Effect of deformation and heat treatment on the distribution of grain boundary characteristics of 825 alloy pipes
Nickel-based Incoloy 825 alloy is a Ti-stabilized Ni-Fe-Cr-Mo-Cu solid solution-strengthening alloy developed by Inconel in 1952. Due to its good stress corrosion cracking resistance, crevice corrosion resistance, oxidation resistance, and non-oxidizing thermal acidity, it has been widely used in machinery, the chemical industry, electronics, boilers, environmental protection, aviation, instrumentation, and other industrial equipment. With the rapid development of the economy, industrial production requirements for equipment and components continue to improve, and further improving the material’s performance has become a matter of concern.
As one of the important structural features of polycrystalline materials, grain boundaries have an important influence on the performance of materials. Many properties of materials are closely related to the properties of grain boundaries, such as grain boundary diffusion, corrosion, intergranular fracture, slip, deflection, etc., which occur in materials that are affected by the grain boundary structure. Coincidence site lattice (CSL) grain boundaries with low sigma values have a high structural order, low energy, and better properties than random grain boundaries. To optimize and improve the properties of materials, Watanabe proposed the concept of “grain boundary design and control,” which was later developed into the research field of grain boundary engineering (GBE), i.e., to improve the proportion of grain boundaries with special structures and control the distribution of grain boundary networks through suitable deformation and heat treatment processes, thus significantly improving the properties related to grain boundaries. The GBE technique can significantly improve the properties of Lehockey et al. treated 625, V-57, and 738 alloys using small deformation cold working combined with annealing at 975-1200°C, resulting in three times lower sigma CSL grain boundary ratios compared to conventional alloys and improved corrosion, creep, and fatigue resistance of the alloys. Thomson and Randle annealed Ni200 alloy at 750°C for 12, 21.5, 24, and 26.5 h after deformation of 6%-7% and found that the alloy’s low ΣCSL grain boundary ratio increased after 24 h of annealing treatment. A large number of Σ3 annealed twins were obtained, which effectively prevented the propagation of intergranular cracks. The low ΣCSL grain boundary ratios were 47.5%, 57.1%, and 72.4% obtained after annealing at 900°C for 1, 2, and 3 h, respectively. Xi et al. annealed the cold-drawn 7% 690 alloy pipe at 1100°C for 5 min, and the low ΣCSL grain boundary ratio was increased to 75%, and the intergranular corrosion resistance was greatly improved. Kumar et al. annealed the 600 alloy at 1000°C for 15 min, repeated this deformation and heat treatment process seven times, and obtained a low ΣCSL grain boundary ratio of 70%. Special grain boundaries blocked the random grain boundary network. Ding Xia and Chen Wenjue annealed cold-rolled 5% 316 stainless steel at 1050°C for 30 min, and the low ΣCSL grain boundary ratio exceeded 80%.
Although many studies have been reported on the process methods to improve the proportion of low ΣCSL grain boundaries in materials, the effects of different deformation and heat treatment process parameters used for different materials on the grain boundary character distribution (GBCD) are also different. The effects of process parameters on GBCD during GBE treatment need the influence of process parameters on GBCD during GBE treatment needs to be further understood. Therefore, to apply GBE technology to production, it is necessary to study the process method to improve the ratio of low sigma CSL grain boundary in the material so that the research results can be extended with the current industrial production process parameters. In this work, the cold-drawing process and annealing treatment of nickel-based 825 alloy pipes with different deformations were carried out with the help of the electron backscatter diffraction (EBSD) technique to study the effects of cold-drawing deformation and heat treatment process on the grain boundary ratio and grain boundary network distribution characteristics of the special structure of 825 alloy.
1. Experimental method
The experimental metal material was a nickel-based 825 alloy pipe with the following chemical composition (mass fraction, %): Cr 21.5, Fe 22.00, C 0.01, Al 0.20, Ti 0.90, Si 0.45, Mo 3.00, Mn 0.80, Cu 2.25, S 0.015, P 0.025, Ni remainder. The solid solution-treated pipe samples were first cold-drawn on the factory YLB-B-5/20-15 cold-drawing machine at 3%, 5%, 7%, and 10%, respectively. Then the samples were uniformly annealed at 1050, 1075, 1100, and 1125°C for 10 min for different cold-drawn deformations, respectively, and quenched quickly after the heating. In this work, a small form variable cold working combined with high temperature (equivalent to solid solution temperature) short time annealing process is used, which can be connected with the production process parameters of the plant.
