- Email: [email protected]
Grain structure evolution, grain boundary sliding and material flow resistance in friction welding of Alloy 718
Author’s Accepted Manuscript Grain structure evolution, grain boundary sliding and material flow resistance in friction welding of Alloy 718 F.C. Liu, T.W. Nelson www.elsevier.com/locate/msea
PII: DOI: Reference:
S0921-5093(17)31420-X https://doi.org/10.1016/j.msea.2017.10.092 MSA35695
To appear in: Materials Science & Engineering A Received date: 11 July 2017 Revised date: 24 October 2017 Accepted date: 26 October 2017 Cite this article as: F.C. Liu and T.W. Nelson, Grain structure evolution, grain boundary sliding and material flow resistance in friction welding of Alloy 718, Materials Science & Engineering A, https://doi.org/10.1016/j.msea.2017.10.092 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Grain structure evolution, grain boundary sliding and material flow resistance in friction welding of Alloy 718
F.C. Liu and T.W. Nelson
Department of Mechanical Engineering, Brigham Young University, 435 CTB, Provo, UT 84602, USA
Alloy 718 tubes were subjected to rotary friction welding to understand to the process fundamental and grain structure evolution during welding. The distribution of grain size, lowangle grain boundaries (LAGBs), and twin boundaries throughout the joints were quantitatively analyzed. The weld power, axial load, and weld temperature were monitored. The grain structure evolution during friction welding was clarified. The grain structure in the recrystallization zone (RXZ) of the weld was a result of competition between dynamic recrystallization and grain boundary sliding (GBS), which is controlled by the local deformation condition. The axial force during welding decreased with reducing the rotation rate from 1000 rpm to 500 rpm. This anomalistic phenomenon can be ascribed that a decrease in rotation rate resulted in finer grain size in the RXZ of the weld, which required lower applied force to enable GBS. Keywords: Superalloy; Recrystallization; EBSD; Friction welding; Power; Axial force 1. Introduction Alloy 718 is a nickel-based austenitic phase alloy. It can be used to make load-bearing structures at the very demanding conditions as this alloy features high strength at elevated temperatures as well as excellent corrosion, oxidation and creep resistance. Increased amount of these materials are being applied in aerospace, power station and chemical industries. Fusion welding of Alloy 718 may cause problems such as weld liquation related cracking and segregation of alloying elements [1, 2]. These problems can be avoid using solid state welding
F.C. Liu, E-mail: [email protected], Tel: +1 801-422-5189, Fax: +1 801-422-0516. T.W. Nelson, E-mail: [email protected], Tel: +1 801-422-6233, Fax: +1 801-422-0516.
1
techniques, such as friction welding (FW) or friction stir welding (FSW) to join these highperformance metals [3-7]. Of the available solid state welding methods, rotary FW (RFW) is usually the best choice for joining cylinder or tubular cross-sectioned components to other workpieces. During our RFW, one tube was rotated at a constant speed and then moved toward another part. After the two components had contacted, they moved toward each other at a constant feeding speed, resulting in a consolidate weld zone after the rotation was terminated. The RFW offer several significant advantages over conventional fusion welding: (1) No melting occurred; (2) The weld zone consisted of fine recrystallized grains rather than a coarse solidification microstructure; (3) The weld operation and monitoring can be performed automatically; (4) RFW is highly efficient and reduce the welding times. (5) RFW is environmentally friendly. Early literature has reported the RFW in the aspects of welding parameter, welding temperature, mechanical properties and as-welded microstructure [4, 6, 8-11]. It is generally established that the material in the weld zone experienced heavy plastic deformation under elevated temperature during RFW, forming a weld zone consisting of fine recrystallized grains as a result of dynamic recrystallization (DRX) [6, 8, 9, 12]. Fine grains developed among deformed grains in the thermo-mechanical affected zone (TMAZ) [6, 12]. During RFW, the strain, strain rate, and temperature vary at different locations across the weld. This made the microstructure evaluation during RFW complicated. The microstructure distribution across the RFW joints has never been quantitatively analyzed in a systematic approach. We are far from a fundamental understanding of the grain structure evolution during RFW and how the microstructure will affect the materials load-bearing capacity.
2
Alloy 718 was chosen as the material to be investigated. It is a γ″ (Ni3Nb) strengthened nickel-based alloy that combines excellent corrosion resistance and high strength at elevated temperatures. This alloy has been widely used in gas turbines, aircraft engines, extrusion dies and chemical processing containers. Fusion welding of Alloy 718 usually associates with some problems such as boron/niobium segregation, and liquation cracking [13, 14], which can be avoided in RFW. In this study, the distribution of grain size, low-angle grain boundaries (LAGBs), and twin boundaries throughout the joints were quantitatively analyzed. The weld power, axial load, and weld temperature were monitored. This provides a comprehensive understanding of grain structure evolution during friction welding and how the microstructure affects the materials loadbearing capacity during welding. 2. Experimental details The outer diameter and inner diameter of the Alloy 718 tube is 25.4 and 20.3 mm, respectively. The chemical composition (wt%) is 51.6 Ni, 18.2 Cr, 5.1 Nb, 3.28 Mo, 1.06 Ti, 0.56 Al, 0.33 V, 0.09 Mn and Fe balance. The welding was performed using a multifunction TTI-RM2 friction stir welding machine. The feed rate and feed distance for both processes are 2 mm/s and 3.6 mm, respectively. The rotation rate of weld I and II are 500 and 1000 rpm, respectively. Thermocouples were welded on the tube wall at distances of 1, 2, 3, and 4 mm away from the end of the tube to record the temperature change during welding. The weld region was removed using electrical discharge machining for optical and EBSD examination. The observation plan and EBSD scanning regions were schematically shown in Fig. 1. The EBSD sample was mechanical polished to 1 m diamond past and then followed by a vibratory polishing using 0.05 m colloidal silica. High-resolution EBSD data were obtained at
3
20 kV using FEI Helios Nanolab 600 with a TSL channel EBSD system. The average confidential index for each EBSD map is higher than 0.5. In order to eliminate spurious boundaries caused by orientation noise, boundaries with misorientation less than 2o were not considered. For twin boundary definition, a tolerance of 5 deg was allowed in both the rotation of twin plane normal and the angular between the twinning planes on either side of the boundary. 3. Results and analysis Fig. 2 shows that an increase in weld rotation rate from 500 to 1000 rpm resulted in a significant increase in weld power. As the two welds used the same amount of time, the higher weld power provides a higher energy input to the weld. Fig. 3 shows that the compression force along the axis direction increased with an increase in tool rotation rate. This phenomenon is interesting as higher energy input (higher tool rotation rate) did not lead to less deformation force under the present deformation conditions. In order to understand this anomalistic phenomenon, the microstructure throughout the joints was quantitatively analyzed. Quantitative microstructure details across the weld zone along the edge (Figs. 4a, b, and c) and center (Figs. 4d, e, and f) of the tube wall of weld I are presented. In general, the microstructure distribution is symmetric about the weld centerline (Fig. 4). The region near the weld centerline was characterized by refined grains (< 2.5 m) and low fraction of LABGs (boundary misorientation < 15%), demonstrating recrystallization has occurred (Figs. 4a and d). This region was therefore defined as recrystallization zone (RXZ). The thickness of RXZ along the edge and center of the tube wall is approximately 0.8 and 0.5 mm, respectively. The average grain size of RXZ at the edge and center of the tube wall is 2.2 and 1.9 m, respectively. The RXZ along the tube wall edge was thicker and consisted of larger grains than that along the tube wall center, indicating high energy absorption in the tube wall edge.
