Grain Boundary Serration in Nickel-Based Superalloy Inconel 600: Generation and Effects on Mechanical Behavior

Grain Boundary Serration in Nickel-Based Superalloy Inconel 600: Generation and Effects on... TOPICAL COLLECTION: SUPERALLOYS AND THEIR APPLICATIONS Grain Boundary Serration in Nickel-Based Superalloy Inconel 600: Generation and Effects on Mechanical Behavior YUANBO T. TANG, ANGUS J. WILKINSON, and ROGER C. REED Grain boundary serration in the superalloy Inconel 600 was studied. Two microstructural variants, one with nonserrated and the other with serrated grain boundaries were generated by altering the heat-treatment conditions, while keeping other aspects of the microstructure unchanged. The effect on the creep response between 700 C and 900 C was measured, and the different failure modes and accumulated damage were quantified using high-angular resolution electron backscatter diffraction analysis in the scanning electron microscope and also by X-ray computed tomography. It is found that serration plays a more crucial role in the high-tem- perature/low-stress regime when an intergranular cracking mechanism involving cavitation is operative; here it plays a role in improving both creep life and creep ductility. Any effect of serration is less prevalent at low temperatures where transgranular failure is dominant. https://doi.org/10.1007/s11661-018-4671-7 The Author(s) 2018 [3–11] I. INTRODUCTION composition. The study of Koul & Gessinger on [3] the Inconel 738, Nimonic 115 and Nimonic 105 alloys, SERRATION of grain boundaries in metals and involving an impressive combination of experimentation alloys, for example the nickel-based superalloys of the 0 and modeling, identified the role of c migration on type studied here, is intriguing for various reasons. First, 0 serration. The role of the c phase has since been there is complexity arising from the inapplicability of the [6,12–15] confirmed by a number of researchers. But what generally accepted picture of smooth or gently curved happens in the absence of the c phase? Alloys contain- grain boundaries, as observed on the microscale. ing very few or no c -forming elements have also been Second, there is the fundamental question of why grain shown to form serrated grain boundaries, for example boundary serration arises in the first place. Third, effects Haynes 230, Inconel 690, Inconel 617, and Nimonic of serration on the mechanical behavior of these [7,10,16,17] 263. In this case, the formation mechanism for materials are not well understood. What are the these alloys involves the triggering of serration by underlying fundamental effects which cause it? At Cr-rich M C carbide. 23 6 present, no truly unequivocable explanation exists. Is the phenomenon of grain boundary serration of Detailed experimentation involving carefully controlled mere academic interest, or are there significant beneficial measurements and high-resolution characterization is effects which confer technological advantage? The needed. time-dependent deformation has been shown to be Serration of grain boundaries in nickel alloys was first improved dramatically by such features, particularly [1] [2] reported by Larson et al. and Miyagawa et al. in the creep ductility/lifetime and crack growth rate. Early 1976; slow cooling from above the c dissolution research conducted on Inconel 792 and 20-11P con- temperature was employed. Since then, attempts have firmed reduced creep rate and increased creep [1,2] been made to elucidate the formation mechanism, strength, although the postmortem analysis was details of heat treatment and the role of alloy limited to the characterization techniques available at that time. Later studies on IN738, Astroloy, PM alloy-10, Waspaloy, and RR1000 have proven similar [18–23] influences. The study up to date on STAL-15 again revealed similar trend—with both extensive experiments YUANBO T. TANG and ANGUS J. WILKINSON are with the [24–26] and modeling efforts. In addition, for alloys with Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, UK. Contact e-mail: yuanbo.tang@materials.ox.ac.uk low or no c phase, creep life and ductility are signifi- [7,27,28] ROGER C. REED is with the Department of Materials, University cantly enhanced. In summary, grain boundary of Oxford and also with the Department of Engineering Science, serration has been demonstrated to improve mechanical University of Oxford, Parks Road, Oxford OX1 3PJ, UK. properties such as creep, low cycle fatigue (LCF), and Manuscript submitted January 23, 2018. dwell fatigue remarkably in various superalloys. A few Article published online June 1, 2018 4324—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A hypotheses have been proposed for taking into account deformation modes which are operative. Attention is of the improved behavior, e.g., prevention of grain paid not only to postmortem analysis, but also to boundary sliding, increased grain boundary diffusion interrupted studies. In the final part of the paper, the path, and impedance of crack growth. However, no implications of the serration effect on the mechanical unambiguous mechanism is yet supported strongly by response of this alloy are considered with an emphasis substantial experimental evidence or modeling. The past on carbide evolution, cavitation, and dislocation sub- efforts have mostly been devoted to the serration effects structures which arise. from different angles in a qualitative manner, such as alloy composition, mechanical properties, and cooling rates, but little study has been attempted to control II. METHODOLOGY AND EXPERIMENTAL other microstructural variables at the same time. The DESCRIPTION most important missing part is that no quantitative measurements have been conducted in order to under- Inconel 600 was chosen for the current study, on stand such phenomenon. account of the relative simplicity of its chemical com- position, Table I. In this alloy, Al or Ti is present at only A further point relates to the complexity of the trace levels so that any influence of the c phase on grain situation. Are the beneficial influences really contributed by grain boundary serration alone, or are there other boundary serration cannot then arise. Moreover, refrac- changes to microstructural features due to slow cooling tory elements such as Ta, Mo, and W are also absent; which are influential? In fact, few researchers have really therefore, grain boundary MC-type carbides enriched addressed this issue, and studies on the serration effect with these elements have no contribution to microstruc- have seldom reported very well-controlled microstruc- ture; neither M C nor M C carbides could be formed 23 6 7 3 ture, obviously due to the nature of the problem. Creep from MC-carbide decomposition. Specimens were taken is one of the most complicated continuum damage from a cold-drawn and annealed bar of circular cross processes, and many factors are well known to be section with a diameter of 16 mm, in which the virgin crucial: grain size, c size, morphology and distribution, grain size was about 15 lm. twin boundary fraction, carbide type/morphology, size, etc. All are likely to be affected by cooling rate to some A. Heat Treatment extent. For example, the c size and distribution can be directly influenced by nucleation density within the c Heat treatment was carried out in two ways in order matrix, which is a function of solution-treatment con- to develop different grain boundary morphologies, i.e., [4,8,23,29–31] dition and cooling rate. The slow cooling serrated and nonserrated (straight or gently curved). A causes larger c size and sometimes a bimodal distribu- solution heat-treatment temperature of 1140 C for 2 tion. Moreover, slow cooling lengthens the total time hours was used throughout, which is significantly higher than the estimated carbide solvus temperature of required for grain growth and partial dissolution of MC [32,33] 1020 C calculated using the ThermoCalc software. carbides, which makes good control over Such solution heat treatment was employed to maintain microstructual variables even harder. Clearly, very the same level of grain size, grain distribution, and R3 carefully designed experimentation is needed. twin boundary fraction in both microstructures, which The research reported in this paper was carried out are considered to be crucial microstructural variables for keeping the above issues in mind. The nickel-based creep and other mechanical properties. The selected heat superalloy Inconel 600 is chosen for study; an advantage is then that the alloying elements Al and Ti are absent, treatment compensated for small variations in total so that any involvement of the c phase Ni ðAl; TiÞ on heat-treatment time caused by different cooling rates, the serration effect cannot then arise. Furthermore, any and it proved the possibility to develop significant grain role by Ta- or W-rich MC carbides is avoided, because boundary serration at the microstructural level. This Ta and W are absent. Instead, as will be shown, Cr-rich was achieved by cooling the material beyond the carbide carbide species become critical to the development of solvus temperature at 12 C/min, measured by N-type grain boundary serration, with the cooling rate through thermocouples, in contrast to water quenching for the carbide solvus temperature also playing a crucial nonserrated grain boundaries. Heat-treatment profiles role. Here, serrated and nonserrated grain boundaries and representative microstructures are illustrated in were first prepared and characterized using scanning Figure 1. electron microscopy, with the grain size distribution and twin boundary fractions having proven to be B. Creep Testing unchanged; the creep response was then tested. Quan- The impact of grain boundary serration on mechan- titative high-angular resolution electron backscatter ical behavior was assessed from 700 C to 900 Cby diffraction (HR-EBSD) and X-ray computed tomogra- tensile creep rupture tests. Creep test specimens were phy (XCT) are used to gain insights into the creep Table I. Nominal Composition of the Alloy Studied in Weight Percent (Ni-Base) Name Ni Cr Fe C Mn Si Co Cu S Inconel 600 74.1 15.9 9.03 0.068 0.21 0.27 0.012 0.0018 0.001 METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4325 Fig. 1—(a) and (b): Heat-treatment profiles for development of nonserrated and serrated grain boundary materials and (c) and (d) corresponding microstructures obtained from heat treatments described in (a) and (b). The curves are drawn to illustrate grain boundary morphologies only and are deliberately offset in order not to obscure the grain boundaries in the micrographs. machined from heat-treated cylindrical bars, diameters condition, a representative sample, i.e., the sample with of which were reduced from 16.4 to 4 mm after heat median rupture life among three, was first taken for treatment to avoid any effect of oxidation damage on fractography analysis. The fractured sample was then specimen gauge section. Creep tests were carried out cut by wire-based electrical discharge machining (EDM) under three conditions for both microsturctures, using from the middle of fracture tip along the axial direction. specimens of 4 mm diameter and 20 mm length (3 tests The sectioned sample was mounted and then prepared under each condition). An Instron 8862 servoelectric via standard metallography procedures, with 2 minutes machine was used, where the strain was measured using colloidal silica for finishing. The SEM is equipped with a an electromechanical actuator equipped in the lower end Bruker EBSD detector, which was used to capture of the crosshead operated at a frequency of 1 Hz with EBSD patterns at a resolution of 600  800 for maps the precision of 1 lm/h. In addition, another six tests collected at a step size of 220 nm. Furthermore, six were interrupted at 5 pct strain, to observe deformation interrupted specimens were also cut by EDM in the and fracture mechanisms under each condition to same way and scanned by HR-EBSD using the same evaluate the influence of serration. The tests were conditions. A cross-correlation-based analysis was used [34–36] carried out under different conditions of significant on the stored EBSD patterns. variations in temperature and stress: 700 C/170 MPa, 815 C/70 MPa, and 900 C/40 MPa. Temperature was D. X-ray Tomography-Based Characterization monitored using K-type thermocouples placed in con- Since the reduction in area and cracking from grain tact with the specimen. Each specimen was held at the boundaries in the crept samples were found to be testing temperature for 1 hour prior to loading. significant, X-ray computed tomography was used to quantify cavity size and its distribution. One represen- C. SEM-Based Characterization tative fracture tip was taken for each creep condition The grain structure and boundary types produced by and tomography analysis carried out using a North Star the two heat treatments were characterized using a Zeiss Imagix system at 150 kV and a current of 27 lA, which Merlin high-resolution field emission gun scanning defined a voxel size of 4.86 lm. For each sample, a electron microscope (FEG-SEM) that was also used height of roughly 10 mm was characterized using 3600 for the fractography and HR-EBSD scans. For each projections and a helical scan used to reconstruct the 3D 4326—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 2—Postmortem X-ray tomography setup and computed results—quantification of cavitation distribution. (a): X-ray technique setup. (b): Volume of the specimen reconstructed by X-ray tomography. (c) Surface reconstruction completed by Avizo with internal cavities labeled in red. (d) A cylindrical volume of 5 mm height from lower end of the fracture tip was used for cavity size and distribution analysis (Color figure online). Fig. 3—Energy dispersive X-ray spectroscopy (EDS) analysis in a region that contains a serrated grain boundary