FIGURE 6: SEM SE IMAGE OF REGION 2 (SAMPLE 1)

5 and 6, sample 4 had the highest percentage of ferrite and, subsequently, the lowest fraction of bainite and martensite. This data suggests that an increase in hold time at 750 C does not increase ferrite content. This observation is counter-intuitive considering that ferrite nucleation is a kinetic process which is temperature-sensitive. One would expect a higher amount of ferrite for sample 6 than sample 4, if transformation from austenite to ferrite continued as a function of hold time at 750 C. The full ITT process is indicated with samples 7, 8 and 9. The microstructures of these samples show a decreasing trend in ferrite content with a corresponding increase in the combined bainite and martensite fraction as hold time at 750 C increased from 0 to 10 mins. The ferrite content of 7, 8 and 9 is very similar to 4, 5 and 6, which again indicates that ferrite formation is not affected by the re-heat process. This suggests that the longer hold time at 750 C combined with a slow secondary cooling rate increases the bainite and martensite content as opposed to a direct quench from 750 C (sample 6).
TABLE 2: QUANTITATIVE MICROSTRUCTURE ANALYSIS

Sample REF 1 2 3 4 5 6 7 8 9

% ferrite 93.4 72.6 76.6 72.4 81.1 75.0 77.2 82.7 79.9 75.0

% bainite & martensite 5.0 23.8 20.8 26.3 18.1 24.6 22.3 13.5 17.1 24.1

% M-A 1.9 3.6 2.6 1.2 0.8 0.4 0.5 3.8 3.1 0.9

FIGURE 7: TEM BF MICROGRAPHS AND DIFFRACTION PATTERNS OF REGION 2 SHOWN IN FIGURE 6

The effect of hold time at the interrupt temperature of 585 C can be determined by comparing samples 1, 2 and 3 in Figure 3 and in Table 2. The ferrite appears roughly equiaxed whereas the martensite and bainite grains are jagged and irregular in shape. The M-A constituent appears as very small round or stringy features that are situated along grain boundaries or at triple points. The phase fractions of each microconstituent remain fairly consistent for these three samples, with the exception that sample 2 has slightly higher amounts of ferrite and a lower combined fraction of bainite and martensite than samples 1 and 3. The phase distribution indicates that the hold time at 585 C had no significant effect on the amounts of microconstituents that form at the interrupt temperature. The effect of hold time at 750 C can be determined by comparing samples 4, 5 and 6 in Figure 3 and Table 2. The grain morphology for these samples, as shown in Figure 3, appears to be fairly consistent as hold time at 750 C increases. Sample 4 has a higher percentage of ferrite and a lower amount of M-A than 1, 2 and 3, although samples 5 and 6 have similar ferrite and combined bainite and martensite fractions as samples 1, 2 and 3. The M-A content of samples 4, 5 and 6 is significantly lower than for 1, 2 and 3. Comparing 4,

As mentioned in the Introduction section of this paper, the microstructural investigation was conducted to determine the effect of thermal history on microstructure development. The next step was to correlate the microstructure to mechanical properties, most notably, work hardening characteristics. The measured hardness values are shown in Table 3. The microhardness results in Table 3 are relatively inconclusive. Despite the difference in microconstituent fractions, the hardness values are very similar. Sample 7 is an exception this sample has the highest hardness value and also has the highest ferrite content other than the reference sample, which has the lowest hardness value. Comparing sample 7 to sample 4, the ferrite contents are almost identical but the amount of bainite and martensite is 5 % lower for sample 7 than for sample 4. This would suggest that sample 7 has a lower hardness (due to the lower amount of hard constituents), but the opposite is true: sample 7 in fact has a hardness value that is approximately 15 HV higher than sample 4. The tensile test results for the ITT trial specimens are plotted as stress-strain curves in Figure 8 and important mechanical parameters (including yield strength, ultimate

Copyright 2012 by ASME

tensile strength, ductility and work hardening coefficient, n) are tabulated in Table 4. The work hardening coefficient was determined from true stress-strain curves. There is no correlation between the hardness values and the UTS shown in Table 4. The reference sample, which has the highest UTS, has the lowest hardness value. This trend is contradictory to expected results as a higher hardness value should correspond with a higher UTS value.
TABLE 3: MICROHARDNESS RESULTS FOR ITT TRIAL SAMPLES

Sample REF 1 2 3 4 5 6 7 8 9

HV 229 1.4 232 1.4 231 2.8 235 2.0 239 1.5 240 2.3 240 2.9 254 2.2 242 1.3 242 3.9

bainite and martensite leads to improved mechanical properties. The stress-strain curve for sample 9 was very similar to sample 7 but vertically shifted down to lower strength values (again, the combined fraction of martensite and bainite for sample 9 is 11% higher than for sample 7 but the strength values are significantly lower). The tensile behaviour cannot be explained by phase distribution alone but rather a detailed investigation into the grain sizes for the various microconstituents must also be conducted to fully understand the effect of thermal treatment on microstructure development and the subsequent relationship to mechanical behaviour.
TABLE 4: TENSILE DATA FOR ITT TRIALS