4
Outside of the RXZ is the TMAZ. The TMAZ contained a higher percentage of LAGBs than the base metal (BM) and the RXZ (Figs. 4b and e). Within the TMAZ, the grain size decreased linearly from the BM to the RXZ (Figs. 4a and d). The TMAZ can be divided into two distinct parts according to the distribution of LAGBs and twin boundaries (Figs. 4b, c, e, and f). In the TMAZ close to the BM, the ratio of LAGBs increased and the percentage of twin boundaries decreased. Such a microstructure change is mainly caused by plastic deformation. In the TMAZ near the RXZ, recrystallization occurred as the metal approached the RXZ, which decreased the ratio of LAGBs and increased the percentage of twin boundaries. As an annealed Alloy 718 was used, the heat affected zone (HAZ) maintained similar grain structure as the base metal, indicating the peak temperature in HAZ is lower than the recrystallization temperature of the base metal. Detailed microstructure analysis was performed on Locations I-VI (Fig. 4a) to understand the microstructure evolution during welding. Fig. 5 shows EBSD maps from Location I to IV (Indicated in Fig. 4). The points in the grain orientation maps (the left maps) were colored according to their crystallography orientation relative to the normal of observation plan. In the orientation spread maps (the right maps), each point was shaded according to the misorientation it made relative to the average orientation of the grain. LAGBs, high-angle grain boundaries (HAGBs > 15o), and twin boundaries were displayed as white, black and red lines, respectively. All the EBSD maps in this study were colored according to the same principle. Most of the grains in the BM were free of LAGBs and have low orientation spread. Annealing twins were extensively within these grains. (Fig. 5a and b). Fig. 5c and d show the grain structure of Location II (Fig. 4a), which is within the TMAZ but closer to the BM. New LAGB segments developed within the grain interiors. The LAGBs
5
segments were short, curved, disconnected and have not developed into continuous grain boundaries. A few fine recrystallized grains (which marked by arrows in Fig. 5c) were observed along the prior grain boundaries. The plastic deformation caused some of the prior twin boundaries to deviate away from the exact twin relationship with the matrix. Similar transform of boundary misorientations has been observed in friction stir welding of annealed stainless steel [15] and explained as strain induced crystallographic rotation of the twin and matrix from their original orientations [16]. As no new twins were generated in this location (Fig. 5c), the disruption of original twin boundaries led to a decrease in the ratio of twin boundaries (Fig. 4c). Enhanced orientation spread was observed in most of the deformed grains (Fig. 5d), but the stored deformation energy was insufficient to activate significant recrystallization in Location II. As material approached the RXZ, it experiences additional plastic deformation at higher temperatures. In Location III which is approximately 200 m away from the RXZ, a high fraction of LAGB segments formed, and the LAGB segments were longer than that in Location II (Fig. 5e). Almost all the prior twins have been eliminated. A number of fine recrystallized grains developed along the boundaries of the deformed coarse grains. New twins are observed in the recrystallized region. Fig. 3f shows a high orientation spread within the deformed coarse grains. In contrast, the recrystallized grains are characterized by low intragranular orientation spread. Thus, the orientation spread map can be used as a rough reference to distinguish the deformed grains and new recrystallized grains. The bright colors in Fig. 3f indicated those locations exhibited high misorientation relative to the average of the grains. However, this does not necessarily mean that these locations experienced more plastic deformation or contained denser dislocations than other locations.
6
Close to the weld centerline, DRX began to control the grain structure evolution (Figs. 5g and h). This reduced the density of LAGBs and the grain size. More new twins developed in the recrystallized regions. In Location IV which is approximately 300 m away from the weld centerline, no grain larger than 5 m was observed and the majority of the grains have low intragranular orientation spread, demonstrating all the coarse parent grains have recrystallized. Fig. 4a shows that the average grain size in RXZ did not monotonically decrease as material approached the weld centerline, despite the increase in strain. The average grain size increased from ~2.0 m (Location IV) to ~2.3 m (Location V) firstly and then decreased to 1.9 m (Location VI) at the weld centerline. When the stain is extremely high and the parent grains have recrystallized, the average grain size of the material strongly affected by the strain-rate ( ̇ and temperature (T) [17, 18]. Similarly, as recrystallized stainless steel rotated around the probe during FSW, it maintained a fairly constant grain size [15] since the grain coarsening and recrystallization reached an equilibrium. Under this extreme deformation condition, the combined effect of ̇ and T are often described by the Zener-Hollomon parameter (Z): ̇
⁄
[19]
where R is the gas constant and Q is the relative activation energy for deformation. Grain refinement can be achieved through increasing the Z value (i.e. an increase in strain rate and/or a decrease in deformation temperature). This study will provide a deeper physical understanding of this phenomenon from the view of grain structure evolution. From Location IV to Location V, the average grain size increased slightly from ~2.0 m to ~2.3 m (Fig. 4a). The grains in Location V were larger and more equiaxed in shape compared to that in Location IV (Figs. 5 and 6). Furthermore, the maximum intragranular orientation 7
spread was reduced from 9.3 in Location IV to 6.9 in Location V, indicating intragranular deformation was lower in Location V. Fig. 6 shows that new recrystallized grains formed among the relatively large grains, indicating that both DRX and grain growth occurred in Location V. The increment of grain size from Location IV to Location V suggested that grain growth had an advantage over DRX in this region. From Location V to Location VI, the grain size decreased from ~2.3 m to ~1.9 m (Figs. 6 and 7), which is contrary to the trend of grain increment from Location IV to Location V. Fig. 7 shows that significant grain refinement occurred within a very narrow region (<80 m) across the weld centerline and the maximum intragranular orientation spread in Location VI increased to 8.3. This indicates the microstructural evolution in Location VI was governed by intragranular deformation and DRX. It is hard to identify the exact location of the weld interface according to the grain structure in Fig. 7 as the weld interface has been totally concealed by the recrystallized grains. Fig. 8 shows the peak temperature at different locations along the axial length of the weld sample. It shows that the deformation temperature increased as material approached the welded centerline, and the deformation temperature in the RXZ is higher than 1330K, which is 0.6 Tm of Alloy 718 (where Tm is the kelvin temperature of the alloy’s melting point). At temperatures close to or higher than 0.5 Tm, grain boundary sliding (GBS) or boundary motion among the equiaxed fine grains becomes active [20-22]. Intergranular deformation instead of intragranular deformation begins to increase with temperature elevation. In addition, GBS needs to be accommodated by other mechanisms, for example, grain boundary migration, diffusion flow or dislocation slip [21, 23-26], which is beneficial to grain growth. Thus, grain growth and reduced
8
intragranular deformation in Location V (Fig. 6) can be ascribed to the GBS at higher deformation temperature. For a specific alloy deformed at a particular temperature, the optimum strain rate for GBS (̇
) is a fixed value [27, 28]. When ̇ > ̇
, the rate of grain boundary migration, diffusion
flow or dislocation slip is insufficient to accommodate GBS. Stress concentration developed at certain sites, such as the grain triple junctions, boundary ledges, and even grain interiors [24], causing intragranular plastic deformation and recrystallization when the material is under compressive stress. Therefore, enhanced intragranular deformation and DRX in Location VI (Fig. 7) is associated the local stress concertation which is caused by accelerated strain rate. In the RXZ, the temperature, strain and strain-rate may increase as material approaches the weld centerline. From Location IV to Location V, the microstructure evolution is more affected by GBS and grain growth, indicating the temperature rise has more effect on grain structure evolution than the strain rate increase in this stage. However, from Location V to Location VI, the microstructure evolution is governed by intragranular deformation and DRX, indicating that the strain rate increase has more effect on grain structure evolution than the temperature rise. These can be further confirmed by the grain size distribution (Fig. 9). When grain structure evolution is controlled by DRX, fine recrystallized grains develop around the relative coarse deformed grains (Fig.5). Such a microstructure evolution will lead to a positively skewed distribution of grain size. Positively skewed distribution of grain size was detected in the TMAZ adjacent to the RXZ, Locations IV and VI (Figs. 9a,b, and d). This verified that these regions are more affected by DRX. On the contrary, GBS and normal grain growth will homogenize the size of recrystallized grains, resulting in a normal distribution of the grains. A nearly normal grain