and planer carbides. (a) and (b) show the regions at two magnifications. (c): Composition maps indicate segregation of Cr and C and depletion of Fe and Ni at the grain boundary. The dark phases are determined as Cr-rich Carbide. volume. ImageJ and Avizo softwares were utilized for and without grain boundary serration. For this, quan- quantification of cavity size and distribution using tification of a 3D cylindrical volume of about 3 mm [37] histogram-based methods. Comparisons of cavity diameter and 5 mm height was selected as illustrated in size and distributions were then made for samples with Figure 2. METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4327 III. RESULTS Grain size and twin boundary fractions can have great influences on creep, and so they were measured for both A. Heat Treatment and Grain Boundary Serration microstructures using EBSD over large sampling areas Development for achieving statistically meaningful estimates. Three random regions of each sample were mapped with The profiles of heat treatment that generated each EBSD to cover over 800 grains for each microstructure. type of grain boundary morphology are shown in The EBSD data were then postprocessed using the Figures 1(a) and (b). Backscatter electron imaging in ESPRIT software, with a threshold grain boundary the FEG-SEM revealed grains and carbides due to the misorientation angle defined as 10 deg. Representative channeling effect and compositional contrast. Typical EBSD maps are illustrated in Figures 4(a) and (b) for microstructures produced from each cooling rate are the nonserrated and serrated microstructures, respec- shown in Figures 1(c) and (d); for illustration purposes, tively. Grain size distribution histograms are shown in curves highlighting the grain boundary morphologies Figures 4(c) and (d), where two distributions overlap are displaced laterally from the actual boundaries. For the nonserrated grain boundary case, no secondary with each other and both are skew with tails on the high phases were observed at the FEG-SEM resolution; grain size side. Mean values of grain size with boundary however, the serrated grain boundaries were always length fraction of R3 (twin) boundaries are given on the distribution plots and show very little variation between associated with intergranular carbides at the microscale. the two microstructures. High-resolution energy dispersive X-ray spectroscopy Postprocessed EBSD results on misorientation angles (EDS) was used to probe the chemical compositions of demonstrated that negligible low-angle (<15 deg) grain such carbides on a serrated grain boundary, and boundaries were present after heat treatment, which Figure 3 shows a typical composition map. This and suggests dislocation arrays/polygonization were mini- other similar maps determined the carbides to be mized. While the mechanism in relation to serration enriched in chromium but depleted in iron and nickel. Fig. 4—(a) and (b) Examples of inverse pole figure maps used for grain size measurements for nonserrated and serrated materials, respectively. (c) and (d) Grain size distributions for each microstructure. Average grain sizes are 376 and 395 lm; twin boundary area fractions are 64.7 and 65.3 pct, respectively, for nonserrated and serrated grain boundary materials (Color figure online). 4328—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 5—Creep data for nonserrated and serrated architectures, (a through c) under conditions of 700 C/170 MPa, 815 C/70 MPa, and 900 C/ 40 MPa. Top row: Representative creep curves for median rupture life samples. (d) demonstrated average rupture lives displayed with error bars. (e) Demonstrated average ductility displayed with error bars. Serrated grain boundary shows no creep life improvement at 700 C, but, enhancements at 815 C and 900 C. Creep ductility was increased by a moderate margin under all conditions (Color figure online). formation is out of this paper’s scope, some measure- Figure 5. The top row presents the creep curves of ments have, however, been performed on serration samples with median rupture life in each condition. The generation. The serration amplitude is found to be a bottom row displays a summary of rupture life and function of cooling rate. For the given solution treat- creep ductility with error bars representing ± one ment used here, the serration amplitude varied between standard deviation. For rupture life, the serrated grain 300 and 600 nm. In addition, although grain sizes were boundary microstructure conferred no life enhancement shown unchanged under both cooling rates, grain at 700 C. However, an increase in creep rupture life was boundary lengths were increased due to the presence measured at the higher test temperatures of 815 C and of local serrations. In total, six random serrated grain 900 C. From the average creep ductility, it is evident boundaries were analyzed using ImageJ software to that serrated grain boundary surpassed the nonserrated compare measurements along the serrated path, the under all conditions by a moderate margin. In addition, actual length to end-to-end straight-line distance. The for both microstructures, a drop in ductility was evident result shows serrations increase by 54 ± 11 pct of the at higher-temperature and lower-stress levels, which grain boundary length compared to perfectly straight suggests a change in fracture mechanism. boundaries. The same procedure was completed for Fracture surfaces of samples under each condition nonserrated case to account for curvature of bound- were characterized by FEG-SEM, for which illustrative aries. The nonserrated boundaries are also longer than results are given in Figure 6. Specimens with both straight-line paths range curvature which increased the microstructures showed similar transition behavior from length by 29 ± 8 pct of the grain boundary length. low to high temperature in deformation. A clear shift from transgranular-dominant features at low tempera- ture to intergranular-dominant cracking at higher tem- B. Creep Tests and Fractography peratures was observed. The largest difference was seen Creep curves measured at 700 C/170 MPa, 815 C/ under the 815 C/70 MPa condition, for which the 70 MPa, and 900 C/40 MPa for both nonserrated and serrated samples remained mostly transgranular, serrated grain boundaries samples are presented in whereas the nonserrated sample showed much more METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4329 Fig. 6—Fractography on representative samples under various conditions with nonserrated grain boundary, (a through c) and with serrated grain boundary, (d through f). (a): Transgranular-dominant deformation with limited signs of intergranular cracking, i.e., a few microvoids. (b) Intergranular-dominant cracking with limited transgranular deformation. (c): Intergranular cracking. (d): Transgranular-dominant deformation (e): Mixed-mode of trans/intergranular cracking. (f): Intergranular cracking. evidence of intergranular processes. This is believed to suggested by previous study on face-centered cubic [39] account for the marginal increase of ductility observed. (FCC) copper. For both microstructures, the cell size Cross sections of fracture tips under each condition increased with the decreasing stress, from 4.4 to 5.2 lm were scanned using EBSD at a low magnification. for the nonserrated case and from 4.8 to 6.8 lm for the Figures 7 and 8 demonstrate crystal orientation and serrated case. Under the same test conditions, the grain boundary type at fracture for nonserrated and nonserrated microstructure always exhibited a smaller serrated samples, respectively. Pattern quality maps cell size than that for the serrated case. were overlain with grain boundary, twin boundary, and subgrain boundary (< 5 deg) next to EBSD maps. It is D. Interrupted Creep Test clear to see the contrast of cavitation behaviors between the two microstructures, particularly under 700 Cand The evolution of the carbide distribution during the 815 C conditions, where notably more cavities were interrupted creep tests was explored using backscattered formed in the nonserrated case. In addition, twin electron (BSE) imaging of sectioned samples. Figure 10 boundaries in each case displayed higher integrity, and shows images of the nonserrated and serrated no cracks were found to propagate through them despite microstructures after creep to 5 pct strain at 900 C. the absence of carbides. Image intensity was used to segment the images and identify the carbides which are shown alongside the BSE images in Figure 10. For the nonserrated microstructure C. HR-EBSD on Fractured Specimens a relatively uniform dispersion of fine carbides develop Cross-correlation HR-EBSD analyses was performed across the grain interiors. In contrast, the carbide on sections through fractured samples that had been dispersion that develops in the serrated microstructure tested under the 815 C/70 MPa and 900 C/40 MPa is heterogeneous and shows much greater number conditions. Geometrically necessary dislocation (GND) density nearer to the grain boundaries than toward the density maps were calculated using the method center of the grains. The average carbide size is larger in described in References 35 and 38. Dislocation cell the serrated case (1000 nm) compared with the nonser- structures consisting of high GND density walls and low rated case (520 nm). The other striking feature is that for GND density interiors were developed within the grains the serrated case, the carbides are clearly dispersed along after the large strain (over 0.3) creep deformation, bands in well-defined directions within each grain. Figure 9. Cell sizes were measured for each specimen by Figure 11 compares carbide distributions in the serrated drawing 10 straight lines across each map (horizontally microstructure after testing at 900 C and 815 C and and vertically) and then counting the number of shows that the bands of carbides occur for both test intersections with cell walls. Cell walls were defined by conditions and are aligned with traces of expected 14 2 f11 1g slip systems as determined by EBSD. a minimum threshold GND density of 2  10 m , 4330—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 7—Inverse pole figures (IPF) corresponding to X-axis (a), (c), and (e) and pattern quality maps (b), (d), and (f) of nonserrated fracture tips under each creep condition. The pattern quality maps are overlain with grain boundary, twin boundary, and subgrain boundary. Cavitations exhibited were mostly of intergranular cracking in all samples. The same circular black spot is displayed both in (c) and (d), caused by surface contamination rather than cavitation (Color figure online). HR-EBSD was also conducted on sectioned samples Alignment of increased GND density regions into from the interrupted tests. At the lowest test tempera- straight linear features is less marked, and there is ture of 700 C, there were quite obvious differences evidence of dislocation cell formation near the grain between GND density distributions for the two boundaries. This suggests a more difficult slip-transfer microstructures, as shown in Figure 12. For the non- process for the serrated grain boundaries. serrated case, there are some linear bands of increased At 815 C, Figure 13, the straight linear features are GND density that are aligned with the f11 1g slip plane suppressed for both microstructures. For the nonser- traces in each grain, but there is little evidence of higher rated case, there is increased GND density and disloca- GND density near the grain boundaries suggesting slip tion cell formation near the triple junctions, and to a transfer was relatively easy. In contrast, for the serrated lesser extent, where twin boundaries intersect with case, there is very marked and obvious accumulation of high-angle grain boundaries. For the serrated case, the increased GND density near the grain boundaries. GND density is markedly higher in regions close to METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4331 Fig. 8—Inverse pole figures (IPF) corresponding to X-axis (a), (c), and (e) and pattern quality maps (b), (d), and (f) of serrated fracture tips under each condition. The pattern quality maps are overlain with grain boundary, twin boundary, and subgrain boundary. Little cavitations were exhibited in 700 C sample, and cavitations mostly of intergranular cracking were exhibited in other samples (Color figure online). high-angle grain boundaries (not twins), with the highest GND density around the intragranular carbides is densities associated with triple junctions. evident, and it is most obvious for the serrated case At the highest test temperature used, Figure 14, the where the carbides are larger and aligned along slip GND density after 5 pct strain is reduced compared to bands. lower temperature tests, and the accumulation of GND density near grain boundaries is less marked. There is E. X-ray Tomography Technique little evidence of GND density accumulation near the For understanding the creep behavior in greater triple junction toward the center of the map for the detail, the X-ray tomography technique was employed nonserrated case, although the density is increased near to characterize cavitation damage in each condition and the intersection of a twin boundary with a general microstructure. 