Effect of annealing temperature on microstructures and mechanical properties of a hot-rolled Ti-6Al-4V-0.5Ni-0.5Nb alloy for offshore applications

Titanium alloys have been increasingly used for downhole tubular and components for corrosive oil wells due to their combination of strength, density, and corrosion resistance. The present study investigated the effect of annealing temperature on microstructures and mechanical properties of a Ti-6Al-4V-0.5Ni-0.5Nb alloy developed for oil well application. The α/β- and β-rolled Ti-6Al-4V-0.5Ni-0.5Nb alloy plates were subjected to annealing at different temperatures ranging from 750 to 900 °C and held for 1 h, and the microstructure and mechanical properties were evaluated. The experimental results show that the developed alloy exhibited both high strength and ductility than that of the conventional Ti-6Al-4V alloy. The α/β- and β-rolled alloy after annealing exhibited bimodal structures and provided well-balanced strength, ductility, and impact toughness, high strength of 950−1000 MPa, elongation of 17%−18%, and impact toughness of 90−120 J cm−2. From a mechanical viewpoint, the new alloy after appropriate annealing is suitable for oil well applications.


Introduction
In the past decades, titanium alloys demand for offshore application have increased tremendously. The offshore oil and gas industries rely heavily on titanium and its alloys for a wide range of applications because of their high strength-to-weight ratio and, excellent corrosion resistance, not only in seawater but also in the petroleum refinery environment [1]. For instance, in recent years, titanium alloys have begun to be used for the making of downhole tubular, and set screws of drills. These are components of oil and gas wells, requiring excellent resistance to corrosion and high strength and toughness. Titanium plays a well-deserved role in the oil and gas industries [2].
Among titanium and its alloys, commercially pure titanium (CP-Ti), i.e., Grade 2, is by far the most used titanium for offshore environments, because of its good corrosion resistance, good formability, and weldability [3,4]. However, unalloyed titanium often exhibits relatively low strength, and thus cannot be used as structural components when high strength is needed [5,6]. Therefore, alloying with a trace amount of elements such as Pd, Ni, or Mo can strengthen pure titanium and also enhance its corrosion resistance. Therefore, Ti-0.05Pd and Ti-0.3Mo-0.8Ni alloy systems were developed for higher strength and corrosion resistance [7,8]. In addition, Ti-3Al-2.5V and Ti-6Al-4V alloys were also used in offshore applications [9] because of their high strength, typically a minimum yield strength of 485 MPa and 760 MPa, respectively. Based on these alloys, Ti-3Al-2.5V-0.1Ru and Ti-6Al-4V-0.1Ru were also developed for higher-temperature applications [10] because Pd or Ru contained titanium alloys offer excellent crevice corrosion resistance in high-temperature saltwater. However, the Pd or Ru is rare and expensive. As a result, cheaper elements that can enhance both the mechanical properties and corrosion resistance are preferable to use as alloying elements. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
A recent study reported that a small amount of Ni and Nb addition can improve the corrosion resistance of titanium and its alloys. For example, the addition of Ni can enhance the strength [11], crevice corrosion resistance, and wear resistance of pure titanium [12]. Perez reported that Nb can prohibit the formation of TiO 2 and significantly reduce the water penetration effect of pure titanium oxidized in water [13]. A recent study also reported that Nb addition significantly promoted the formation of continuous Al 2 O 3 layers in the oxide scales because the addition of Nb increased the formation energy barrier of TiO 2 , while it decreased the formation energy barrier of Al 2 O 3 [14].
Based on these studies, a Ti-6Al-4V-0.5Ni-0.5Nb (wt%) alloy was developed for the purpose of offshore applications, such as oil country tubular goods and fasteners, using a small amount of Ni and Nb as the alloying elements. Ni and Nb are relatively cheaper elements than Pd or Ru. However, as a newly designed alloy, the relationship between microstructures and mechanical properties of the Ti-6Al-4V-0.5Ni-0.5Nb alloy remained unknown. The effect of hot working and subsequent heat treatments on the microstructural evolution and resulting mechanical properties also needs further investigation. Therefore, in this study, we investigated the effect of heat treatments on the microstructures and mechanical properties, tensile properties, and impact toughness, of the hot-rolled alloy samples. This study would provide a reference and guidance for the alloy design and engineering application of cost-effective and high-strength titanium alloys for offshore applications, such as oil country tubular goods and set screws of drills.

Materials and experimental 2.1. Materials
The raw material had a nominal composition of Ti-6Al-4V-0.5Ni-0.5Nb (wt%). An ingot with a diameter of 600 mm and a length of 800 mm was smelted by using vacuum arc remelting process. The consumable electrodes were remelted three times to obtain a homogeneous composition. The microstructure of the alloy ingot is shown in figure 1, which is typical of widmannstatten structure of as-cast titanium alloys, consisting of α colonies and thin β layers between α laths. The β-transus temperature of this alloy was measured to be 950 ± 5°C, by using a metallurgical method by observing the microstructures (martensite formation) of the samples heat treated at 930, 940, 950, 960, and 970°C for 1 h and followed by water quench. The ingot was first homogenized at 1200°C and then hot forged into a billet of Φ200 mm multiple times. A piece of billet with a dimension of Φ200 × 300 mm was cut from the forged billet and then hot forged at 930°C into a plate of 60 mm in thickness and 400 mm in width. The forged plate was then hot rolled at 920°C and 1020°C into a thinner plate of 30 mm in thickness and 400 mm in width. They are hereafter referred to as the α/β rolling and β rolling samples, respectively. The hot rolled plates were used as the initial material in this study. The as-rolled samples were then annealed in a preheated furnace at different temperatures of 750, 800, 850, and 900°C and held for 1 h and then followed by air cooling.