9
size distribution appeared in Location V (Fig. 9c), demonstrating GBS and normal grain growth governed the grain structure evolution. Fig. 10 shows quantitative microstructure details along the weld centerline. The examination region was marked in Fig. 1b. From the center to the edge of the tube wall, the grain size increased slightly from ~1.8 m to ~2.1 m (Fig. 10a). The ratio of twin boundaries and LAGBs remained in the range of 3-4% and 9-10%, respectively (Figs. 10b and c). During FSW, the hot material at the center of the tube wall was extruded toward the edge of the tube wall and finally expelled away from the tube wall to form flash. The heat generated by plastic deformation moved toward the edge of the tube wall along with the extruded material. This increased the welding temperature at the tube wall edge, resulting in slightly increased grain size in that region. Fig. 11 shows the influence of rotation rate on microstructure distribution in RFW joints. With increasing weld rotation rate from 500 to 1000 rpm, the average grain size of the RXZ in the center and edge of the tube wall increased from ~1.9 and ~2.2 m to ~2.6 and ~3.0 m, respectively. The thickness of the RXZ in the center of the tube wall maintained ~400 m regardless the change in weld rotation rate. The thickness of the RXZ in the edge of the tube wall increased from ~800 m, to 1000 m after an increase in rotation rate from 500 to 1000 rpm. The change of rotation rate only led to a slight variation in the LAGBs and the twin boundaries in the RXZ. The present microstructural investigation shows that the major change caused by increasing tool rotation rate is the significant grain coarsening in the RXZ. The thin metal layer near the weld interface (the rubbing surface) is expected to flow to some extent along the tube rotation direction during RFW. From Fig. 2, the increase in power was roughly 60% as the rotation rate increased from 500 rpm to 1000 rpm. According to the definition of torque: torque = power /angular speed, an increase the rotation rate from 500 rpm to
10
1000 rpm should result in a reduction in torque, and therefore, a reduced friction force along the weld interface considering the tube dimension of the two welds are unchanged. Available data show that the flow stress of Alloy 718 increases significantly with an increase in strain rate at elevated temperatures [29, 30]. The reduced friction force at 1000 rpm indicated that the strain rate of the thin metal along the rotation direction did not increase after increasing tubing rotation rate. In addition, the two welds were produced at the same feed rate and the thickness of RXZ in the two welds was similar (Fig. 11). The increase in rotation rate from 500 rpm to 1000 rpm did not lead to a significant increase in strain in either the rotation direction or axial direction. The larger compression force at high rotation rate cannot be mainly ascribed to the strain rate of the alloy. Beside strain rate and temperature, grain size also has a significant effect on the flow stress of fine-grained metals (grain size ≤ 10 m). Grain refinement increases the density of grain boundaries, promoting GBS at elevated temperatures, which reduces the flow stress of the metal [21, 31, 32]. A previous study showed that the flow stress of a Al-Mg-Sc alloy with average grain size of 16 m deformed at 525oC and 1 × 10-1 s-1 is almost three times higher than that of another Al-Mg-Sc alloy with average grain size of 3.2 m deformed at 475oC and 1 × 10-1 s-1 [27]. It has been well documented that the flow stress of fine-grained alloys has a liner relationship with grain size at elevated temperatures when other deformation conditions are constant [27, 31, 33]. The RXZ of the two welds consisted of fine equiaxed grains, which allow GBS at elevated temperatures. The average grain size in RXZ of the weld increased by at least ~35% after the tool rotation rate increased from 500 rpm to 1000 rpm. Considering the finer grains have a faster grain coarsening rate during the process of sample quenching, the difference of grain size during
11
friction welding should be larger than that observed in the as-welded samples. Thus, the lower axis welding force in sample welded at 500 rpm can be attributed to the finer grain size in the RXZ which required lower applied force to enable GBS.
4. Conclusions 1. The weld made at 500 rpm exhibited lower axial force during welding than the weld at 1000 rpm. The lower axis welding force can be attributed to the finer grain size in the RXZ which requires lower applied force to enable GBS. 2. The average grain size in the RXZ was not decreased monotonically as material approached weld centerline. This is because the grain structure in the RXZ was a result of the competition between DRX and GBS. The GBS caused grain growth at elevated temperatures while the DRX refined the grains. 3. In the TMAZ near the RXZ, the deformed coarse grains exhibited high orientation spread. DRX developed along the boundaries of deformed coarse grains and began to control grain structure evolution. Almost all the prior twin boundaries have been destroyed and new twins generated in the recrystallization regions. 4. From the center to the edge of the tube wall, both the thickness of RXZ and the average grain size in RXZ increased. This is because the heat generated by plastic deformation at the center of the tube wall is moved toward the edge of the tube wall along with the extruded material.
Acknowledgement: Funding for this work was provided by General Electric Global Research, Niskayuna, NY.
12
References [1]
[2] [3]
[4] [5] [6]
[7] [8] [9] [10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
N. J. Harrison, I. Todd, and K. Mumtaz, "Reduction of micro-cracking in nickel superalloys processed by Selective Laser Melting: A fundamental alloy design approach," Acta Materialia, vol. 94, pp. 59-68, 2015. Huang Xiao, C. M.C., R. N.L., and J. J., "<1997 The effect of grain boundary segregation of boron in cast alloy 718 on HAZ microfissuring.pdf>," Acta Mater, vol. 45, 1997 1997. F. C. Liu and Z. Y. Ma, "Influence of Tool Dimension and Welding Parameters on Microstructure and Mechanical Properties of Friction-Stir-Welded 6061-T651 Aluminum Alloy," Metallurgical and Materials Transactions A, vol. 39, pp. 2378-2388, 2008. R. Damodaram, S. G. S. Raman, and K. P. Rao, "Microstructure and mechanical properties of friction welded alloy 718," Materials Science and Engineering: A, vol. 560, pp. 781-786, 2013. C. Mary and M. Jahazi, "Multi-Scale Analysis of IN-718 Microstructure Evolution During Linear Friction Welding," Advanced Engineering Materials, vol. 10, pp. 573-578, 2008. R. Damodaram, S. Ganesh Sundara Raman, and K. Prasad Rao, "Effect of post-weld heat treatments on microstructure and mechanical properties of friction welded alloy 718 joints," Materials & Design, vol. 53, pp. 954-961, 2014. K. R. Vishwakarma, O. A. Ojo, P. Wanjara, and M. C. Chaturvedi, "Microstructural Analysis of Linear Friction-Welded 718 Plus Superalloy," Jom, vol. 66, pp. 2525-2534, 2014. W. Li, A. Vairis, M. Preuss, and T. Ma, "Linear and rotary friction welding review," International Materials Reviews, vol. 61, pp. 71-100, 2016. A. Chamanfar, M. Jahazi, and J. Cormier, "A Review on Inertia and Linear Friction Welding of NiBased Superalloys," Metallurgical and Materials Transactions A, vol. 46, pp. 1639-1669, 2015. N. Iqbal, J. Rolph, R. Moat, D. Hughes, M. Hofmann, J. Kelleher, et al., "A Comparison of Residual Stress Development in Inertia Friction Welded Fine Grain and Coarse Grain Nickel-Base Superalloy," Metallurgical and Materials Transactions a-Physical Metallurgy and Materials Science, vol. 42A, pp. 4056-4063, Dec 2011. Z. W. Huang, H. Y. Li, G. Baxter, S. Bray, and P. Bowen, "Electron microscopy characterization of the weld line zones of an inertia friction welded superalloy," Journal of Materials Processing Technology, vol. 211, pp. 1927-1936, Dec 2011. A. Chamanfar, M. Jahazi, J. Gholipour, P. Wanjara, and S. Yue, "Mechanical Property and Microstructure of Linear Friction Welded WASPALOY," Metallurgical and Materials Transactions A, vol. 42, pp. 729-744, 2010. W. Chen, M. C. Chaturvedi, and N. L. Richards, "Effect of boron segregation at grain boundaries on heat-affected zone cracking in wrought INCONEL 718," Metallurgical and Materials Transactions a-Physical Metallurgy and Materials Science, vol. 32, pp. 931-939, Apr 2001. S. L. West, W. A. Baeslack, and T. J. Kelly, "MORPHOLOGY OF WELD HEAT-AFFECTED ZONE LIQUATION CRACKING IN TA-MODIFIED CAST ALLOY 718," Metallography, vol. 23, pp. 219-229, Nov 1989. F. C. Liu and T. W. Nelson, "In-situ grain structure and texture evolution during friction stir welding of austenite stainless steel," Materials & Design, vol. 115, pp. 467-478, 2017. Y. V. Nesterova and V. V. Rybin, "MECHANICAL TWINNING AND FRAGMENTATION OF TECHNICALLY PURE TITANIUM ON DEVELOPED PLASTICA DEFORMATION STAGE," Fizika Metallov I Metallovedenie, vol. 59, pp. 395-406, 1985. Y. S. Li, Y. Zhang, N. R. Tao, and K. Lu, "Effect of the Zener-Hollomon parameter on the microstructures and mechanical properties of Cu subjected to plastic deformation," Acta Materialia, vol. 57, pp. 761-772, Feb 2009.