3D reconstructed surfaces of each high-angle grain boundary. Accumulation of increased 4332—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 9—(a), (b), (e) and (f) displayed GND density maps obtained from 815 C/70 MPa and 900 C/40 MPa for both materials, along with the inverse pole figure map corresponding to Y-axis, see (c), (d), (g) and (h). Dislocation substructures are clearly revealed in each map. Cell sizes increase with stress level. In particular, nonserrated specimens display smaller cell size than serrated ones in both conditions (Color figure online). fracture tip showing cavitation (labeled red) distribu- damage caused by cavitation quantitatively, two metrics tions across the specimen are displayed in Figures 15(a) are considered here: cavitation frequency and total through (c) for nonserrated, and (d) through (f) for volume of cavitation. Cavitation frequency—con- serrated microstructures. Figure 16 shows the number tributed predominantly by small cavities—increased density of cavitation vs distance from fracture tip. In dramatically from 700 C to 900 C, i.e., from 714 to general, cavities are at higher density and coarser near 3813 counts for the nonserrated and from 46 to 6194 the fracture tip, and the damage declines slightly over counts for the serrated. On the other hand, cavitation distance from its fracture surface. To describe the volume are contributed mostly by larger cavities. It METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4333 Fig. 10—Carbide distribution within grain interior and in vicinity of grain boundaries. ImageJ has been utilized for outlining carbides. (a) and (c) are the backscattered image of the nonserrated and serrated material from creep tests interrupted at 5 pct strain. (b) and (d) are the graphs with outlined carbides. Carbides are fine and evenly distributed in nonserrated grain boundary material. By contrast, carbides are coarser and elongated in serrated material, predominantly distributed near grain boundary region with specific orientations. implies the degree of porosity in the material and distributions affect the movement and accumulation of effectiveness of small-crack coalescence, which reduced dislocations during creep. gauge area and accelerated the fracture. Thus, this is considered to be the better metric of cavitation damage. A. Carbide Distribution and Its Evolution For the interrupted condition, the carbide distribu- tions were distinct for the two microstructures—homo- IV. DISCUSSION geneous fine carbides in the nonserrated case and The heat treatments applied to Inconel 600 in the heterogeneous, coarse carbides in the serrated case. This study have successfully generated two distinctive grain is a phenomenon caused by a combination of the heat boundary morphologies, i.e., serrated and nonserrated, treatment and subsequent creep deformation. In the while maintaining the other microstructural aspects conventional quenched nonserrated case, carbon atoms same, thus allowing the isolation of the effects of are relatively uniformly dispersed, so a uniform disper- serration on creep deformation. In summary, the results sion of carbides during thermal exposure in testing was clearly demonstrate three key findings. First, for both obtained. The nonconventional slow cooling leads to strong segregation of carbon atoms to grain boundaries microstructures, a transition from transgranular-to- and triggered grain boundary serration with intergran- intergranular fracture was observed with the increasing ular carbides. In addition, the carbides in serrated temperature and the decreasing stress. This transition samples are seen to grow in a preferred crystallographic was offset to higher temperatures by grain boundary serration. Second, grain boundary serration has demon- orientation. This is demonstrated by means of diffrac- strated great impact on creep properties in Inconel 600; tion patterns obtained by EBSD. Figure 11 demon- in particular, it promotes enhancement in creep rupture strates that carbide growth direction is aligned with the life at high temperatures. However, at the lowest test f11 1g slip planes. temperature, no measurable difference was found. Investigation of the carbide distribution in the ser- Third, the carbide distribution evolves significantly rated case was carried out under all test conditions. during creep testing and is distinctively different in SEM image intensities were used to segment carbide the two microstructures. These different carbide particles and their centers of mass were recorded relative 4334—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 11—Serrated grain boundary sample interrupted by 5 pct creep strain at (a) 815 C/70 MPa and (c) 900 C/40 MPa. Carbides are outlined in (b) and (d) where it demonstrated preferred growth orientations. Unit cell orientations were extracted from EBSD data and stitched next to each grain. f11 0g directions on h11 1i planes are labeled with yellow arrows. These directions are parallel to the carbides. It suggests carbide growth is facilitated by mass transport on slip systems (Color figure online). to the positions of the grain boundary. The number explicitly evaluate the effects of grain boundary serra- densities of carbides as a function of distance from the tion and precipitates that formed in correspondence is ground boundary were averaged over three representa- not feasible. Admittedly, the coexistence of precipitate/ tive boundaries and are shown in Figure 17.No serration may not always be beneficial, as the final carbides were observed at 700 C, but carbides were properties might be sacrificed by way of losing some frequently observed at higher test temperatures. The intragranular strength. width of the carbide region is more extended at greater The carbide formation near grain boundaries might temperature, around 35 lm at 815 C and 85 lmat be facilitated by two possible mechanisms. One is due to 900 C. With the increasing temperature, the carbides faster diffusion path (pipe diffusion) where the nucle- also become lowered in frequency; this suggests the ation sites are provided by dislocations. Another pos- carbides formed in the vicinity of grain boundary sible mechanism may be mass transport by dislocation 0 [40] regions are originally from intergranular carbides. climb, analogous to c phase rafting, when a gliding Furthermore, magnified images of initial creep curves dislocation being trapped by an intergranular carbide in are shown in Figure 5. It is interesting to note that the first grain, it climbs at matrix/carbide interfaces and nonserrated samples possess lower minimum creep rates eventually escapes. The latter is believed to be dominat- in the beginning, despite their lower final creep life. ing during the process, because with decreasing stress, From the investigation on carbide distribution under slip bands are less pronounced from 170 to 40 MPa, as interrupted conditions, this is believed to be contributed for the number of nucleation sites. In contrast, slip by in situ precipitation of smaller and higher densities of transfer becomes easier, as evidenced by the GND intragranular type carbides that provided precipitation density maps. Therefore, at higher temperatures, slip strengthening. On the contrary, much of carbon in transfer is thought to require less effort despite the serrated samples was consumed via precipitation of pinning carbides. grain boundary carbides, and less intragranular carbides were able to form. It is also interesting to note the B. Dislocation Accumulation coexistence of grain boundary precipitates and serration is always the case for the present study, which agrees Creep testing was shown to induce dislocation cell well with other findings in the literature. The phe- structures in all cases. Average cell sizes were smaller for nomenon of serration is de facto always associated with the higher-stress/lower-temperature tests, and were con- some types of grain boundary precipitates. Therefore, to sistently slightly smaller for the serrated microstructure. METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4335 Fig. 12—GND density maps of interrupted specimens of (a) nonserrated and (b) serrated specimens at 700 C/170 MPa, together with inverse pole figure and schmid factor map, which shows the locations of the grains. GND density values are accumulated near grain boundary in serrated sample, whereas slip transfer seems to be fairly easy for nonserrated specimen (Color figure online). For the tests interrupted at 5 pct strain, the dislocation hence are more prone to slip. In the case of nonserrated cell structures had not yet formed fully, but there are grain boundaries at 815 C, the high GND density was signs of this beginning preferentially in regions close to localized at regions where notable schmid factor differ- grain boundaries and triple junctions. ences were found, i.e., the triple junction region in this At 700 C, the GND density is more localized around case. This is understandbly due to the incompatibility of grain boundary regions with the serrated architecture, plasticity of different grains. However, despite the which suggests a role of carbides in preventing slip differences in the serrated case, GND density localiza- transfer. At 815 C, for nonserrated samples, the inter- tion seems mostly independent of schmid values. It is sections, between three grain boundaries or between again more pronounced under 900 C conditions, where twin-grain boundaries, were shown to be associated with the twin-grain boundary intersection regions in the high GND density. In contrast, the serrated samples nonserrated case share a similar value of schmid factor, showed no particular dislocation accumulation in such although this shows a severe GND density concentra- grain intersections; instead, they show entanglement of tion. In contrast, the twin-grain boundary intersection GND with coarse carbides near the whole grain shows more difference in schmid values, but was found boundary region. At 900 C, such intersections were to have a negligible accumulation of GND densities. more pronounced where twin boundary and grain The high GND density intersections of this kind in boundary intersect. However, no GND accumulations nonserrated microstructures are believed to act as were found at twin-grain boundary intersections in dislocation sources during creep, and possibly cavita- serrated grain boundary samples under any conditions. tion-initiation sites. Therefore, further SEM study was Schmid factor maps displayed next to each GND conducted near twin-grain boundary intersections in density map further supports the observation. In both microstructures. As confirmed by a number of Figures 12, 13, and 14, lighter grains are the ones different locations, triple junctions and twin-grain possessing higher magnitude of schmid factors, and boundary intersections were found to be prone to 4336—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 13—GND density maps of interrupted specimens of (a) nonserrated and (b) serrated specimens at 815 C/70 MPa, together with inverse pole figure and schmid factor map. Dislocation cell structures has initiated. GND density were highest at triple junction and twin-grain boundary intersection in nonserrated specimen, in contrast, GND density were highest near entire grain boundary region where carbides located (Color figure online). cavitation in the nonserrated case. As shown in number of internal cracks, which mostly correspond to Figure 18, small cracks were frequently found to be grain boundaries adjacent to triple junctions or twin/ associated with those intersections—either twin-grain grain boundary intersections. boundary intersections or triple junctions. However, no such cracks/voids were observed in any interrupted tests C. Assessment of Degree of Cavitation on serrated grain boundary architectures, in agreement with the HR-EBSD results. Hence, nonserrated grain In terms of total cavitation volume, for nonserrated boundaries are more prone to initiate cavities, particu- grain boundary specimens, the extent measured is larly near intersections between twin and grain bound- always remarkably higher—almost twice that in the aries, where high stress concentrations are likely to build serrated ones—more damage due to cavitation is up. expected. Furthermore, an interesting point is that Previous study by Carter et al. also found the same serrated specimens are more frequently shown in tendency of such intersections using a Digital Image 815 C to 900 C conditions—1690 and 6154 counts [41] Correlation (DIC) approach. The reason why ser- (serrated) compared to 1053 and 3813 counts (nonser- rated grain boundaries do not suffer from strain rated)—which is primarily contributed by the small size accumulation probably relates to the presence of coarse cavities. This is believed mostly contributed by increased carbides near and on grain boundaries. See, for example, actual grain boundary length due to the nature of Figure 14, where dislocation loops were identified where serration that provided more cavity nucleation sites. In the carbides are located. This suggests the Orowan-loop- summary, serrated specimens have more but smaller ing and related mechanisms are operative where coarser cavities, whereas nonserrated ones have fewer but larger carbides are more easily bowed around and circum- ones. Moreover, the size and frequency in both vented. In reflection of this phenomenon, the EBSD microstructures reflects a relation of direct proportion- image of fractured cross sections further supports the ality over significant variations in temperatures and stresses, see Figure 19. The cavitation volume is plotted observation. Figures 7 and 8 both display a large METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4337 Fig. 14—GND density maps of interrupted specimens of (a) nonserrated and (b) serrated specimens at 900 C/40 MPa, together with inverse pole figure and schmid factor map. Dislocation cell structures started to grow. GND density were again accumulated at twin-grain boundary intersection for nonserrated specimen. Serrated specimen shows only higher GND density near carbides (Color figure online). against frequency, so the slope represents average cavity stresses is considered here. The X-ray computed tomog- volume is consistent in all conditions, 1:6  10 and raphy makes it clear that all the samples exhibit 4 3 considerable porosity. This means that the more the 6:5  10 mm for nonserrated and serrated, respec- cavitation damage a