The samples were prepared by electrolytic polishing with 20% HClO4+80% CH3COOH (v/v) and polished at room temperature for about 90 s using a 30 V DC power supply, followed by metallographic etching with an etchant of HNO3: HCl: H2O=1:1:1 (v/v) for about 5-10 s to reveal the grain boundaries. The sample surface was scanned point by point and line by line using the HKL-EBSD attachment equipped with a CamS-can Apollo 300 thermal field emission gun scanning electron microscope (SEM) with a scan step of 3 μm and a scan area of 1050 μm×660 μm. Information. The test results were analyzed and processed by HKL-Channel5 software, and the measurement system was used to determine the grain boundary type using the Palumbo and Aust standards (Δqmax=15°Σ-5/6, where Δqmax refers to the maximum deviation angle between the actual measured CSL orientation relationship in the experiment and the CSL orientation relationship in the traditional geometric sense). The etched sample’s surface was observed using a VHX-100 digital metallographic microscope (OM).
2. Experimental results and discussion
Figure 1 shows the solid solution OM images, the distribution of different grain boundaries, and the grain orientation distribution of the nickel-based 825 alloy pipes. The average grain size of the sample is 10.5 μm using the equivalent circle diameter method in HKL Channel5 software (in this work, twins are counted in the grain size calculation). The percentage of low ΣCSL grain boundaries in this solid solution pipe is 43.4% (length fraction, same below), most of which are Σ3 twin boundaries, 41.6%. In comparison, the percentage of multiple twin boundaries Σ9+Σ27 is 1.4%, which is very low.
The EBSD of the 825 alloy after cold-drawing and annealing at different temperatures with different small deformation variables is shown in Fig. 2. It can be seen that the low ΣCSL grain boundary ratio of the GBE-treated samples was significantly increased compared with the solid solution samples, in which the low ΣCSL grain boundary ratio of all the samples reached more than 70.0% except for the samples annealed at 1050-1125°C after 10% cold-drawing deformation, and the Σ3 grain boundary ratio exceeded 60.0%, while the low ΣCSL grain boundary ratio of the samples annealed at 1050-1125°C after 10% cold-drawing deformation was significantly increased compared with the solid solution samples. The annealed samples’ low ΣCSL grain boundary ratio also accounted for about 60.0%, and the Σ3 grain boundary ratio exceeded 50.0%. Whether the samples were annealed at 1050°C after 3% cold-drawing deformation or annealed at 1125°C after 10% cold-drawing deformation, the low ΣCSL grain boundaries were mainly Σ3 grain boundaries, accounting for about 87.0% of the overall low ΣCSL grain boundaries, followed by multiple twin boundaries Σ9 and Σ27. The content of other low ΣCSL grain boundaries was very small, indicating that after cold-drawing with small deformation, The percentage of Σ3 grain boundaries in the samples annealed at 1125°C after 10% cold-drawing deformation was only 53.7%, which was significantly lower than the percentage of Σ3 grain boundaries in the other groups of deformation annealed samples. Secondly, for the samples deformed by 3% cold-drawing, the low ΣCSL grain boundary ratio after annealing at 1050°C was significantly higher than that after annealing at higher temperatures; the low ΣCSL grain boundary ratio after 10% cold-drawing and annealing at 1125°C was significantly lower than that after annealing for the other deformed samples; while the low ΣCSL grain boundary ratio after 5% cold-drawing and annealing at different temperatures was significantly lower than that after annealing for the other deformed samples. The samples’ low ΣCSL grain boundary ratios after 5% cold-drawn deformation and annealing were significantly higher than those after 7% or 10% cold-drawn and annealing. Therefore, for the 825 alloy pipe in this work, 5% cold-drawn deformation and annealing at different temperatures for 10 min is an ideal GBE treatment. The sample’s low ΣCSL grain boundary ratio can be increased to more than 75.0%, and the influence of annealing temperature is less. Of course, the deformation and annealing processes applied in the GBE treatment are also related to the original microstructure state, such as the original grain size, carbide precipitation state, and weave structure. In addition, it can be seen from Fig. 2c that the highest percentage of low ΣCSL grain boundaries was obtained when the 825 alloy was cold drawn and deformed by 3% and annealed at 1050°C for 10 min. The percentage of low ΣCSL grain boundaries decreased as the annealing temperature increased. On the one hand, it is because the smaller the deformation variables in the GBE treatment, the lower the recrystallization nucleation density in the subsequent annealing process, the larger the potential space available for nuclei to grow, and the constant generation of annealing twins when the migration of grain boundaries at the recrystallization front sweeps away the deformed matrix, which is conducive to the formation of a high proportion of low ΣCSL grain boundaries; on the other hand, because the high-temperature conditions promote the grain growth when treated above 1050°C (Fig. 2d). On the other hand, since the high-temperature condition promotes the grain growth when treated above 1050°C (Fig. 2d), after the recrystallization is completed, during the grain growth process. Generally, the large-angle grain boundaries are migrated, and the existing low ΣCSL grain boundaries are swept away so that the proportion of low ΣCSL grain boundaries decreases.