13
[18]
[19] [20] [21]
[22] [23] [24]
[25]
[26] [27] [28]
[29]
[30]
[31]
[32]
[33]
C. I. Chang, C. J. Lee, and J. C. Huang, "Relationship between grain size and Zener-Holloman parameter during friction stir processing in AZ31 Mg alloys," Scripta Materialia, vol. 51, pp. 509514, Sep 2004. C. Zener and J. H. Hollomon, "Effect of strain rate upon plastic flow of steel," Journal of Applied Physics, vol. 15, 1944. F. C. Liu and Z. Y. Ma, "Contribution of grain boundary sliding in low-temperature superplasticity of ultrafine-grained aluminum alloys," Scripta Materialia, vol. 62, pp. 125-128, 2010. Z. Y. Ma, F. C. Liu, and R. S. Mishra, "Superplastic deformation mechanism of an ultrafinegrained aluminum alloy produced by friction stir processing," Acta Materialia, vol. 58, pp. 46934704, 2010. F. C. Liu and Z. Y. Ma, "Low-temperature superplasticity of friction stir processed Al–Zn–Mg–Cu alloy," Scripta Materialia, vol. 58, pp. 667-670, 2008. P. Wang, Y. Feng, F. Liu, L. Wu, and S. Guan, "Microstructure and mechanical properties of Ti-ZrCr biomedical alloys," Mater Sci Eng C Mater Biol Appl, vol. 51, pp. 148-52, Jun 2015. F. C. Liu, P. Xue, and Z. Y. Ma, "Microstructural evolution in recrystallized and unrecrystallized Al–Mg–Sc alloys during superplastic deformation," Materials Science and Engineering: A, vol. 547, pp. 55-63, 2012. R. C. Gifkins, "GRAIN-BOUNDARY SLIDING AND ITS ACCOMMODATION DURING CREEP AND SUPERPLASTICITY," Metallurgical Transactions a-Physical Metallurgy and Materials Science, vol. 7, pp. 1225-1232, 1976. M. F. Ashby and R. A. Verrall, "DIFFUSION-ACCOMMODATED FLOW AND SUPERPLASTICITY," Acta Metallurgica, vol. 21, pp. 149-163, 1973. F. C. Liu and Z. Y. Ma, "Superplasticity governed by effective grain size and its distribution in finegrained aluminum alloys," Materials Science and Engineering: A, vol. 530, pp. 548-558, 2011. K. Wang, F. C. Liu, P. Xue, B. L. Xiao, and Z. Y. Ma, "Effects of heating rates on microstructure and superplastic behavior of friction stir processed 7075 aluminum alloy," Journal of Materials Science, vol. 50, pp. 1006-1015, 2014. M. Azarbarmas, M. Aghaie-Khafri, J. M. Cabrera, and J. Calvo, "Microstructural evolution and constitutive equations of Inconel 718 alloy under quasi-static and quasi-dynamic conditions," Materials & Design, vol. 94, pp. 28-38, Mar 2016. Y. C. Lin, K. K. Li, H. B. Li, J. Chen, X. M. Chen, and D. X. Wen, "New constitutive model for hightemperature deformation behavior of inconel 718 superalloy," Materials & Design, vol. 74, pp. 108-118, Jun 2015. K. Wang, F. C. Liu, P. Xue, D. Wang, B. L. Xiao, and Z. Y. Ma, "Superplastic Constitutive Equation Including Percentage of High-Angle Grain Boundaries as a Microstructural Parameter," Metallurgical and Materials Transactions A, vol. 47, pp. 546-559, 2015. F. C. Liu and Z. Y. Ma, "Achieving exceptionally high superplasticity at high strain rates in a micrograined Al–Mg–Sc alloy produced by friction stir processing," Scripta Materialia, vol. 59, pp. 882-885, 2008. T. G. Langdon, "Grain boundary sliding revisited: Developments in sliding over four decades," Journal of Materials Science, vol. 41, pp. 597-609, Feb 2006.
14
Fig. 1. EBSD examination regions: (a) examination cross-section relative to RFW joint; (b) EBSD scanning regions relative to observing cross-section.
15
Fig. 2. Variation of welding power with time for the joints processed at different rotation rate.
16
Fig. 3. Variation of welding power with time for the joints processed at different rotation rate.
17
Fig. 4. Microstructure distribution across weld centerline of weld I (500 rpm): (a) grain size, (b) ratio of LAGBs and (c) percentage of twin boundaries along edge of tube wall; (d) grain size, (e) ratio of LAGBs and (f) percentage of twin boundaries along center of tube wall.
18
Fig. 5. EBSD maps of Locations I-V. Grain orientation maps of (a) Location I, (c) Location II, (e) Location III, (g) Location IV. Orientation spread maps of (b) Location I, (d) Location II, (f) Location III, (h) Location IV.
19
Fig. 6. Grain orientation map (a) and orientation spread map (b) of Location V.
20
Fig. 7. Grain orientation map (a) and orientation spread map (b) of Location VI.
21
Fig. 8. Peak temperatures at on the tube wall at different distances from the end of tube.
22
Fig. 9. Grain size distribution of (a) TMAZ adjacent to RXZ, (b) Location IV, (c) Location V, and (d) Location VI (Curve fitting were presented as red lines).
23
Fig. 10. Microstructure distribution along weld centerline of weld I: (a) grain size, (b) ratio of LAGBs and (c) percentage of twin boundaries.
24
Fig. 11. Microstructure distribution across the weld centerline as a function of weld rotation rate: (a) grain size, (b) ratio of LAGBs and (c) percentage of twin boundaries along edge of tube wall; (d) grain size, (e) ratio of LAGBs and (f) percentage of twin boundaries along center of tube wall.