material is experiencing, the higher tively. It is then deduced that the average size of the true stress that is experienced due to a reduction in cavitation is independent of the creep conditions, but a the load-carrying cross section. Since the cavitation function of grain boundary morphologies. A further volume is overwhelmingly higher in the nonserrated point to be made here is that the growth behavior is microstructure, the true stresses are indeed larger significantly depressed due to serration. Rupture life and compared with the serrated case. ductility improvements are gained by inhibiting cavity A comprehensive study on fracture and deformation link-ups. The current study is the first quantitative mechanisms has been completed by Frost & Ashby on cavitation analysis completed in 3D on understanding [42,43] nichrome (Ni-18Cr wt pct), a very similar alloy, the serration effect, which has allowed greater insight which can be used to rationalize the fracture mode in into cavitation growth resistance. the present material. The fracture mode determined The cell structures developed at 815 C and 900 C under 700 C/170 MPa condition is a mixture of further support the argument above on cavitation transgranular and intergranular types, consistent with volume and corresponding damage that it represents. the power law regime. Under the 815 C to 900 C The cell sizes are known to be inversely related to the conditions, the deformation is still power-law con- stress that specimen experiences; therefore, an increase trolled, but fracture mode is largely intergranular. This in cell size reflects lower stress applied, which is is in agreement with current observations made by consistent in both microstructures. Now, while we fractography. However, the serrated microstructure consider the cases in serrated and nonserrated mor- always exhibits more transgranular features than the phologies, we observe that under the same condition, the nonserrated case at lower test temperatures, particu- serrated microstructure always has a larger cell size than larly at 815 C. The observation suggests a higher that of the nonserrated cases. The difference in true 4338—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 15—3D reconstruction of fractured creep specimens from side and top views showing cavitation distribution, where red represents cavities. Two metrics were used for damage assessment, cavitation counts (frequency) and volume, which are labeled next to each specimen. The former takes account of small size cavitation, the latter takes account of ultimate damage. (a through c) displayed nonserrated samples and (d through f) for serrated ones. In higher temperature and lower stress regime, cavitation content is increased. Serration has strong resistance in cavitation propagation, as demonstrated with reduced cavitation volume by inhibiting coalescence of small cracks (Color figure online). Fig. 16—Cavitation number density vs distance from fracture tip. (a) and (b) represents a specific example of such a fracture tip and its cavitation profile over distance. (c) Represents cavitation profile over distance for both microstructures at all creep conditions (Color figure online). transition temperature is caused by serration due to could be active regardless of grain boundary morphol- cavitation resistance. It is well understood in the ogy, and hence, no marked change in creep life was literature—and also verified in the current creep observed. Consequently, with the increasing tempera- tests—that the cavitation limits ductility significantly. ture and decreasing stress, a transition to intergranular Although some intergranular features are determined fracture emerged in the present experiment, where to be active during creep, impact of cavitation is cavitation plays a more influential role. limited at 700 C, (the most affective damaging mech- To summarize, in the current study, grain boundary anisms being contributed by plasticity), obviously due serration has been proven to enhance intergranular to the significant creep strain. Twelve-slip systems cracking resistance, both by stagnation of cavitation METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4339 Fig. 17—Carbide distribution in relation to distance from grain boundary in the serrated case. (a) and (b) represent a specific example of a grain boundary used for analysis taken from interrupted specimen tested under 900 C/40 MPa condition. Its number density frequency vs distance is represented in (b). Accumulated carbide distribution measured for several grain boundaries for 815 C and 900 C are exhibited in (c) and (d), respectively (Color figure online). nucleation and its subsequent growth. Serration has microstructures with identical grain size and twin conferred no beneficial effect on transgranular defor- boundary fraction. mation, but it is not detrimental either. 2. The creep responses of serrated and nonserrated microstructures were assessed. A transition in deformation mode from mainly transgranular at 700 C/170 MPa to intergranular-dominant crack- V. SUMMARY AND CONCLUSIONS ing at 900 C/40 MPa was observed, when the creep ductility was much reduced in comparison. Rupture The following specific conclusions can be drawn from lives were enhanced by serrated grain boundaries in this study. the high-temperature/low-stress regimes. 1. In alloy Inconel 600, the degree of grain boundary 3. Under the 815 C/70 MPa and 900 C/40 MPa serration is shown to be sensitive to the details of conditions, serrations improved the creep life by heat treatment and in particular, to the rate of about 40 pct; however, no marked improvement cooling beyond the carbide solvus temperature; by was observed under 700 C/170 MPa. Thus, ser- varying this, significant differences in the degree of rated grain boundaries conferred no beneficial serration occur. Hence, it demonstrated that it is effects when transgranular-type deformation is possible to develop microstructures of different dominating, but were more effective for intergran- degrees of serration, but also of other ular-type fractures. 4340—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 18—Typical cavities observed in nonserrated material during interrupted creep test under (a) and (b) 815 C/70 MPa and (c) and (d) 900 C/ 40 MPa conditions, at different magnifications. It shows that triple junction and twin-grain boundary intersections are the nucleation points of cavities. 5. Cavitation damage becomes more severe with the increasing temperature. Under the higher-tempera- ture conditions employed, serrated samples produce a large number of small cavities, but with small cavitation volumes overall; nonserrated samples produce larger cavities and bigger cavitation vol- umes in total. Thus, the growth of cavities is greatly suppressed by serration. 6. Cavitation size in Inconel 600 is sensitive to microstructure; its volume and frequency are directly proportional regardless of the creep condi- tions, for both nonserrated and serrated grain boundary materials. However, average cavitation volume that can be obtained—determined by the slope of this relation—is related to grain boundary morphology. 7. Subgrain boundary cell dislocation structures were characterized and measured using cross-correla- tion-method-based HR-EBSD. The cells were larger Fig. 19—Cavitation volume against frequency curve for both grain for serrated rather than nonserrated samples under boundary morphology materials. The cavity volume and frequency the same creep condition. Cavitation had an impact are directly proportional, and the slope represents the average on cell sizes by increasing porosity, which eventually cavitation volume within each microstructure. The cavity volume is increased the true stress thus accelerating the creep independent of creep conditions but a function of microstructures. rate. 8. The current study has confirmed life enhancement 4. Under the 815 C/70 MPa and 900 C/40 MPa in time-dependent intergranular deformation on the conditions, twin-grain boundary intersections have account of grain boundary serration. The knowl- been shown—consistent with their high GND edge generated is believed to be transferable to a densities—to be the initiation points for dislocation wide class of engineering alloys that are being used sources and cavitation. Serrated grain boundaries nowadays; with simple alteration in heat treatment, do not suffer so greatly from this effect and appear the properties of an alloy can be re-engineered. to present greater resistance to cavity formation. METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4341 16. J.G. Yoon, H.W. Jeong, Y.S. Yoo, and H.U. Hong: Mater. ACKNOWLEDGMENTS Charact., 2015, vol. 101, pp. 49–57. 17. J. Choi, J. Lee, J. Lee, H. Hong, and D. Kim: Korean J. Met. The use of facilities funded by EPSRC Grants EP/ Mater., 2015, vol. 53, pp. 1–12. M02833X/1 and EP/J013501/1 is gratefully acknowl- 18. J. Beddoes and W. Wallace: Metallography, 1980, vol. 13, edged. Roger Reed acknowledges financial support pp. 185–94. from EPSRC Grant EP/M005607/01. We would like 19. G. Van Drunen, J. Liurdi, J. Lib. Wallace, and T. Terada: Con- to show our gratitude to Junliang Liu and Dr Andrew ference on Advanced Fabrication Processes. 20. H. Loyer Danflou, M. Marty, and A. Walder: Superalloys, 1992, Lui for their assistance with Avizo software and X-ray pp. 63–72. computed tomography setup. 21. D. Rice, P. Kantzos, B. Hann, J. Neumann, and R. Helmink: Superalloys 2008, pp. 139 – 147. 22. A. Wisniewski and J. Beddoes: Mater. Sci. Eng. A, 2009, vol. 510, OPEN ACCESS 511, pp. 266–72. 23. H.Y. Li, J.F. Sun, M.C. Hardy, H.E. Evans, S.J. Williams, This article is distributed under the terms of the T.J.A. Doel, and P. Bowen: Acta Mater., 2015, vol. 90, pp. 355– Creative Commons Attribution 4.0 International 24. P. Kontis, H.A. Mohd Yusof, S. Pedrazzini, M. Danaie, K.L. License (http://creativecommons.org/licenses/by/4.0/), Moore, P.A.J. Bagot, M.P. Moody, C.R.M Grovenor, and R.C. which permits unrestricted use, distribution, and Reed: Acta Mater., 2016, vol. 103, pp. 688–99. reproduction in any medium, provided you give 25. P. Kontis, E. Alabort, D. Barba, D.M. Collins, A.J. Wilkinson, and R.C. Reed: Acta Mater., 2017, vol. 124, pp. 489–500. appropriate credit to the original author(s) and the 26. E. Alabort, D. Barba, S. Sulzer, M. Liner, N. Petrinic, and R.C. source, provide a link to the Creative Commons Reed: Acta Mater., 2018. license, and indicate if changes were made. 27. A.C. Yeh, K.W. Lu, C.M. Kuo, H.Y. Bor, and C.N. Wei: Mater. Sci. Eng. A, 2011, vol. 530, pp. 525–29. 28. J.W. Lee, D.J. Kim, and H.U. Hong: Mater. Sci. Eng. A, 2015, vol. 625, pp. 164–68. REFERENCES 29. Y.Q. Chen, E. Francis, J. Robson, M. Preuss, and S.J. Haigh: Acta 1. J.M. Larson: Metall. Mater. Trans. A, 1976, vol. 7A, Mater., 2015, vol. 85, pp. 199–206. pp. 1497–1502. 30. R.J. Mitchell, M.C. Hardy, M. Preuss, and S. Tin: Superalloys, 2. O Miyagawa, M Yamanoto, and M Kobayashi: Superalloys, 1976, 2004, pp. 361–70. pp. 245–54. 31. R.C. Reed: The Superalloys: Fundamentals and Applications, 3. A.K. Koul and G.H. Gessinger: Acta Metall., 1983, vol. 31 (7), Cambridge University Press, Cambridge, 2006. pp. 1061–69. 32. K.K. Mehta, R. Mitra, and S. Chawla: Mater. Sci. Eng. A, 2014, 4. P.R. Bhowal, E.F. Wright, and E.L. Raymond: Metall. Mater. vol. 611, pp. 280–89. Trans. A, 1990, vol. 21A, pp. 1709–17. 33. J.R. Davis: ASM Specialty Handbook Heat-Resistant Material. 5. M.F. Henry, Y.S. Yoo, D.Y. Yoon, and J. Choi: Metall. Mater. ASM International, 1997. Trans. A, 1993, vol. 24A, pp. 1733–43. 34. A.J. Wilkinson, G. Meaden, and D.J. Dingley: Ultramicroscopy, 2006, vol. 106 (4), pp. 307–13. 6. H. Loyer Danflou, M. Macia, T.H. Sanders, and T. Khan: Su- 35. A.J. Wilkinson and D. Randman: Philos. Mag., 2010, vol. 90 (9), peralloys 1996, pp. 119–27. 7. H.U. Hong, I.S. Kim, B.G. Choi, M.Y. Kim, and C.Y. Jo: Mater. pp. 1159–77. Sci. Eng. A, 2009, vol. 517 (12), pp. 125–31. 36. T.B. Britton and A.J. Wilkinson: Ultramicroscopy, 2012, vol. 114, 8. A.R.P. Singh, S. Nag, J.Y. Hwang, G.B. Viswanathan, J. Tiley, R. pp. 82–95. Srinivasan, H.L. Fraser, and R. Banerjee: Mater. Charact., 2011, 37. T.J.A. Slater, R.S. Bradley, G. Bertali, R. Geurts, S.M. vol. 62 (9), pp. 878–86. Northover, M.G. Burke, S.J. Haigh, T.L. Burnett, and P.J. 9. L. Jiang, H. Rui, H. Kou, J. Li, G. Bai, and F. Hengzhi: Mater. Withers: Sci. Rep., 2017. Sci. Eng. A, 2012, vol. 536, pp. 37–44. 38. P.S. Karamched and A.J. Wilkinson: Acta Mater., 2011, vol. 59 10. Y.S. Lim, D.J. Kim, S.S. Hwang, and H.P. Kim: Mater. Charact., (1), pp. 263–72. 2014, vol. 96, pp. 28–39. 39. J. Jiang, T. Ben Britton, and A.J. Wilkinson: Int. J. Plast., 2015, 11. M. Terner, H.U. Hong, J.H. Lee, and B.G. Choi: Int. J. Mater. vol. 69, pp. 102–17. Res., 2016, vol. 107, pp. 229–38. 40. Z. Zhu, H. Basoalto, N. Warnken, and R.C. Reed: Acta Mater., 12. A.K. Koul and R. Thamburaj: Metall. Mater. Trans. A, 1985, 2012, vol. 60 (12), pp. 4888–4900. vol. 16A, pp. 17–26. 41. J.L.W. Carter, M.W. Kuper, M.D. Uchic, and M.J. Mills: Mater. 13. R.J. Mitchell, H.Y. Li, and Z.W. Huang: J. Mater. Process. Sci. Eng. A, 2014, vol. 605, pp. 127–36. Technol., 2009, vol. 209 (2), pp. 1011–17. 42. M.F. Ashby, C. Gandhi, and D.M.R. Taplin: Acta Metall., 1979, 14. X.D. Lu, Q. Deng, J.H. Du, J.L. Qu, J.Y. Zhuang, and Z.Y. vol. 27 (5), pp. 699–729. Zhong: J. Alloys Compd., 2009, vol. 477 (1), pp. 100–03. 43. H.J. Frost and M.F. Ashby: Deformation-Mechanism Maps, The 15. X.X. Yao, Y. Fang, H.T. Kim, and J. Choi: Mater. Charact., 1997, Plasticity and Creep of Metals and Ceramics, Pregamon Press, New vol. 38 (2), pp. 97–102. York, 1982. 4342—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Metallurgical and Materials Transactions A Springer Journals

Grain Boundary Serration in Nickel-Based Superalloy Inconel 600: Generation and Effects on Mechanical Behavior

Free
19 pages

Loading next page...