Figure.1 OM image, grain boundary distribution, and grain orientation distribution of 825 alloy pipe in solid solution
After the GBE treatment, the distinctive feature of the grain boundary network is the formation of a large number of annealed twins, and there is also a specific orientation relationship between the annealed twins, and multiple twin boundaries (Σ9 and Σ27, etc.) are derived when the twins on different {111} surfaces meet. Therefore, the sum of Σ9 and Σ27 grain boundary ratios should have a good correspondence with the Σ3 twin ratio, which means that when the Σ3 ratio is high, the sum of Σ9 and Σ27 grain boundary ratios is also relatively high. From the variation curves of the low ΣCSL grain boundary ratio with annealing temperature given in Fig. 2, it can be seen that after cold-drawn deformation of 825 alloy at 3%, 5%, 7%, and 10% and annealed at different temperatures for 10 min, the variation trend of the sum of Σ9 and Σ27 grain boundary ratios is similar to that of Σ3 grain boundary ratio with annealing temperature. The ratio of Σ3 grain boundaries decreased with the increase of annealing temperature after 3% cold-drawing deformation of the specimens. Similarly, the sum of Σ9 and Σ27 grain boundary ratios also decreases with the increase of annealing temperature.
Figure.2 Effect of cold-drawing deformation and annealing temperature on the distribution of grain boundary characteristics of 825 alloy
Figure 2d gives the variation curves of the average grain size of 825 alloy with the cold-drawing deformation and annealing temperature. It can be seen that, compared with the solid solution samples, the grain size of the samples increased slightly after different cold-drawing deformation and annealing treatments and decreased with the increase of cold-drawing deformation before annealing and increased with the increase of annealing temperature. Chen et al. found that in the evolution of the annealed twins of 95% cold-rolled high-purity Ni with different annealing temperatures, the change of Σ3 twin boundary ratio during annealing was divided into two stages; the proportion of Σ3 twin boundaries increased with the increase of grain size during recrystallization, however, when recrystallization was completed, the proportion of Σ3 twin boundaries decreased significantly with the further increase of grain size due to the formation of new small-angle grain boundaries during the subsequent grain growth phase. In this experiment, it can be seen from Fig. 2c and d that the percentage of low ΣCSL grain boundaries decreases with increasing grain size when the specimens are annealed at 1050-1125°C after 3%, 7% and 10% cold-drawn deformation; however, the percentage of low ΣCSL grain boundaries do not change significantly with the apparent increase in grain size when the specimens are cold-drawn deformed and annealed at 5%. The higher the ratio of (Σ9+Σ27)/Σ3, the more fully developed the multiple twins are. The ratios of (Σ9+Σ27)/Σ3 grain boundary ratios of 825 alloy after different cold-drawn deformation and annealing temperatures are presented in Figure 3. The (Σ9+Σ27)/Σ3 ratio of the sample after 3% or 10% cold-drawn deformation gradually decreases with increasing annealing temperature. In comparison, the (Σ9+Σ27)/Σ3 ratio after 5% or 7% cold-drawn deformation and annealing treatment at different temperatures is higher than that after 3% or 10% cold-drawn deformation annealing treatment and at 1050-1125°C, the (Σ9 +Σ27)/Σ3 ratios at 1050-1125°C, indicating that the multiple twins are more fully developed in 825 alloy pipes after 5% or 7% cold-drawn deformation before annealing. Figure 4 shows the different grain boundaries of 825 alloy after different cold-drawn deformation and annealing temperature treatments, respectively. It can be seen that the GBE-treated sample contains many annealed twins and their derived multiple twin boundaries, which are connected and form a large number of Σ3n-type trigonal grain boundaries (n=1,2,3…), thus forming a network of grain boundaries characterized by the microstructure of large-size “clusters of grains with mutual Σ3n orientation”. In these clusters, all grains have Σ3n orientation differences whether they are adjacent to each other or not, thus forming a large number of interconnected Σ3n type trigonal grain boundaries, the most important of which are Σ3-Σ3-Σ9 and Σ3-Σ9-Σ27, and the grain clusters are composed of random grain boundaries. In Fig. 5, grains 1, 2, and 3 are any three grains in cluster C1 that are not adjacent. The HKL-Channel5 software can calculate the orientation relationship between any two grains by counting each grain’s average orientation Euler angles in the scanned area. q represents the rotation angle, [HKL] represents the rotation axis, Σ represents the inverse of the dot density of the relocation array, and Δq represents the maximum deviation angle between the experimentally measured CSL orientation relationship and the CSL orientation relationship in the traditional geometric sense. Statistically, the orientation relationship between grains 1 and 2 is Σ81b/0.84°; the orientation relationship between grains 1 and 3 is 38.4° [011], Σ9/0.89°; the orientation relationship between grains 2 and 3 is 54.5° [322], Σ81c/0.26° (b, c denotes the CSL grain boundaries with the same value obtained by rotating the crystal axes by different angles), indicating that they all have Σ3n orientation difference relationship.