25
PII: DOI: Reference:
S0921-5093(17)31420-X https://doi.org/10.1016/j.msea.2017.10.092 MSA35695
To appear in: Materials Science & Engineering A Received date: 11 July 2017 Revised date: 24 October 2017 Accepted date: 26 October 2017 Cite this article as: F.C. Liu and T.W. Nelson, Grain structure evolution, grain boundary sliding and material flow resistance in friction welding of Alloy 718, Materials Science & Engineering A, https://doi.org/10.1016/j.msea.2017.10.092 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Grain structure evolution, grain boundary sliding and material flow resistance in friction welding of Alloy 718
F.C. Liu and T.W. Nelson
Department of Mechanical Engineering, Brigham Young University, 435 CTB, Provo, UT 84602, USA
Alloy 718 tubes were subjected to rotary friction welding to understand to the process fundamental and grain structure evolution during welding. The distribution of grain size, lowangle grain boundaries (LAGBs), and twin boundaries throughout the joints were quantitatively analyzed. The weld power, axial load, and weld temperature were monitored. The grain structure evolution during friction welding was clarified. The grain structure in the recrystallization zone (RXZ) of the weld was a result of competition between dynamic recrystallization and grain boundary sliding (GBS), which is controlled by the local deformation condition. The axial force during welding decreased with reducing the rotation rate from 1000 rpm to 500 rpm. This anomalistic phenomenon can be ascribed that a decrease in rotation rate resulted in finer grain size in the RXZ of the weld, which required lower applied force to enable GBS. Keywords: Superalloy; Recrystallization; EBSD; Friction welding; Power; Axial force 1. Introduction Alloy 718 is a nickel-based austenitic phase alloy. It can be used to make load-bearing structures at the very demanding conditions as this alloy features high strength at elevated temperatures as well as excellent corrosion, oxidation and creep resistance. Increased amount of these materials are being applied in aerospace, power station and chemical industries. Fusion welding of Alloy 718 may cause problems such as weld liquation related cracking and segregation of alloying elements [1, 2]. These problems can be avoid using solid state welding
F.C. Liu, E-mail: [email protected], Tel: +1 801-422-5189, Fax: +1 801-422-0516. T.W. Nelson, E-mail: [email protected], Tel: +1 801-422-6233, Fax: +1 801-422-0516.
1
techniques, such as friction welding (FW) or friction stir welding (FSW) to join these highperformance metals [3-7]. Of the available solid state welding methods, rotary FW (RFW) is usually the best choice for joining cylinder or tubular cross-sectioned components to other workpieces. During our RFW, one tube was rotated at a constant speed and then moved toward another part. After the two components had contacted, they moved toward each other at a constant feeding speed, resulting in a consolidate weld zone after the rotation was terminated. The RFW offer several significant advantages over conventional fusion welding: (1) No melting occurred; (2) The weld zone consisted of fine recrystallized grains rather than a coarse solidification microstructure; (3) The weld operation and monitoring can be performed automatically; (4) RFW is highly efficient and reduce the welding times. (5) RFW is environmentally friendly. Early literature has reported the RFW in the aspects of welding parameter, welding temperature, mechanical properties and as-welded microstructure [4, 6, 8-11]. It is generally established that the material in the weld zone experienced heavy plastic deformation under elevated temperature during RFW, forming a weld zone consisting of fine recrystallized grains as a result of dynamic recrystallization (DRX) [6, 8, 9, 12]. Fine grains developed among deformed grains in the thermo-mechanical affected zone (TMAZ) [6, 12]. During RFW, the strain, strain rate, and temperature vary at different locations across the weld. This made the microstructure evaluation during RFW complicated. The microstructure distribution across the RFW joints has never been quantitatively analyzed in a systematic approach. We are far from a fundamental understanding of the grain structure evolution during RFW and how the microstructure will affect the materials load-bearing capacity.
2
Alloy 718 was chosen as the material to be investigated. It is a γ″ (Ni3Nb) strengthened nickel-based alloy that combines excellent corrosion resistance and high strength at elevated temperatures. This alloy has been widely used in gas turbines, aircraft engines, extrusion dies and chemical processing containers. Fusion welding of Alloy 718 usually associates with some problems such as boron/niobium segregation, and liquation cracking [13, 14], which can be avoided in RFW. In this study, the distribution of grain size, low-angle grain boundaries (LAGBs), and twin boundaries throughout the joints were quantitatively analyzed. The weld power, axial load, and weld temperature were monitored. This provides a comprehensive understanding of grain structure evolution during friction welding and how the microstructure affects the materials loadbearing capacity during welding. 2. Experimental details The outer diameter and inner diameter of the Alloy 718 tube is 25.4 and 20.3 mm, respectively. The chemical composition (wt%) is 51.6 Ni, 18.2 Cr, 5.1 Nb, 3.28 Mo, 1.06 Ti, 0.56 Al, 0.33 V, 0.09 Mn and Fe balance. The welding was performed using a multifunction TTI-RM2 friction stir welding machine. The feed rate and feed distance for both processes are 2 mm/s and 3.6 mm, respectively. The rotation rate of weld I and II are 500 and 1000 rpm, respectively. Thermocouples were welded on the tube wall at distances of 1, 2, 3, and 4 mm away from the end of the tube to record the temperature change during welding. The weld region was removed using electrical discharge machining for optical and EBSD examination. The observation plan and EBSD scanning regions were schematically shown in Fig. 1. The EBSD sample was mechanical polished to 1 m diamond past and then followed by a vibratory polishing using 0.05 m colloidal silica. High-resolution EBSD data were obtained at
3
20 kV using FEI Helios Nanolab 600 with a TSL channel EBSD system. The average confidential index for each EBSD map is higher than 0.5. In order to eliminate spurious boundaries caused by orientation noise, boundaries with misorientation less than 2o were not considered. For twin boundary definition, a tolerance of 5 deg was allowed in both the rotation of twin plane normal and the angular between the twinning planes on either side of the boundary. 3. Results and analysis Fig. 2 shows that an increase in weld rotation rate from 500 to 1000 rpm resulted in a significant increase in weld power. As the two welds used the same amount of time, the higher weld power provides a higher energy input to the weld. Fig. 3 shows that the compression force along the axis direction increased with an increase in tool rotation rate. This phenomenon is interesting as higher energy input (higher tool rotation rate) did not lead to less deformation force under the present deformation conditions. In order to understand this anomalistic phenomenon, the microstructure throughout the joints was quantitatively analyzed. Quantitative microstructure details across the weld zone along the edge (Figs. 4a, b, and c) and center (Figs. 4d, e, and f) of the tube wall of weld I are presented. In general, the microstructure distribution is symmetric about the weld centerline (Fig. 4). The region near the weld centerline was characterized by refined grains (< 2.5 m) and low fraction of LABGs (boundary misorientation < 15%), demonstrating recrystallization has occurred (Figs. 4a and d). This region was therefore defined as recrystallization zone (RXZ). The thickness of RXZ along the edge and center of the tube wall is approximately 0.8 and 0.5 mm, respectively. The average grain size of RXZ at the edge and center of the tube wall is 2.2 and 1.9 m, respectively. The RXZ along the tube wall edge was thicker and consisted of larger grains than that along the tube wall center, indicating high energy absorption in the tube wall edge.