Sample REF 1 2 3 4 5 6 7 8 9

UTS (MPa) 788 707 510 540 694 699 692 576 526 431

YS (MPa) 548 428 373 300 463 445 445 354 297 244

YS/UTS 0.70 0.60 0.61 0.56 0.67 0.64 0.64 0.61 0.56 0.53

n 0.13 0.14 0.17 0.17 0.11 0.12 0.12 0.15 0.14 0.15

Ductility 0.20 0.09 0.07 0.08 0.19 0.21 0.22 0.14 0.19 0.15

The reference sample had the highest strength and ductility values and the ITT trial samples did not improve the YS or UTS. The changes in microstructure as a result of the thermal treatments resulted in decreased mechanical property performance. The work hardening coefficient remained largely unchanged and only slight decreases in YS/UTS were observed as a result of the ITT process. Increasing the hold time at 585C resulted in decreased strength and toughness. The microstructures of sample 1 and sample 3 are very similar (the phase fractions are within roughly 1% of each other); however, the engineering stress-strain curve for sample 3 is almost identical to sample 2 even though sample 2 has a different phase distribution than sample 3. Samples 4, 5 and 6 had similar elongation and UTS values to the reference sample although the microstructure quantification yielded very different phase compositions (more specifically, the reference sample consisted mostly of ferrite while samples 4, 5 and 6 had significant contributions from bainite and martensite). The YS/UTS ratio was successfully decreased for samples 4, 5 and 6 as compared with the reference sample; however, this result was not achieved by increasing the UTS but rather by decreasing both the YS and the UTS simultaneously. The full ITT samples (7, 8 and 9) also exhibited similar ductility to the reference sample but with significantly lower strength values. Sample 7 had the highest strength of the full ITT process, indicating that a 0 min hold time at 750 C achieved the best mechanical properties compared with hold times of 5 and 10 mins at the elevated temperature. Furthermore, sample 7 had the highest ferrite and the lowest bainite and martensite content of samples 7, 8 and 9, thereby indicating that a lower phase fraction of

CONCLUSIONS The effect of thermal treatment on microstructure development and tensile behaviour was investigated by applying an interrupted thermal treatment (ITT) process to X80 microalloyed steel. The results of this study are listed below: 1) The ITT procedure did not significantly change the microstructure of X80 steel. The hold time at 585 C did not affect the amount of microconstituents that formed, whereas the hold time at 750 C caused slight differences in the ferrite and combined martensite and bainite fractions with no evident trend. 2) The ITT samples did not perform as well as the reference sample in tensile testing, although the samples held at 750 C for 0, 5 and 10 minutes before quenching to room temperature had similar ductility and only slightly reduced UTS values. 3) The YS/UTS ratio was successfully minimized as a result of the ITT process (samples that were heated to 750 C and held for 0, 5 and 10 minutes before cooling to room temperature at 3 C/s); however, this was achieved by reducing both the YS and UTS values rather than increasing the UTS and maintaining a constant YS value. 4) The work hardening coefficient was largely unchanged as a result of the ITT schedule. ACKNOWLEDGMENTS This work was supported by Evraz Inc. NA and the Natural Sciences and Engineering Research Council (NSERC)

Copyright 2012 by ASME

FIGURE 8: ENGINEERING STRESS-STRAIN CURVES FOR ITT SAMPLES

of Canada. The authors would like to thank M. Kupsta (National Research Council of Canadas National Institute for Nanotechnology (NRC-NINT)) for his assistance with FIB sample preparation and Mr. L. Good from Evraz Inc. NA for performing Gleeble tests and assisting with SEM analysis. REFERENCES [1] Gladman, T., 1997, The Physical Metallurgy of Microalloyed Steels, The University Press, Cambridge, UK, Chap 1. [2] 2005, ASM Handbooks Volume 1. High-Strength Structural and High-Strength Low-Alloy Steels. United States of America. [3] Nobuyuki, I., Endo, S., Kondo, J., 2006, High Performance UOE Linepipes, JFE Technical Report, 7, pp 20-26. [4] Mitsuhiro, O., Nobuo, S., Kondo, J. 2008. Development of a High-Deformability Linepipe with Resistance to Strainaged Hardening by HOP (Heat-treatment On-line Process), JFE Technical Report, 12, pp 8-14. [5] 2003, ASTM 8M-04: Standard Tests Methods for Tension Testing of Metallic Materials, United States of America.

Copyright 2012 by ASME