 
/lp/springer_journal/grain-boundary-serration-in-nickel-based-superalloy-inconel-600-l8Emg4kuag
Publisher
Springer US
Copyright
Copyright © 2018 by The Author(s)
Subject
Materials Science; Metallic Materials; Characterization and Evaluation of Materials; Structural Materials; Surfaces and Interfaces, Thin Films; Nanotechnology
ISSN
1073-5623
eISSN
1543-1940
D.O.I.
10.1007/s11661-018-4671-7
Publisher site
See Article on Publisher Site

Abstract

TOPICAL COLLECTION: SUPERALLOYS AND THEIR APPLICATIONS Grain Boundary Serration in Nickel-Based Superalloy Inconel 600: Generation and Effects on Mechanical Behavior YUANBO T. TANG, ANGUS J. WILKINSON, and ROGER C. REED Grain boundary serration in the superalloy Inconel 600 was studied. Two microstructural variants, one with nonserrated and the other with serrated grain boundaries were generated by altering the heat-treatment conditions, while keeping other aspects of the microstructure unchanged. The effect on the creep response between 700 C and 900 C was measured, and the different failure modes and accumulated damage were quantified using high-angular resolution electron backscatter diffraction analysis in the scanning electron microscope and also by X-ray computed tomography. It is found that serration plays a more crucial role in the high-tem- perature/low-stress regime when an intergranular cracking mechanism involving cavitation is operative; here it plays a role in improving both creep life and creep ductility. Any effect of serration is less prevalent at low temperatures where transgranular failure is dominant. https://doi.org/10.1007/s11661-018-4671-7 The Author(s) 2018 [3–11] I. INTRODUCTION composition. The study of Koul & Gessinger on [3] the Inconel 738, Nimonic 115 and Nimonic 105 alloys, SERRATION of grain boundaries in metals and involving an impressive combination of experimentation alloys, for example the nickel-based superalloys of the 0 and modeling, identified the role of c migration on type studied here, is intriguing for various reasons. First, 0 serration. The role of the c phase has since been there is complexity arising from the inapplicability of the [6,12–15] confirmed by a number of researchers. But what generally accepted picture of smooth or gently curved happens in the absence of the c phase? Alloys contain- grain boundaries, as observed on the microscale. ing very few or no c -forming elements have also been Second, there is the fundamental question of why grain shown to form serrated grain boundaries, for example boundary serration arises in the first place. Third, effects Haynes 230, Inconel 690, Inconel 617, and Nimonic of serration on the mechanical behavior of these [7,10,16,17] 263. In this case, the formation mechanism for materials are not well understood. What are the these alloys involves the triggering of serration by underlying fundamental effects which cause it? At Cr-rich M C carbide. 23 6 present, no truly unequivocable explanation exists. Is the phenomenon of grain boundary serration of Detailed experimentation involving carefully controlled mere academic interest, or are there significant beneficial measurements and high-resolution characterization is effects which confer technological advantage? The needed. time-dependent deformation has been shown to be Serration of grain boundaries in nickel alloys was first improved dramatically by such features, particularly [1] [2] reported by Larson et al. and Miyagawa et al. in the creep ductility/lifetime and crack growth rate. Early 1976; slow cooling from above the c dissolution research conducted on Inconel 792 and 20-11P con- temperature was employed. Since then, attempts have firmed reduced creep rate and increased creep [1,2] been made to elucidate the formation mechanism, strength, although the postmortem analysis was details of heat treatment and the role of alloy limited to the characterization techniques available at that time. Later studies on IN738, Astroloy, PM alloy-10, Waspaloy, and RR1000 have proven similar [18–23] influences. The study up to date on STAL-15 again revealed similar trend—with both extensive experiments YUANBO T. TANG and ANGUS J. WILKINSON are with the [24–26] and modeling efforts. In addition, for alloys with Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, UK. Contact e-mail: yuanbo.tang@materials.ox.ac.uk low or no c phase, creep life and ductility are signifi- [7,27,28] ROGER C. REED is with the Department of Materials, University cantly enhanced. In summary, grain boundary of Oxford and also with the Department of Engineering Science, serration has been demonstrated to improve mechanical University of Oxford, Parks Road, Oxford OX1 3PJ, UK. properties such as creep, low cycle fatigue (LCF), and Manuscript submitted January 23, 2018. dwell fatigue remarkably in various superalloys. A few Article published online June 1, 2018 4324—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A hypotheses have been proposed for taking into account deformation modes which are operative. Attention is of the improved behavior, e.g., prevention of grain paid not only to postmortem analysis, but also to boundary sliding, increased grain boundary diffusion interrupted studies. In the final part of the paper, the path, and impedance of crack growth. However, no implications of the serration effect on the mechanical unambiguous mechanism is yet supported strongly by response of this alloy are considered with an emphasis substantial experimental evidence or modeling. The past on carbide evolution, cavitation, and dislocation sub- efforts have mostly been devoted to the serration effects structures which arise. from different angles in a qualitative manner, such as alloy composition, mechanical properties, and cooling rates, but little study has been attempted to control II. METHODOLOGY AND EXPERIMENTAL other microstructural variables at the same time. The DESCRIPTION most important missing part is that no quantitative measurements have been conducted in order to under- Inconel 600 was chosen for the current study, on stand such phenomenon. account of the relative simplicity of its chemical com- position, Table I. In this alloy, Al or Ti is present at only A further point relates to the complexity of the trace levels so that any influence of the c phase on grain situation. Are the beneficial influences really contributed by grain boundary serration alone, or are there other boundary serration cannot then arise. Moreover, refrac- changes to microstructural features due to slow cooling tory elements such as Ta, Mo, and W are also absent; which are influential? In fact, few researchers have really therefore, grain boundary MC-type carbides enriched addressed this issue, and studies on the serration effect with these elements have no contribution to microstruc- have seldom reported very well-controlled microstruc- ture; neither M C nor M C carbides could be formed 23 6 7 3 ture, obviously due to the nature of the problem. Creep from MC-carbide decomposition. Specimens were taken is one of the most complicated continuum damage from a cold-drawn and annealed bar of circular cross processes, and many factors are well known to be section with a diameter of 16 mm, in which the virgin crucial: grain size, c size, morphology and distribution, grain size was about 15 lm. twin boundary fraction, carbide type/morphology, size, etc. All are likely to be affected by cooling rate to some A. Heat Treatment extent. For example, the c size and distribution can be directly influenced by nucleation density within the c Heat treatment was carried out in two ways in order matrix, which is a function of solution-treatment con- to develop different grain boundary morphologies, i.e., [4,8,23,29–31] dition and cooling rate. The slow cooling serrated and nonserrated (straight or gently curved). A causes larger c size and sometimes a bimodal distribu- solution heat-treatment temperature of 1140 C for 2 tion. Moreover, slow cooling lengthens the total time hours was used throughout, which is significantly higher than the estimated carbide solvus temperature of required for grain growth and partial dissolution of MC [32,33] 1020 C calculated using the ThermoCalc software. carbides, which makes good control over Such solution heat treatment was employed to maintain microstructual variables even harder. Clearly, very the same level of grain size, grain distribution, and R3 carefully designed experimentation is needed. twin boundary fraction in both microstructures, which The research reported in this paper was carried out are considered to be crucial microstructural variables for keeping the above issues in mind. The nickel-based creep and other mechanical properties. The selected heat superalloy Inconel 600 is chosen for study; an advantage is then that the alloying elements Al and Ti are absent, treatment compensated for small variations in total so that any involvement of the c phase Ni ðAl; TiÞ on heat-treatment time caused by different cooling rates, the serration effect cannot then arise. Furthermore, any and it proved the possibility to develop significant grain role by Ta- or W-rich MC carbides is avoided, because boundary serration at the microstructural level. This Ta and W are absent. Instead, as will be shown, Cr-rich was achieved by cooling the material beyond the carbide carbide species become critical to the development of solvus temperature at 12 C/min, measured by N-type grain boundary serration, with the cooling rate through thermocouples, in contrast to water quenching for the carbide solvus temperature also playing a crucial nonserrated grain boundaries. Heat-treatment profiles role. Here, serrated and nonserrated grain boundaries and representative microstructures are illustrated in were first prepared and characterized using scanning Figure 1. electron microscopy, with the grain size distribution and twin boundary fractions having proven to be B. Creep Testing unchanged; the creep response was then tested. Quan- The impact of grain boundary serration on mechan- titative high-angular resolution electron backscatter ical behavior was assessed from 700 C to 900 Cby diffraction (HR-EBSD) and X-ray computed tomogra- tensile creep rupture tests. Creep test specimens were phy (XCT) are used to gain insights into the creep Table I. Nominal Composition of the Alloy Studied in Weight Percent (Ni-Base) Name Ni Cr Fe C Mn Si Co Cu S Inconel 600 74.1 15.9 9.03 0.068 0.21 0.27 0.012 0.0018 0.001 METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4325 Fig. 1—(a) and (b): Heat-treatment profiles for development of nonserrated and serrated grain boundary materials and (c) and (d) corresponding microstructures obtained from heat treatments described in (a) and (b). The curves are drawn to illustrate grain boundary morphologies only and are deliberately offset in order not to obscure the grain boundaries in the micrographs. machined from heat-treated cylindrical bars, diameters condition, a representative sample, i.e., the sample with of which were reduced from 16.4 to 4 mm after heat median rupture life among three, was first taken for treatment to avoid any effect of oxidation damage on fractography analysis. The fractured sample was then specimen gauge section. Creep tests were carried out cut by wire-based electrical discharge machining (EDM) under three conditions for both microsturctures, using from the middle of fracture tip along the axial direction. specimens of 4 mm diameter and 20 mm length (3 tests The sectioned sample was mounted and then prepared under each condition). An Instron 8862 servoelectric via standard metallography procedures, with 2 minutes machine was used, where the strain was measured using colloidal silica for finishing. The SEM is equipped with a an electromechanical actuator equipped in the lower end Bruker EBSD detector, which was used to capture of the crosshead operated at a frequency of 1 Hz with EBSD patterns at a resolution of 600  800 for maps the precision of 1 lm/h. In addition, another six tests collected at a step size of 220 nm. Furthermore, six were interrupted at 5 pct strain, to observe deformation interrupted specimens were also cut by EDM in the and fracture mechanisms under each condition to same way and scanned by HR-EBSD using the same evaluate the influence of serration. The tests were conditions. A cross-correlation-based analysis was used [34–36] carried out under different conditions of significant on the stored EBSD patterns. variations in temperature and stress: 700 C/170 MPa, 815 C/70 MPa, and 900 C/40 MPa. Temperature was D. X-ray Tomography-Based Characterization monitored using K-type thermocouples placed in con- Since the reduction in area and cracking from grain tact with the specimen. Each specimen was held at the boundaries in the crept samples were found to be testing temperature for 1 hour prior to loading. significant, X-ray computed tomography was used to quantify cavity size and its distribution. One represen- C. SEM-Based Characterization tative fracture tip was taken for each creep condition The grain structure and boundary types produced by and tomography analysis carried out using a North Star the two heat treatments were characterized using a Zeiss Imagix system at 150 kV and a current of 27 lA, which Merlin high-resolution field emission gun scanning defined a voxel size of 4.86 lm. For each sample, a electron microscope (FEG-SEM) that was also used height of roughly 10 mm was characterized using 3600 for the fractography and HR-EBSD scans. For each projections and a helical scan used to reconstruct the 3D 4326—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 2—Postmortem X-ray tomography setup and computed results—quantification of cavitation distribution. (a): X-ray technique setup. (b): Volume of the specimen reconstructed by X-ray tomography. (c) Surface reconstruction completed by Avizo with internal cavities labeled in red. (d) A cylindrical volume of 5 mm height from lower end of the fracture