4
Outside of the RXZ is the TMAZ. The TMAZ contained a higher percentage of LAGBs than the base metal (BM) and the RXZ (Figs. 4b and e). Within the TMAZ, the grain size decreased linearly from the BM to the RXZ (Figs. 4a and d). The TMAZ can be divided into two distinct parts according to the distribution of LAGBs and twin boundaries (Figs. 4b, c, e, and f). In the TMAZ close to the BM, the ratio of LAGBs increased and the percentage of twin boundaries decreased. Such a microstructure change is mainly caused by plastic deformation. In the TMAZ near the RXZ, recrystallization occurred as the metal approached the RXZ, which decreased the ratio of LAGBs and increased the percentage of twin boundaries. As an annealed Alloy 718 was used, the heat affected zone (HAZ) maintained similar grain structure as the base metal, indicating the peak temperature in HAZ is lower than the recrystallization temperature of the base metal. Detailed microstructure analysis was performed on Locations I-VI (Fig. 4a) to understand the microstructure evolution during welding. Fig. 5 shows EBSD maps from Location I to IV (Indicated in Fig. 4). The points in the grain orientation maps (the left maps) were colored according to their crystallography orientation relative to the normal of observation plan. In the orientation spread maps (the right maps), each point was shaded according to the misorientation it made relative to the average orientation of the grain. LAGBs, high-angle grain boundaries (HAGBs > 15o), and twin boundaries were displayed as white, black and red lines, respectively. All the EBSD maps in this study were colored according to the same principle. Most of the grains in the BM were free of LAGBs and have low orientation spread. Annealing twins were extensively within these grains. (Fig. 5a and b). Fig. 5c and d show the grain structure of Location II (Fig. 4a), which is within the TMAZ but closer to the BM. New LAGB segments developed within the grain interiors. The LAGBs
5
segments were short, curved, disconnected and have not developed into continuous grain boundaries. A few fine recrystallized grains (which marked by arrows in Fig. 5c) were observed along the prior grain boundaries. The plastic deformation caused some of the prior twin boundaries to deviate away from the exact twin relationship with the matrix. Similar transform of boundary misorientations has been observed in friction stir welding of annealed stainless steel [15] and explained as strain induced crystallographic rotation of the twin and matrix from their original orientations [16]. As no new twins were generated in this location (Fig. 5c), the disruption of original twin boundaries led to a decrease in the ratio of twin boundaries (Fig. 4c). Enhanced orientation spread was observed in most of the deformed grains (Fig. 5d), but the stored deformation energy was insufficient to activate significant recrystallization in Location II. As material approached the RXZ, it experiences additional plastic deformation at higher temperatures. In Location III which is approximately 200 m away from the RXZ, a high fraction of LAGB segments formed, and the LAGB segments were longer than that in Location II (Fig. 5e). Almost all the prior twins have been eliminated. A number of fine recrystallized grains developed along the boundaries of the deformed coarse grains. New twins are observed in the recrystallized region. Fig. 3f shows a high orientation spread within the deformed coarse grains. In contrast, the recrystallized grains are characterized by low intragranular orientation spread. Thus, the orientation spread map can be used as a rough reference to distinguish the deformed grains and new recrystallized grains. The bright colors in Fig. 3f indicated those locations exhibited high misorientation relative to the average of the grains. However, this does not necessarily mean that these locations experienced more plastic deformation or contained denser dislocations than other locations.
6
Close to the weld centerline, DRX began to control the grain structure evolution (Figs. 5g and h). This reduced the density of LAGBs and the grain size. More new twins developed in the recrystallized regions. In Location IV which is approximately 300 m away from the weld centerline, no grain larger than 5 m was observed and the majority of the grains have low intragranular orientation spread, demonstrating all the coarse parent grains have recrystallized. Fig. 4a shows that the average grain size in RXZ did not monotonically decrease as material approached the weld centerline, despite the increase in strain. The average grain size increased from ~2.0 m (Location IV) to ~2.3 m (Location V) firstly and then decreased to 1.9 m (Location VI) at the weld centerline. When the stain is extremely high and the parent grains have recrystallized, the average grain size of the material strongly affected by the strain-rate ( ̇ and temperature (T) [17, 18]. Similarly, as recrystallized stainless steel rotated around the probe during FSW, it maintained a fairly constant grain size [15] since the grain coarsening and recrystallization reached an equilibrium. Under this extreme deformation condition, the combined effect of ̇ and T are often described by the Zener-Hollomon parameter (Z): ̇
⁄
[19]
where R is the gas constant and Q is the relative activation energy for deformation. Grain refinement can be achieved through increasing the Z value (i.e. an increase in strain rate and/or a decrease in deformation temperature). This study will provide a deeper physical understanding of this phenomenon from the view of grain structure evolution. From Location IV to Location V, the average grain size increased slightly from ~2.0 m to ~2.3 m (Fig. 4a). The grains in Location V were larger and more equiaxed in shape compared to that in Location IV (Figs. 5 and 6). Furthermore, the maximum intragranular orientation 7
spread was reduced from 9.3 in Location IV to 6.9 in Location V, indicating intragranular deformation was lower in Location V. Fig. 6 shows that new recrystallized grains formed among the relatively large grains, indicating that both DRX and grain growth occurred in Location V. The increment of grain size from Location IV to Location V suggested that grain growth had an advantage over DRX in this region. From Location V to Location VI, the grain size decreased from ~2.3 m to ~1.9 m (Figs. 6 and 7), which is contrary to the trend of grain increment from Location IV to Location V. Fig. 7 shows that significant grain refinement occurred within a very narrow region (<80 m) across the weld centerline and the maximum intragranular orientation spread in Location VI increased to 8.3. This indicates the microstructural evolution in Location VI was governed by intragranular deformation and DRX. It is hard to identify the exact location of the weld interface according to the grain structure in Fig. 7 as the weld interface has been totally concealed by the recrystallized grains. Fig. 8 shows the peak temperature at different locations along the axial length of the weld sample. It shows that the deformation temperature increased as material approached the welded centerline, and the deformation temperature in the RXZ is higher than 1330K, which is 0.6 Tm of Alloy 718 (where Tm is the kelvin temperature of the alloy’s melting point). At temperatures close to or higher than 0.5 Tm, grain boundary sliding (GBS) or boundary motion among the equiaxed fine grains becomes active [20-22]. Intergranular deformation instead of intragranular deformation begins to increase with temperature elevation. In addition, GBS needs to be accommodated by other mechanisms, for example, grain boundary migration, diffusion flow or dislocation slip [21, 23-26], which is beneficial to grain growth. Thus, grain growth and reduced
8
intragranular deformation in Location V (Fig. 6) can be ascribed to the GBS at higher deformation temperature. For a specific alloy deformed at a particular temperature, the optimum strain rate for GBS (̇
) is a fixed value [27, 28]. When ̇ > ̇
, the rate of grain boundary migration, diffusion
flow or dislocation slip is insufficient to accommodate GBS. Stress concentration developed at certain sites, such as the grain triple junctions, boundary ledges, and even grain interiors [24], causing intragranular plastic deformation and recrystallization when the material is under compressive stress. Therefore, enhanced intragranular deformation and DRX in Location VI (Fig. 7) is associated the local stress concertation which is caused by accelerated strain rate. In the RXZ, the temperature, strain and strain-rate may increase as material approaches the weld centerline. From Location IV to Location V, the microstructure evolution is more affected by GBS and grain growth, indicating the temperature rise has more effect on grain structure evolution than the strain rate increase in this stage. However, from Location V to Location VI, the microstructure evolution is governed by intragranular deformation and DRX, indicating that the strain rate increase has more effect on grain structure evolution than the temperature rise. These can be further confirmed by the grain size distribution (Fig. 9). When grain structure evolution is controlled by DRX, fine recrystallized grains develop around the relative coarse deformed grains (Fig.5). Such a microstructure evolution will lead to a positively skewed distribution of grain size. Positively skewed distribution of grain size was detected in the TMAZ adjacent to the RXZ, Locations IV and VI (Figs. 9a,b, and d). This verified that these regions are more affected by DRX. On the contrary, GBS and normal grain growth will homogenize the size of recrystallized grains, resulting in a normal distribution of the grains. A nearly normal grain