Figure.3 Ratio of (Σ9+Σ27)/Σ3 grain boundaries after different cold-drawing deformation and annealing temperature treatments
They are comparing Figs. 4a-d, it can be seen that the ratio of low ΣCSL grain boundaries is increased to different degrees after cold-drawing with small deformation and annealing at 1050°C, resulting in the formation of large size “clusters of grains with mutual Σ3n orientation” microstructures. However, the grain cluster sizes differed significantly when the cold-drawn deformation differed. The grain cluster sizes of the samples annealed at 1050°C after 3%, 5%, 7%, and 10% cold drawing deformation are shown in Figure 6. The sample with the highest percentage of low ΣCSL grain boundaries and the largest grain cluster size after annealing at 1050°C for 10 min after 3% cold-drawn deformation, and the sample with the lowest percentage of low ΣCSL grain boundaries and the smallest grain cluster size after annealing at 1050°C for 10 min after 10% cold-drawn deformation. With the increase of cold-drawn deformation before annealing, the grain cluster size decreases, and the percentage of low ΣCSL grain boundaries decreases. However, it still contains some Σ3n type of grain boundaries. Numerous experimental results show that the grain clusters grow from a single recrystallized nucleus during the recrystallization process and then form Σ3n grain boundaries during the growth process by forming a series of annealed twins and multiple, and the growth meeting between different orders of twins. After the deformation of the material, the work done by the external force will be stored in the form of distortion energy inside the material, which will be recovered, recrystallized, and grain growth will occur after reheating and annealing. Recrystallization is replacing all the deformed tissue with new distortion-free grains driven by the unreleased stored energy of the twisted metal after recovery. During the GBE process, the samples with smaller deformation have less deformation storage energy. Hence, the recrystallization nucleation density is low during the subsequent annealing process, and the potential space for nucleation growth is larger. For fcc materials with low lamellar dislocation energy, annealing twins are easily formed during the growth of new grains. The recrystallized grains continuously produce annealed twins as they grow, forming first-generation, second-generation, and higher-generation twins, thus constituting a very long twin chain that can form many interconnected Σ3n grain boundaries. The new grain clusters continuously engulf the deformed substrate and grow up. When they have finished engulfing the deformed substrate, the recrystallization process is completed, and a grain cluster of a certain size is formed. However, suppose the deformation amount is too large. In that case, the recrystallization nucleation density is very high, and the space for grain growth is very small, so it is difficult to form a large grain cluster microstructure. Thus, the ratio of low ΣCSL grain boundary is relatively low. Therefore, small deformation and high-temperature annealing can obtain large-size grain clusters and a high percentage of low ΣCSL grain boundaries.
Figure.4 Different types of grain boundaries after different cold drawing deformation and annealing temperature treatment
Figure.5 Different types of grain boundaries and grain orientation distribution of grain cluster C1 in Figure 4a
Figure.6 Effect of cold-drawing deformation on grain cluster size of 825 alloy after recrystallization and annealing at 1050°C
3. Conclusion
- (1) After cold-drawn deformation of 5%, the proportion of low ΣCSL grain boundaries in 825 alloy pipe increased significantly after 10 min of recrystallization annealing at 1050°C, while large size “clusters of grains with mutual Σ3n orientation difference” were formed, and the size of grains within the clusters was relatively small.
- (2) With the increase of cold-drawing deformation before recrystallization and annealing, the size of grain clusters decreases, the proportion of annealed twins decreases, and the proportion of low ΣCSL grain boundaries also decreases.
- (3) The effect of annealing temperature on the distribution of grain boundary characteristics was small for alloy 825 pipe after 5% cold-drawing deformation treatment. The percentage of low ΣCSL grain boundaries decreased with increased annealing temperature for samples after 3%, 7%, and 10% cold-drawing deformation treatments.
Authors: Qing Zhao