tip was used for cavity size and distribution analysis (Color figure online). Fig. 3—Energy dispersive X-ray spectroscopy (EDS) analysis in a region that contains a serrated grain boundary and planer carbides. (a) and (b) show the regions at two magnifications. (c): Composition maps indicate segregation of Cr and C and depletion of Fe and Ni at the grain boundary. The dark phases are determined as Cr-rich Carbide. volume. ImageJ and Avizo softwares were utilized for and without grain boundary serration. For this, quan- quantification of cavity size and distribution using tification of a 3D cylindrical volume of about 3 mm [37] histogram-based methods. Comparisons of cavity diameter and 5 mm height was selected as illustrated in size and distributions were then made for samples with Figure 2. METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4327 III. RESULTS Grain size and twin boundary fractions can have great influences on creep, and so they were measured for both A. Heat Treatment and Grain Boundary Serration microstructures using EBSD over large sampling areas Development for achieving statistically meaningful estimates. Three random regions of each sample were mapped with The profiles of heat treatment that generated each EBSD to cover over 800 grains for each microstructure. type of grain boundary morphology are shown in The EBSD data were then postprocessed using the Figures 1(a) and (b). Backscatter electron imaging in ESPRIT software, with a threshold grain boundary the FEG-SEM revealed grains and carbides due to the misorientation angle defined as 10 deg. Representative channeling effect and compositional contrast. Typical EBSD maps are illustrated in Figures 4(a) and (b) for microstructures produced from each cooling rate are the nonserrated and serrated microstructures, respec- shown in Figures 1(c) and (d); for illustration purposes, tively. Grain size distribution histograms are shown in curves highlighting the grain boundary morphologies Figures 4(c) and (d), where two distributions overlap are displaced laterally from the actual boundaries. For the nonserrated grain boundary case, no secondary with each other and both are skew with tails on the high phases were observed at the FEG-SEM resolution; grain size side. Mean values of grain size with boundary however, the serrated grain boundaries were always length fraction of R3 (twin) boundaries are given on the distribution plots and show very little variation between associated with intergranular carbides at the microscale. the two microstructures. High-resolution energy dispersive X-ray spectroscopy Postprocessed EBSD results on misorientation angles (EDS) was used to probe the chemical compositions of demonstrated that negligible low-angle (<15 deg) grain such carbides on a serrated grain boundary, and boundaries were present after heat treatment, which Figure 3 shows a typical composition map. This and suggests dislocation arrays/polygonization were mini- other similar maps determined the carbides to be mized. While the mechanism in relation to serration enriched in chromium but depleted in iron and nickel. Fig. 4—(a) and (b) Examples of inverse pole figure maps used for grain size measurements for nonserrated and serrated materials, respectively. (c) and (d) Grain size distributions for each microstructure. Average grain sizes are 376 and 395 lm; twin boundary area fractions are 64.7 and 65.3 pct, respectively, for nonserrated and serrated grain boundary materials (Color figure online). 4328—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 5—Creep data for nonserrated and serrated architectures, (a through c) under conditions of 700 C/170 MPa, 815 C/70 MPa, and 900 C/ 40 MPa. Top row: Representative creep curves for median rupture life samples. (d) demonstrated average rupture lives displayed with error bars. (e) Demonstrated average ductility displayed with error bars. Serrated grain boundary shows no creep life improvement at 700 C, but, enhancements at 815 C and 900 C. Creep ductility was increased by a moderate margin under all conditions (Color figure online). formation is out of this paper’s scope, some measure- Figure 5. The top row presents the creep curves of ments have, however, been performed on serration samples with median rupture life in each condition. The generation. The serration amplitude is found to be a bottom row displays a summary of rupture life and function of cooling rate. For the given solution treat- creep ductility with error bars representing ± one ment used here, the serration amplitude varied between standard deviation. For rupture life, the serrated grain 300 and 600 nm. In addition, although grain sizes were boundary microstructure conferred no life enhancement shown unchanged under both cooling rates, grain at 700 C. However, an increase in creep rupture life was boundary lengths were increased due to the presence measured at the higher test temperatures of 815 C and of local serrations. In total, six random serrated grain 900 C. From the average creep ductility, it is evident boundaries were analyzed using ImageJ software to that serrated grain boundary surpassed the nonserrated compare measurements along the serrated path, the under all conditions by a moderate margin. In addition, actual length to end-to-end straight-line distance. The for both microstructures, a drop in ductility was evident result shows serrations increase by 54 ± 11 pct of the at higher-temperature and lower-stress levels, which grain boundary length compared to perfectly straight suggests a change in fracture mechanism. boundaries. The same procedure was completed for Fracture surfaces of samples under each condition nonserrated case to account for curvature of bound- were characterized by FEG-SEM, for which illustrative aries. The nonserrated boundaries are also longer than results are given in Figure 6. Specimens with both straight-line paths range curvature which increased the microstructures showed similar transition behavior from length by 29 ± 8 pct of the grain boundary length. low to high temperature in deformation. A clear shift from transgranular-dominant features at low tempera- ture to intergranular-dominant cracking at higher tem- B. Creep Tests and Fractography peratures was observed. The largest difference was seen Creep curves measured at 700 C/170 MPa, 815 C/ under the 815 C/70 MPa condition, for which the 70 MPa, and 900 C/40 MPa for both nonserrated and serrated samples remained mostly transgranular, serrated grain boundaries samples are presented in whereas the nonserrated sample showed much more METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4329 Fig. 6—Fractography on representative samples under various conditions with nonserrated grain boundary, (a through c) and with serrated grain boundary, (d through f). (a): Transgranular-dominant deformation with limited signs of intergranular cracking, i.e., a few microvoids. (b) Intergranular-dominant cracking with limited transgranular deformation. (c): Intergranular cracking. (d): Transgranular-dominant deformation (e): Mixed-mode of trans/intergranular cracking. (f): Intergranular cracking. evidence of intergranular processes. This is believed to suggested by previous study on face-centered cubic [39] account for the marginal increase of ductility observed. (FCC) copper. For both microstructures, the cell size Cross sections of fracture tips under each condition increased with the decreasing stress, from 4.4 to 5.2 lm were scanned using EBSD at a low magnification. for the nonserrated case and from 4.8 to 6.8 lm for the Figures 7 and 8 demonstrate crystal orientation and serrated case. Under the same test conditions, the grain boundary type at fracture for nonserrated and nonserrated microstructure always exhibited a smaller serrated samples, respectively. Pattern quality maps cell size than that for the serrated case. were overlain with grain boundary, twin boundary, and subgrain boundary (< 5 deg) next to EBSD maps. It is D. Interrupted Creep Test clear to see the contrast of cavitation behaviors between the two microstructures, particularly under 700 Cand The evolution of the carbide distribution during the 815 C conditions, where notably more cavities were interrupted creep tests was explored using backscattered formed in the nonserrated case. In addition, twin electron (BSE) imaging of sectioned samples. Figure 10 boundaries in each case displayed higher integrity, and shows images of the nonserrated and serrated no cracks were found to propagate through them despite microstructures after creep to 5 pct strain at 900 C. the absence of carbides. Image intensity was used to segment the images and identify the carbides which are shown alongside the BSE images in Figure 10. For the nonserrated microstructure C. HR-EBSD on Fractured Specimens a relatively uniform dispersion of fine carbides develop Cross-correlation HR-EBSD analyses was performed across the grain interiors. In contrast, the carbide on sections through fractured samples that had been dispersion that develops in the serrated microstructure tested under the 815 C/70 MPa and 900 C/40 MPa is heterogeneous and shows much greater number conditions. Geometrically necessary dislocation (GND) density nearer to the grain boundaries than toward the density maps were calculated using the method center of the grains. The average carbide size is larger in described in References 35 and 38. Dislocation cell the serrated case (1000 nm) compared with the nonser- structures consisting of high GND density walls and low rated case (520 nm). The other striking feature is that for GND density interiors were developed within the grains the serrated case, the carbides are clearly dispersed along after the large strain (over 0.3) creep deformation, bands in well-defined directions within each grain. Figure 9. Cell sizes were measured for each specimen by Figure 11 compares carbide distributions in the serrated drawing 10 straight lines across each map (horizontally microstructure after testing at 900 C and 815 C and and vertically) and then counting the number of shows that the bands of carbides occur for both test intersections with cell walls. Cell walls were defined by conditions and are aligned with traces of expected 14 2 f11 1g slip systems as determined by EBSD. a minimum threshold GND density of 2  10 m , 4330—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 7—Inverse pole figures (IPF) corresponding to X-axis (a), (c), and (e) and pattern quality maps (b), (d), and (f) of nonserrated fracture tips under each creep condition. The pattern quality maps are overlain with grain boundary, twin boundary, and subgrain boundary. Cavitations exhibited were mostly of intergranular cracking in all samples. The same circular black spot is displayed both in (c) and (d), caused by surface contamination rather than cavitation (Color figure online). HR-EBSD was also conducted on sectioned samples Alignment of increased GND density regions into from the interrupted tests. At the lowest test tempera- straight linear features is less marked, and there is ture of 700 C, there were quite obvious differences evidence of dislocation cell formation near the grain between GND density distributions for the two boundaries. This suggests a more difficult slip-transfer microstructures, as shown in Figure 12. For the non- process for the serrated grain boundaries. serrated case, there are some linear bands of increased At 815 C, Figure 13, the straight linear features are GND density that are aligned with the f11 1g slip plane suppressed for both microstructures. For the nonser- traces in each grain, but there is little evidence of higher rated case, there is increased GND density and disloca- GND density near the grain boundaries suggesting slip tion cell formation near the triple junctions, and to a transfer was relatively easy. In contrast, for the serrated lesser extent, where twin boundaries intersect with case, there is very marked and obvious accumulation of high-angle grain boundaries. For the serrated case, the increased GND density near the grain boundaries. GND density is markedly higher in regions close to METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4331 Fig. 8—Inverse pole figures (IPF) corresponding to X-axis (a), (c), and (e) and pattern quality maps (b), (d), and (f) of serrated fracture tips under each condition. The pattern quality maps are overlain with grain boundary, twin boundary, and subgrain boundary. Little cavitations were exhibited in 700 C sample, and cavitations mostly of intergranular cracking were exhibited in other samples (Color figure online). high-angle grain boundaries (not twins), with the highest GND density around the intragranular carbides is densities associated with triple junctions. evident, and it is most obvious for the serrated case At the highest test temperature used, Figure 14, the where the carbides are larger and aligned along slip GND density after 5 pct strain is reduced compared to bands. lower temperature tests, and the accumulation of GND density near grain boundaries is less marked. There is E. X-ray Tomography Technique little evidence of GND density accumulation near the For understanding the creep behavior in greater triple junction toward the center of the map for the detail, the X-ray tomography technique was employed nonserrated case, although the density is increased near to characterize cavitation damage in each condition and the intersection of a twin boundary with a general microstructure. 