9
size distribution appeared in Location V (Fig. 9c), demonstrating GBS and normal grain growth governed the grain structure evolution. Fig. 10 shows quantitative microstructure details along the weld centerline. The examination region was marked in Fig. 1b. From the center to the edge of the tube wall, the grain size increased slightly from ~1.8 m to ~2.1 m (Fig. 10a). The ratio of twin boundaries and LAGBs remained in the range of 3-4% and 9-10%, respectively (Figs. 10b and c). During FSW, the hot material at the center of the tube wall was extruded toward the edge of the tube wall and finally expelled away from the tube wall to form flash. The heat generated by plastic deformation moved toward the edge of the tube wall along with the extruded material. This increased the welding temperature at the tube wall edge, resulting in slightly increased grain size in that region. Fig. 11 shows the influence of rotation rate on microstructure distribution in RFW joints. With increasing weld rotation rate from 500 to 1000 rpm, the average grain size of the RXZ in the center and edge of the tube wall increased from ~1.9 and ~2.2 m to ~2.6 and ~3.0 m, respectively. The thickness of the RXZ in the center of the tube wall maintained ~400 m regardless the change in weld rotation rate. The thickness of the RXZ in the edge of the tube wall increased from ~800 m, to 1000 m after an increase in rotation rate from 500 to 1000 rpm. The change of rotation rate only led to a slight variation in the LAGBs and the twin boundaries in the RXZ. The present microstructural investigation shows that the major change caused by increasing tool rotation rate is the significant grain coarsening in the RXZ. The thin metal layer near the weld interface (the rubbing surface) is expected to flow to some extent along the tube rotation direction during RFW. From Fig. 2, the increase in power was roughly 60% as the rotation rate increased from 500 rpm to 1000 rpm. According to the definition of torque: torque = power /angular speed, an increase the rotation rate from 500 rpm to
10
1000 rpm should result in a reduction in torque, and therefore, a reduced friction force along the weld interface considering the tube dimension of the two welds are unchanged. Available data show that the flow stress of Alloy 718 increases significantly with an increase in strain rate at elevated temperatures [29, 30]. The reduced friction force at 1000 rpm indicated that the strain rate of the thin metal along the rotation direction did not increase after increasing tubing rotation rate. In addition, the two welds were produced at the same feed rate and the thickness of RXZ in the two welds was similar (Fig. 11). The increase in rotation rate from 500 rpm to 1000 rpm did not lead to a significant increase in strain in either the rotation direction or axial direction. The larger compression force at high rotation rate cannot be mainly ascribed to the strain rate of the alloy. Beside strain rate and temperature, grain size also has a significant effect on the flow stress of fine-grained metals (grain size ≤ 10 m). Grain refinement increases the density of grain boundaries, promoting GBS at elevated temperatures, which reduces the flow stress of the metal [21, 31, 32]. A previous study showed that the flow stress of a Al-Mg-Sc alloy with average grain size of 16 m deformed at 525oC and 1 × 10-1 s-1 is almost three times higher than that of another Al-Mg-Sc alloy with average grain size of 3.2 m deformed at 475oC and 1 × 10-1 s-1 [27]. It has been well documented that the flow stress of fine-grained alloys has a liner relationship with grain size at elevated temperatures when other deformation conditions are constant [27, 31, 33]. The RXZ of the two welds consisted of fine equiaxed grains, which allow GBS at elevated temperatures. The average grain size in RXZ of the weld increased by at least ~35% after the tool rotation rate increased from 500 rpm to 1000 rpm. Considering the finer grains have a faster grain coarsening rate during the process of sample quenching, the difference of grain size during
11
friction welding should be larger than that observed in the as-welded samples. Thus, the lower axis welding force in sample welded at 500 rpm can be attributed to the finer grain size in the RXZ which required lower applied force to enable GBS.
4. Conclusions 1. The weld made at 500 rpm exhibited lower axial force during welding than the weld at 1000 rpm. The lower axis welding force can be attributed to the finer grain size in the RXZ which requires lower applied force to enable GBS. 2. The average grain size in the RXZ was not decreased monotonically as material approached weld centerline. This is because the grain structure in the RXZ was a result of the competition between DRX and GBS. The GBS caused grain growth at elevated temperatures while the DRX refined the grains. 3. In the TMAZ near the RXZ, the deformed coarse grains exhibited high orientation spread. DRX developed along the boundaries of deformed coarse grains and began to control grain structure evolution. Almost all the prior twin boundaries have been destroyed and new twins generated in the recrystallization regions. 4. From the center to the edge of the tube wall, both the thickness of RXZ and the average grain size in RXZ increased. This is because the heat generated by plastic deformation at the center of the tube wall is moved toward the edge of the tube wall along with the extruded material.
Acknowledgement: Funding for this work was provided by General Electric Global Research, Niskayuna, NY.
12
References [1]
[2] [3]
[4] [5] [6]
[7] [8] [9] [10]
[11]
[12]
[13]
[14]
[15] [16]
[17]
N. J. Harrison, I. Todd, and K. Mumtaz, "Reduction of micro-cracking in nickel superalloys processed by Selective Laser Melting: A fundamental alloy design approach," Acta Materialia, vol. 94, pp. 59-68, 2015. Huang Xiao, C. M.C., R. N.L., and J. J., "<1997 The effect of grain boundary segregation of boron in cast alloy 718 on HAZ microfissuring.pdf>," Acta Mater, vol. 45, 1997 1997. F. C. Liu and Z. Y. Ma, "Influence of Tool Dimension and Welding Parameters on Microstructure and Mechanical Properties of Friction-Stir-Welded 6061-T651 Aluminum Alloy," Metallurgical and Materials Transactions A, vol. 39, pp. 2378-2388, 2008. R. Damodaram, S. G. S. Raman, and K. P. Rao, "Microstructure and mechanical properties of friction welded alloy 718," Materials Science and Engineering: A, vol. 560, pp. 781-786, 2013. C. Mary and M. Jahazi, "Multi-Scale Analysis of IN-718 Microstructure Evolution During Linear Friction Welding," Advanced Engineering Materials, vol. 10, pp. 573-578, 2008. R. Damodaram, S. Ganesh Sundara Raman, and K. Prasad Rao, "Effect of post-weld heat treatments on microstructure and mechanical properties of friction welded alloy 718 joints," Materials & Design, vol. 53, pp. 954-961, 2014. K. R. Vishwakarma, O. A. Ojo, P. Wanjara, and M. C. Chaturvedi, "Microstructural Analysis of Linear Friction-Welded 718 Plus Superalloy," Jom, vol. 66, pp. 2525-2534, 2014. W. Li, A. Vairis, M. Preuss, and T. Ma, "Linear and rotary friction welding review," International Materials Reviews, vol. 61, pp. 71-100, 2016. A. Chamanfar, M. Jahazi, and J. Cormier, "A Review on Inertia and Linear Friction Welding of NiBased Superalloys," Metallurgical and Materials Transactions A, vol. 46, pp. 1639-1669, 2015. N. Iqbal, J. Rolph, R. Moat, D. Hughes, M. Hofmann, J. Kelleher, et al., "A Comparison of Residual Stress Development in Inertia Friction Welded Fine Grain and Coarse Grain Nickel-Base Superalloy," Metallurgical and Materials Transactions a-Physical Metallurgy and Materials Science, vol. 42A, pp. 4056-4063, Dec 2011. Z. W. Huang, H. Y. Li, G. Baxter, S. Bray, and P. Bowen, "Electron microscopy characterization of the weld line zones of an inertia friction welded superalloy," Journal of Materials Processing Technology, vol. 211, pp. 1927-1936, Dec 2011. A. Chamanfar, M. Jahazi, J. Gholipour, P. Wanjara, and S. Yue, "Mechanical Property and Microstructure of Linear Friction Welded WASPALOY," Metallurgical and Materials Transactions A, vol. 42, pp. 729-744, 2010. W. Chen, M. C. Chaturvedi, and N. L. Richards, "Effect of boron segregation at grain boundaries on heat-affected zone cracking in wrought INCONEL 718," Metallurgical and Materials Transactions a-Physical Metallurgy and Materials Science, vol. 32, pp. 931-939, Apr 2001. S. L. West, W. A. Baeslack, and T. J. Kelly, "MORPHOLOGY OF WELD HEAT-AFFECTED ZONE LIQUATION CRACKING IN TA-MODIFIED CAST ALLOY 718," Metallography, vol. 23, pp. 219-229, Nov 1989. F. C. Liu and T. W. Nelson, "In-situ grain structure and texture evolution during friction stir welding of austenite stainless steel," Materials & Design, vol. 115, pp. 467-478, 2017. Y. V. Nesterova and V. V. Rybin, "MECHANICAL TWINNING AND FRAGMENTATION OF TECHNICALLY PURE TITANIUM ON DEVELOPED PLASTICA DEFORMATION STAGE," Fizika Metallov I Metallovedenie, vol. 59, pp. 395-406, 1985. Y. S. Li, Y. Zhang, N. R. Tao, and K. Lu, "Effect of the Zener-Hollomon parameter on the microstructures and mechanical properties of Cu subjected to plastic deformation," Acta Materialia, vol. 57, pp. 761-772, Feb 2009.
13
[18]
[19] [20] [21]
[22] [23] [24]
[25]
[26] [27] [28]
[29]
[30]
[31]
[32]
[33]
C. I. Chang, C. J. Lee, and J. C. Huang, "Relationship between grain size and Zener-Holloman parameter during friction stir processing in AZ31 Mg alloys," Scripta Materialia, vol. 51, pp. 509514, Sep 2004. C. Zener and J. H. Hollomon, "Effect of strain rate upon plastic flow of steel," Journal of Applied Physics, vol. 15, 1944. F. C. Liu and Z. Y. Ma, "Contribution of grain boundary sliding in low-temperature superplasticity of ultrafine-grained aluminum alloys," Scripta Materialia, vol. 62, pp. 125-128, 2010. Z. Y. Ma, F. C. Liu, and R. S. Mishra, "Superplastic deformation mechanism of an ultrafinegrained aluminum alloy produced by friction stir processing," Acta Materialia, vol. 58, pp. 46934704, 2010. F. C. Liu and Z. Y. Ma, "Low-temperature superplasticity of friction stir processed Al–Zn–Mg–Cu alloy," Scripta Materialia, vol. 58, pp. 667-670, 2008. P. Wang, Y. Feng, F. Liu, L. Wu, and S. Guan, "Microstructure and mechanical properties of Ti-ZrCr biomedical alloys," Mater Sci Eng C Mater Biol Appl, vol. 51, pp. 148-52, Jun 2015. F. C. Liu, P. Xue, and Z. Y. Ma, "Microstructural evolution in recrystallized and unrecrystallized Al–Mg–Sc alloys during superplastic deformation," Materials Science and Engineering: A, vol. 547, pp. 55-63, 2012. R. C. Gifkins, "GRAIN-BOUNDARY SLIDING AND ITS ACCOMMODATION DURING CREEP AND SUPERPLASTICITY," Metallurgical Transactions a-Physical Metallurgy and Materials Science, vol. 7, pp. 1225-1232, 1976. M. F. Ashby and R. A. Verrall, "DIFFUSION-ACCOMMODATED FLOW AND SUPERPLASTICITY," Acta Metallurgica, vol. 21, pp. 149-163, 1973. F. C. Liu and Z. Y. Ma, "Superplasticity governed by effective grain size and its distribution in finegrained aluminum alloys," Materials Science and Engineering: A, vol. 530, pp. 548-558, 2011. K. Wang, F. C. Liu, P. Xue, B. L. Xiao, and Z. Y. Ma, "Effects of heating rates on microstructure and superplastic behavior of friction stir processed 7075 aluminum alloy," Journal of Materials Science, vol. 50, pp. 1006-1015, 2014. M. Azarbarmas, M. Aghaie-Khafri, J. M. Cabrera, and J. Calvo, "Microstructural evolution and constitutive equations of Inconel 718 alloy under quasi-static and quasi-dynamic conditions," Materials & Design, vol. 94, pp. 28-38, Mar 2016. Y. C. Lin, K. K. Li, H. B. Li, J. Chen, X. M. Chen, and D. X. Wen, "New constitutive model for hightemperature deformation behavior of inconel 718 superalloy," Materials & Design, vol. 74, pp. 108-118, Jun 2015. K. Wang, F. C. Liu, P. Xue, D. Wang, B. L. Xiao, and Z. Y. Ma, "Superplastic Constitutive Equation Including Percentage of High-Angle Grain Boundaries as a Microstructural Parameter," Metallurgical and Materials Transactions A, vol. 47, pp. 546-559, 2015. F. C. Liu and Z. Y. Ma, "Achieving exceptionally high superplasticity at high strain rates in a micrograined Al–Mg–Sc alloy produced by friction stir processing," Scripta Materialia, vol. 59, pp. 882-885, 2008. T. G. Langdon, "Grain boundary sliding revisited: Developments in sliding over four decades," Journal of Materials Science, vol. 41, pp. 597-609, Feb 2006.
14
Fig. 1. EBSD examination regions: (a) examination cross-section relative to RFW joint; (b) EBSD scanning regions relative to observing cross-section.
15
Fig. 2. Variation of welding power with time for the joints processed at different rotation rate.
16
Fig. 3. Variation of welding power with time for the joints processed at different rotation rate.
17
Fig. 4. Microstructure distribution across weld centerline of weld I (500 rpm): (a) grain size, (b) ratio of LAGBs and (c) percentage of twin boundaries along edge of tube wall; (d) grain size, (e) ratio of LAGBs and (f) percentage of twin boundaries along center of tube wall.
18
Fig. 5. EBSD maps of Locations I-V. Grain orientation maps of (a) Location I, (c) Location II, (e) Location III, (g) Location IV. Orientation spread maps of (b) Location I, (d) Location II, (f) Location III, (h) Location IV.
19
Fig. 6. Grain orientation map (a) and orientation spread map (b) of Location V.
20
Fig. 7. Grain orientation map (a) and orientation spread map (b) of Location VI.
21
Fig. 8. Peak temperatures at on the tube wall at different distances from the end of tube.
22
Fig. 9. Grain size distribution of (a) TMAZ adjacent to RXZ, (b) Location IV, (c) Location V, and (d) Location VI (Curve fitting were presented as red lines).
23
Fig. 10. Microstructure distribution along weld centerline of weld I: (a) grain size, (b) ratio of LAGBs and (c) percentage of twin boundaries.
24
Fig. 11. Microstructure distribution across the weld centerline as a function of weld rotation rate: (a) grain size, (b) ratio of LAGBs and (c) percentage of twin boundaries along edge of tube wall; (d) grain size, (e) ratio of LAGBs and (f) percentage of twin boundaries along center of tube wall.
25