3D reconstructed surfaces of each high-angle grain boundary. Accumulation of increased 4332—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 9—(a), (b), (e) and (f) displayed GND density maps obtained from 815 C/70 MPa and 900 C/40 MPa for both materials, along with the inverse pole figure map corresponding to Y-axis, see (c), (d), (g) and (h). Dislocation substructures are clearly revealed in each map. Cell sizes increase with stress level. In particular, nonserrated specimens display smaller cell size than serrated ones in both conditions (Color figure online). fracture tip showing cavitation (labeled red) distribu- damage caused by cavitation quantitatively, two metrics tions across the specimen are displayed in Figures 15(a) are considered here: cavitation frequency and total through (c) for nonserrated, and (d) through (f) for volume of cavitation. Cavitation frequency—con- serrated microstructures. Figure 16 shows the number tributed predominantly by small cavities—increased density of cavitation vs distance from fracture tip. In dramatically from 700 C to 900 C, i.e., from 714 to general, cavities are at higher density and coarser near 3813 counts for the nonserrated and from 46 to 6194 the fracture tip, and the damage declines slightly over counts for the serrated. On the other hand, cavitation distance from its fracture surface. To describe the volume are contributed mostly by larger cavities. It METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4333 Fig. 10—Carbide distribution within grain interior and in vicinity of grain boundaries. ImageJ has been utilized for outlining carbides. (a) and (c) are the backscattered image of the nonserrated and serrated material from creep tests interrupted at 5 pct strain. (b) and (d) are the graphs with outlined carbides. Carbides are fine and evenly distributed in nonserrated grain boundary material. By contrast, carbides are coarser and elongated in serrated material, predominantly distributed near grain boundary region with specific orientations. implies the degree of porosity in the material and distributions affect the movement and accumulation of effectiveness of small-crack coalescence, which reduced dislocations during creep. gauge area and accelerated the fracture. Thus, this is considered to be the better metric of cavitation damage. A. Carbide Distribution and Its Evolution For the interrupted condition, the carbide distribu- tions were distinct for the two microstructures—homo- IV. DISCUSSION geneous fine carbides in the nonserrated case and The heat treatments applied to Inconel 600 in the heterogeneous, coarse carbides in the serrated case. This study have successfully generated two distinctive grain is a phenomenon caused by a combination of the heat boundary morphologies, i.e., serrated and nonserrated, treatment and subsequent creep deformation. In the while maintaining the other microstructural aspects conventional quenched nonserrated case, carbon atoms same, thus allowing the isolation of the effects of are relatively uniformly dispersed, so a uniform disper- serration on creep deformation. In summary, the results sion of carbides during thermal exposure in testing was clearly demonstrate three key findings. First, for both obtained. The nonconventional slow cooling leads to strong segregation of carbon atoms to grain boundaries microstructures, a transition from transgranular-to- and triggered grain boundary serration with intergran- intergranular fracture was observed with the increasing ular carbides. In addition, the carbides in serrated temperature and the decreasing stress. This transition samples are seen to grow in a preferred crystallographic was offset to higher temperatures by grain boundary serration. Second, grain boundary serration has demon- orientation. This is demonstrated by means of diffrac- strated great impact on creep properties in Inconel 600; tion patterns obtained by EBSD. Figure 11 demon- in particular, it promotes enhancement in creep rupture strates that carbide growth direction is aligned with the life at high temperatures. However, at the lowest test f11 1g slip planes. temperature, no measurable difference was found. Investigation of the carbide distribution in the ser- Third, the carbide distribution evolves significantly rated case was carried out under all test conditions. during creep testing and is distinctively different in SEM image intensities were used to segment carbide the two microstructures. These different carbide particles and their centers of mass were recorded relative 4334—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 11—Serrated grain boundary sample interrupted by 5 pct creep strain at (a) 815 C/70 MPa and (c) 900 C/40 MPa. Carbides are outlined in (b) and (d) where it demonstrated preferred growth orientations. Unit cell orientations were extracted from EBSD data and stitched next to each grain. f11 0g directions on h11 1i planes are labeled with yellow arrows. These directions are parallel to the carbides. It suggests carbide growth is facilitated by mass transport on slip systems (Color figure online). to the positions of the grain boundary. The number explicitly evaluate the effects of grain boundary serra- densities of carbides as a function of distance from the tion and precipitates that formed in correspondence is ground boundary were averaged over three representa- not feasible. Admittedly, the coexistence of precipitate/ tive boundaries and are shown in Figure 17.No serration may not always be beneficial, as the final carbides were observed at 700 C, but carbides were properties might be sacrificed by way of losing some frequently observed at higher test temperatures. The intragranular strength. width of the carbide region is more extended at greater The carbide formation near grain boundaries might temperature, around 35 lm at 815 C and 85 lmat be facilitated by two possible mechanisms. One is due to 900 C. With the increasing temperature, the carbides faster diffusion path (pipe diffusion) where the nucle- also become lowered in frequency; this suggests the ation sites are provided by dislocations. Another pos- carbides formed in the vicinity of grain boundary sible mechanism may be mass transport by dislocation 0 [40] regions are originally from intergranular carbides. climb, analogous to c phase rafting, when a gliding Furthermore, magnified images of initial creep curves dislocation being trapped by an intergranular carbide in are shown in Figure 5. It is interesting to note that the first grain, it climbs at matrix/carbide interfaces and nonserrated samples possess lower minimum creep rates eventually escapes. The latter is believed to be dominat- in the beginning, despite their lower final creep life. ing during the process, because with decreasing stress, From the investigation on carbide distribution under slip bands are less pronounced from 170 to 40 MPa, as interrupted conditions, this is believed to be contributed for the number of nucleation sites. In contrast, slip by in situ precipitation of smaller and higher densities of transfer becomes easier, as evidenced by the GND intragranular type carbides that provided precipitation density maps. Therefore, at higher temperatures, slip strengthening. On the contrary, much of carbon in transfer is thought to require less effort despite the serrated samples was consumed via precipitation of pinning carbides. grain boundary carbides, and less intragranular carbides were able to form. It is also interesting to note the B. Dislocation Accumulation coexistence of grain boundary precipitates and serration is always the case for the present study, which agrees Creep testing was shown to induce dislocation cell well with other findings in the literature. The phe- structures in all cases. Average cell sizes were smaller for nomenon of serration is de facto always associated with the higher-stress/lower-temperature tests, and were con- some types of grain boundary precipitates. Therefore, to sistently slightly smaller for the serrated microstructure. METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4335 Fig. 12—GND density maps of interrupted specimens of (a) nonserrated and (b) serrated specimens at 700 C/170 MPa, together with inverse pole figure and schmid factor map, which shows the locations of the grains. GND density values are accumulated near grain boundary in serrated sample, whereas slip transfer seems to be fairly easy for nonserrated specimen (Color figure online). For the tests interrupted at 5 pct strain, the dislocation hence are more prone to slip. In the case of nonserrated cell structures had not yet formed fully, but there are grain boundaries at 815 C, the high GND density was signs of this beginning preferentially in regions close to localized at regions where notable schmid factor differ- grain boundaries and triple junctions. ences were found, i.e., the triple junction region in this At 700 C, the GND density is more localized around case. This is understandbly due to the incompatibility of grain boundary regions with the serrated architecture, plasticity of different grains. However, despite the which suggests a role of carbides in preventing slip differences in the serrated case, GND density localiza- transfer. At 815 C, for nonserrated samples, the inter- tion seems mostly independent of schmid values. It is sections, between three grain boundaries or between again more pronounced under 900 C conditions, where twin-grain boundaries, were shown to be associated with the twin-grain boundary intersection regions in the high GND density. In contrast, the serrated samples nonserrated case share a similar value of schmid factor, showed no particular dislocation accumulation in such although this shows a severe GND density concentra- grain intersections; instead, they show entanglement of tion. In contrast, the twin-grain boundary intersection GND with coarse carbides near the whole grain shows more difference in schmid values, but was found boundary region. At 900 C, such intersections were to have a negligible accumulation of GND densities. more pronounced where twin boundary and grain The high GND density intersections of this kind in boundary intersect. However, no GND accumulations nonserrated microstructures are believed to act as were found at twin-grain boundary intersections in dislocation sources during creep, and possibly cavita- serrated grain boundary samples under any conditions. tion-initiation sites. Therefore, further SEM study was Schmid factor maps displayed next to each GND conducted near twin-grain boundary intersections in density map further supports the observation. In both microstructures. As confirmed by a number of Figures 12, 13, and 14, lighter grains are the ones different locations, triple junctions and twin-grain possessing higher magnitude of schmid factors, and boundary intersections were found to be prone to 4336—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 13—GND density maps of interrupted specimens of (a) nonserrated and (b) serrated specimens at 815 C/70 MPa, together with inverse pole figure and schmid factor map. Dislocation cell structures has initiated. GND density were highest at triple junction and twin-grain boundary intersection in nonserrated specimen, in contrast, GND density were highest near entire grain boundary region where carbides located (Color figure online). cavitation in the nonserrated case. As shown in number of internal cracks, which mostly correspond to Figure 18, small cracks were frequently found to be grain boundaries adjacent to triple junctions or twin/ associated with those intersections—either twin-grain grain boundary intersections. boundary intersections or triple junctions. However, no such cracks/voids were observed in any interrupted tests C. Assessment of Degree of Cavitation on serrated grain boundary architectures, in agreement with the HR-EBSD results. Hence, nonserrated grain In terms of total cavitation volume, for nonserrated boundaries are more prone to initiate cavities, particu- grain boundary specimens, the extent measured is larly near intersections between twin and grain bound- always remarkably higher—almost twice that in the aries, where high stress concentrations are likely to build serrated ones—more damage due to cavitation is up. expected. Furthermore, an interesting point is that Previous study by Carter et al. also found the same serrated specimens are more frequently shown in tendency of such intersections using a Digital Image 815 C to 900 C conditions—1690 and 6154 counts [41] Correlation (DIC) approach. The reason why ser- (serrated) compared to 1053 and 3813 counts (nonser- rated grain boundaries do not suffer from strain rated)—which is primarily contributed by the small size accumulation probably relates to the presence of coarse cavities. This is believed mostly contributed by increased carbides near and on grain boundaries. See, for example, actual grain boundary length due to the nature of Figure 14, where dislocation loops were identified where serration that provided more cavity nucleation sites. In the carbides are located. This suggests the Orowan-loop- summary, serrated specimens have more but smaller ing and related mechanisms are operative where coarser cavities, whereas nonserrated ones have fewer but larger carbides are more easily bowed around and circum- ones. Moreover, the size and frequency in both vented. In reflection of this phenomenon, the EBSD microstructures reflects a relation of direct proportion- image of fractured cross sections further supports the ality over significant variations in temperatures and stresses, see Figure 19. The cavitation volume is plotted observation. Figures 7 and 8 both display a large METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4337 Fig. 14—GND density maps of interrupted specimens of (a) nonserrated and (b) serrated specimens at 900 C/40 MPa, together with inverse pole figure and schmid factor map. Dislocation cell structures started to grow. GND density were again accumulated at twin-grain boundary intersection for nonserrated specimen. Serrated specimen shows only higher GND density near carbides (Color figure online). against frequency, so the slope represents average cavity stresses is considered here. The X-ray computed tomog- volume is consistent in all conditions, 1:6  10 and raphy makes it clear that all the samples exhibit 4 3 considerable porosity. This means that the more the 6:5  10 mm for nonserrated and serrated, respec- cavitation damage a material is experiencing, the higher tively. It is then deduced that the average size of the true stress that is experienced due to a reduction in cavitation is independent of the creep conditions, but a the load-carrying cross section. Since the cavitation function of grain boundary morphologies. A further volume is overwhelmingly higher in the nonserrated point to be made here is that the growth behavior is microstructure, the true stresses are indeed larger significantly depressed due to serration. Rupture life and compared with the serrated case. ductility improvements are gained by inhibiting cavity A comprehensive study on fracture and deformation link-ups. The current study is the first quantitative mechanisms has been completed by Frost & Ashby on cavitation analysis completed in 3D on understanding [42,43] nichrome (Ni-18Cr wt pct), a very similar alloy, the serration effect, which has allowed greater insight which can be used to rationalize the fracture mode in into cavitation growth resistance. the present material. The fracture mode determined The cell structures developed at 815 C and 900 C under 700 C/170 MPa condition is a mixture of further support the argument above on cavitation transgranular and intergranular types, consistent with volume and corresponding damage that it represents. the power law regime. Under the 815 C to 900 C The cell sizes are known to be inversely related to the conditions, the deformation is still power-law con- stress that specimen experiences; therefore, an increase trolled, but fracture mode is largely intergranular. This in cell size reflects lower stress applied, which is is in agreement with current observations made by consistent in both microstructures. Now, while we fractography. However, the serrated microstructure consider the cases in serrated and nonserrated mor- always exhibits more transgranular features than the phologies, we observe that under the same condition, the nonserrated case at lower test temperatures, particu- serrated microstructure always has a larger cell size than larly at 815 C. The observation suggests a higher that of the nonserrated cases. The difference in true 4338—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 15—3D reconstruction of fractured creep specimens from side and top views showing cavitation distribution, where red represents cavities. Two metrics were used for damage assessment, cavitation counts (frequency) and volume, which are labeled next to each specimen. The former takes account of small size cavitation, the latter takes account of ultimate damage. (a through c) displayed nonserrated samples and (d through f) for serrated ones. In higher temperature and lower stress regime, cavitation content is increased. Serration has strong resistance in cavitation propagation, as demonstrated with reduced cavitation volume by inhibiting coalescence of small cracks (Color figure online). Fig. 16—Cavitation number density vs distance from fracture tip. (a) and (b) represents a specific example of such a fracture tip and its cavitation profile over distance. (c) Represents cavitation profile over distance for both microstructures at all creep conditions (Color figure online). transition temperature is caused by serration due to could be active regardless of grain boundary morphol- cavitation resistance. It is well understood in the ogy, and hence, no marked change in creep life was literature—and also verified in the current creep observed. Consequently, with the increasing tempera- tests—that the cavitation limits ductility significantly. ture and decreasing stress, a transition to intergranular Although some intergranular features are determined fracture emerged in the present experiment, where to be active during creep, impact of cavitation is cavitation plays a more influential role. limited at 700 C, (the most affective damaging mech- To summarize, in the current study, grain boundary anisms being contributed by plasticity), obviously due serration has been proven to enhance intergranular to the significant creep strain. Twelve-slip systems cracking resistance, both by stagnation of cavitation METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4339 Fig. 17—Carbide distribution in relation to distance from grain boundary in the serrated case. (a) and (b) represent a specific example of a grain boundary used for analysis taken from interrupted specimen tested under 900 C/40 MPa condition. Its number density frequency vs distance is represented in (b). Accumulated carbide distribution measured for several grain boundaries for 815 C and 900 C are exhibited in (c) and (d), respectively (Color figure online). nucleation and its subsequent growth. Serration has microstructures with identical grain size and twin conferred no beneficial effect on transgranular defor- boundary fraction. mation, but it is not detrimental either. 2. The creep responses of serrated and nonserrated microstructures were assessed. A transition in deformation mode from mainly transgranular at 700 C/170 MPa to intergranular-dominant crack- V. SUMMARY AND CONCLUSIONS ing at 900 C/40 MPa was observed, when the creep ductility was much reduced in comparison. Rupture The following specific conclusions can be drawn from lives were enhanced by serrated grain boundaries in this study. the high-temperature/low-stress regimes. 1. In alloy Inconel 600, the degree of grain boundary 3. Under the 815 C/70 MPa and 900 C/40 MPa serration is shown to be sensitive to the details of conditions, serrations improved the creep life by heat treatment and in particular, to the rate of about 40 pct; however, no marked improvement cooling beyond the carbide solvus temperature; by was observed under 700 C/170 MPa. Thus, ser- varying this, significant differences in the degree of rated grain boundaries conferred no beneficial serration occur. Hence, it demonstrated that it is effects when transgranular-type deformation is possible to develop microstructures of different dominating, but were more effective for intergran- degrees of serration, but also of other ular-type fractures. 4340—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A Fig. 18—Typical cavities observed in nonserrated material during interrupted creep test under (a) and (b) 815 C/70 MPa and (c) and (d) 900 C/ 40 MPa conditions, at different magnifications. It shows that triple junction and twin-grain boundary intersections are the nucleation points of cavities. 5. Cavitation damage becomes more severe with the increasing temperature. Under the higher-tempera- ture conditions employed, serrated samples produce a large number of small cavities, but with small cavitation volumes overall; nonserrated samples produce larger cavities and bigger cavitation vol- umes in total. Thus, the growth of cavities is greatly suppressed by serration. 6. Cavitation size in Inconel 600 is sensitive to microstructure; its volume and frequency are directly proportional regardless of the creep condi- tions, for both nonserrated and serrated grain boundary materials. However, average cavitation volume that can be obtained—determined by the slope of this relation—is related to grain boundary morphology. 7. Subgrain boundary cell dislocation structures were characterized and measured using cross-correla- tion-method-based HR-EBSD. The cells were larger Fig. 19—Cavitation volume against frequency curve for both grain for serrated rather than nonserrated samples under boundary morphology materials. The cavity volume and frequency the same creep condition. Cavitation had an impact are directly proportional, and the slope represents the average on cell sizes by increasing porosity, which eventually cavitation volume within each microstructure. The cavity volume is increased the true stress thus accelerating the creep independent of creep conditions but a function of microstructures. rate. 8. The current study has confirmed life enhancement 4. Under the 815 C/70 MPa and 900 C/40 MPa in time-dependent intergranular deformation on the conditions, twin-grain boundary intersections have account of grain boundary serration. The knowl- been shown—consistent with their high GND edge generated is believed to be transferable to a densities—to be the initiation points for dislocation wide class of engineering alloys that are being used sources and cavitation. Serrated grain boundaries nowadays; with simple alteration in heat treatment, do not suffer so greatly from this effect and appear the properties of an alloy can be re-engineered. to present greater resistance to cavity formation. METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 49A, SEPTEMBER 2018—4341 16. J.G. Yoon, H.W. Jeong, Y.S. Yoo, and H.U. Hong: Mater. ACKNOWLEDGMENTS Charact., 2015, vol. 101, pp. 49–57. 17. J. Choi, J. Lee, J. Lee, H. Hong, and D. Kim: Korean J. Met. The use of facilities funded by EPSRC Grants EP/ Mater., 2015, vol. 53, pp. 1–12. M02833X/1 and EP/J013501/1 is gratefully acknowl- 18. J. Beddoes and W. Wallace: Metallography, 1980, vol. 13, edged. Roger Reed acknowledges financial support pp. 185–94. from EPSRC Grant EP/M005607/01. We would like 19. G. Van Drunen, J. Liurdi, J. Lib. Wallace, and T. Terada: Con- to show our gratitude to Junliang Liu and Dr Andrew ference on Advanced Fabrication Processes. 20. H. Loyer Danflou, M. Marty, and A. Walder: Superalloys, 1992, Lui for their assistance with Avizo software and X-ray pp. 63–72. computed tomography setup. 21. D. Rice, P. Kantzos, B. Hann, J. Neumann, and R. Helmink: Superalloys 2008, pp. 139 – 147. 22. A. Wisniewski and J. Beddoes: Mater. Sci. Eng. A, 2009, vol. 510, OPEN ACCESS 511, pp. 266–72. 23. H.Y. Li, J.F. Sun, M.C. Hardy, H.E. Evans, S.J. Williams, This article is distributed under the terms of the T.J.A. Doel, and P. Bowen: Acta Mater., 2015, vol. 90, pp. 355– Creative Commons Attribution 4.0 International 24. P. Kontis, H.A. Mohd Yusof, S. Pedrazzini, M. Danaie, K.L. License (http://creativecommons.org/licenses/by/4.0/), Moore, P.A.J. Bagot, M.P. Moody, C.R.M Grovenor, and R.C. which permits unrestricted use, distribution, and Reed: Acta Mater., 2016, vol. 103, pp. 688–99. reproduction in any medium, provided you give 25. P. Kontis, E. Alabort, D. Barba, D.M. Collins, A.J. Wilkinson, and R.C. Reed: Acta Mater., 2017, vol. 124, pp. 489–500. appropriate credit to the original author(s) and the 26. E. Alabort, D. Barba, S. Sulzer, M. Liner, N. Petrinic, and R.C. source, provide a link to the Creative Commons Reed: Acta Mater., 2018. license, and indicate if changes were made. 27. A.C. Yeh, K.W. Lu, C.M. Kuo, H.Y. Bor, and C.N. Wei: Mater. Sci. Eng. A, 2011, vol. 530, pp. 525–29. 28. J.W. Lee, D.J. Kim, and H.U. Hong: Mater. Sci. Eng. A, 2015, vol. 625, pp. 164–68. REFERENCES 29. Y.Q. Chen, E. Francis, J. Robson, M. Preuss, and S.J. Haigh: Acta 1. J.M. Larson: Metall. Mater. Trans. A, 1976, vol. 7A, Mater., 2015, vol. 85, pp. 199–206. pp. 1497–1502. 30. R.J. Mitchell, M.C. Hardy, M. Preuss, and S. Tin: Superalloys, 2. O Miyagawa, M Yamanoto, and M Kobayashi: Superalloys, 1976, 2004, pp. 361–70. pp. 245–54. 31. R.C. Reed: The Superalloys: Fundamentals and Applications, 3. A.K. Koul and G.H. Gessinger: Acta Metall., 1983, vol. 31 (7), Cambridge University Press, Cambridge, 2006. pp. 1061–69. 32. K.K. Mehta, R. Mitra, and S. Chawla: Mater. Sci. Eng. A, 2014, 4. P.R. Bhowal, E.F. Wright, and E.L. Raymond: Metall. Mater. vol. 611, pp. 280–89. Trans. A, 1990, vol. 21A, pp. 1709–17. 33. J.R. Davis: ASM Specialty Handbook Heat-Resistant Material. 5. M.F. Henry, Y.S. Yoo, D.Y. Yoon, and J. Choi: Metall. Mater. ASM International, 1997. Trans. A, 1993, vol. 24A, pp. 1733–43. 34. A.J. Wilkinson, G. Meaden, and D.J. Dingley: Ultramicroscopy, 2006, vol. 106 (4), pp. 307–13. 6. H. Loyer Danflou, M. Macia, T.H. Sanders, and T. Khan: Su- 35. A.J. Wilkinson and D. Randman: Philos. Mag., 2010, vol. 90 (9), peralloys 1996, pp. 119–27. 7. H.U. Hong, I.S. Kim, B.G. Choi, M.Y. Kim, and C.Y. Jo: Mater. pp. 1159–77. Sci. Eng. A, 2009, vol. 517 (12), pp. 125–31. 36. T.B. Britton and A.J. Wilkinson: Ultramicroscopy, 2012, vol. 114, 8. A.R.P. Singh, S. Nag, J.Y. Hwang, G.B. Viswanathan, J. Tiley, R. pp. 82–95. Srinivasan, H.L. Fraser, and R. Banerjee: Mater. Charact., 2011, 37. T.J.A. Slater, R.S. Bradley, G. Bertali, R. Geurts, S.M. vol. 62 (9), pp. 878–86. Northover, M.G. Burke, S.J. Haigh, T.L. Burnett, and P.J. 9. L. Jiang, H. Rui, H. Kou, J. Li, G. Bai, and F. Hengzhi: Mater. Withers: Sci. Rep., 2017. Sci. Eng. A, 2012, vol. 536, pp. 37–44. 38. P.S. Karamched and A.J. Wilkinson: Acta Mater., 2011, vol. 59 10. Y.S. Lim, D.J. Kim, S.S. Hwang, and H.P. Kim: Mater. Charact., (1), pp. 263–72. 2014, vol. 96, pp. 28–39. 39. J. Jiang, T. Ben Britton, and A.J. Wilkinson: Int. J. Plast., 2015, 11. M. Terner, H.U. Hong, J.H. Lee, and B.G. Choi: Int. J. Mater. vol. 69, pp. 102–17. Res., 2016, vol. 107, pp. 229–38. 40. Z. Zhu, H. Basoalto, N. Warnken, and R.C. Reed: Acta Mater., 12. A.K. Koul and R. Thamburaj: Metall. Mater. Trans. A, 1985, 2012, vol. 60 (12), pp. 4888–4900. vol. 16A, pp. 17–26. 41. J.L.W. Carter, M.W. Kuper, M.D. Uchic, and M.J. Mills: Mater. 13. R.J. Mitchell, H.Y. Li, and Z.W. Huang: J. Mater. Process. Sci. Eng. A, 2014, vol. 605, pp. 127–36. Technol., 2009, vol. 209 (2), pp. 1011–17. 42. M.F. Ashby, C. Gandhi, and D.M.R. Taplin: Acta Metall., 1979, 14. X.D. Lu, Q. Deng, J.H. Du, J.L. Qu, J.Y. Zhuang, and Z.Y. vol. 27 (5), pp. 699–729. Zhong: J. Alloys Compd., 2009, vol. 477 (1), pp. 100–03. 43. H.J. Frost and M.F. Ashby: Deformation-Mechanism Maps, The 15. X.X. Yao, Y. Fang, H.T. Kim, and J. Choi: Mater. Charact., 1997, Plasticity and Creep of Metals and Ceramics, Pregamon Press, New vol. 38 (2), pp. 97–102. York, 1982. 4342—VOLUME 49A, SEPTEMBER 2018 METALLURGICAL AND MATERIALS TRANSACTIONS A

Journal

Metallurgical and Materials Transactions ASpringer Journals

Published: Jun 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off