Microstructural observations
The samples for microstructural observations were mechanically ground using abrasive papers and then polished using 5 μm, and 1.5 μm diamond paste and 0.04 μm colloidal silica suspension solution. The polished The optical microstructures of the as-rolled and the annealed samples were observed by using an optical microscope (Axiovert A1). A field-emission scanning electron microscope (FESEM: TESCAN MIRA III) equipped with an electron backscatter diffraction (EBSD) detector was also employed to detect the microstructures and grain orientations. The EBSD mapping was performed at a voltage of 20 kV, with a working distance of 15 mm and a step size of 2 μm.

Mechanical properties
The samples for tensile testing and Charpy impact testing were cut from the as-rolled and heat-treated plates, the length direction of the samples paralleled to the rolling direction. The room-temperature tensile tests and Charpy impact tests were carried out as per GB/T 228.1-2010 and HB 5144-96, respectively. The samples for tensile testing had a gauge dimension of 5 mm in diameter and 35 mm in length. The samples were tested at room temperature with a strain rate of 0.001 s −1 using an MTS E45.0 tensile machine. The samples for Charpy impact testing were machined to a standard dimension of 10 × 10 × 55 mm and, a 2 mm V-notch. Charpy impact tests were performed at room temperature by using a NI300C instrumented Charpy impact testing equipment. Three samples of each condition of the tests were measured for average values to ensure the consistency and repeatability of experimental results.

Results and discussion
3.1. As-rolled microstructure Figure 2 shows the initial microstructures of the α/βand β-rolling samples. It is observed that the α/β-rolling sample exhibited an equiaxed grain structure with a large fraction of equiaxed α grains, approximately 65%, and the equiaxed α in the α/β-rolling sample had an average size of approximately 20 μm, as shown in figure 2(a). In the β-rolling sample, owing to the rolling at the temperature above the β transus temperature, the prior β grains are clearly observed, with a grain size of approximately 200 μm, as shown in figure 2(b). In addition, grain boundary α layers are formed along the prior β boundaries, as shown in figure 2(b). The β-rolling sample exhibited a lamellar structure having many α colonies and a very small amount of equiaxed α grains (approximately 8%). The grain size of the equiaxed α grains is approximately 10 μm. Figure 3 shows the microstructures of the α/βand β-rolling samples after annealing at different temperatures. The α/β-rolling samples after annealing at this temperature range exhibited a standard bimodal structure that consisted of both equiaxed α and very fine α laths. The β-rolling samples after annealing also exhibited a bimodal structure; however, they contained slightly finer equiaxed α but coarser lamellar α. Grain boundary α can be also observed in these samples, as shown in figures 3(e)−(h). Additionally, the total fractions of equiaxed α are much lower than those of the α/β-rolling samples after annealing. It is observed in figure 4(a) that the fraction of equiaxed α in the α/β-rolling samples after annealing exhibited a range from 50% to 70%, and the equiaxed α fraction in the β-rolling samples after annealing ranged from 20% to 30%. In addition, annealing results in grain refinement of equiaxed αin the samples, as shown in figure 4(b); however, increasing the annealing temperature did not significantly change the grain size of equiaxed α. Figure 5 shows the EBSD IPF maps of the annealed samples. When annealed at 800°C, grains with orientation gradients can be observed as indicated by arrows in figures 5(a) and (c). This implies that the recrystallization did not accomplish at this temperature. However, the samples annealed at 900°C are fully recrystallized, because in-grain orientation gradients are hardly observed as shown in figures 5(b) and (d). The βrolling samples exhibited finer grain sizes than those of the α/β-rolling samples when annealed at the same temperature. This is consistent with the optical microstructures in figure 3. Additionally, the α/β-rolling and βrolling samples after annealing at 800 and 900°C exhibited significantly different textures. 〈1120〉α tends to align with the rolling direction in the α/β-rolling samples, while 〈0001〉α aligns with the rolling direction in the β-rolling samples. This finding is consistent with the previous study on Ti-6Al-4V [15]. Figure 6 shows the mechanical properties of the samples as a function of annealing temperature. The α/βrolling and β-rolling samples exhibited UTSs of 975 MPa and 1027 MPa, YSs of 955 MPa and 995 MPa, Els of  17% and 18%, respectively. Note that both the strength and ductility are higher than that of Ti-6Al-4V alloy [16]. The β-rolling sample exhibited a similar ductility to the α/β-rolling sample, but higher strength. because the β-rolling sample had the finer equiaxed α and a larger amount of lamellar α. Although the two samples had a similar ductility, the α/β-rolling sample exhibited a much higher impact toughness, 72.3 J cm −2 . In general, Ti-6Al-4V with lamellar structure shows larger impact toughness than that of the alloy with equiaxed structure because lamella α could significantly deflect the crack [17]. The present results showed an opposite phenomenon, probably because the thickness of α lath in the α/β-rolling sample is much smaller than that of the β-rolling sample, as shown in figure 2.

Mechanical properties
After annealing, the strengths of the α/β-rolling and β-rolling samples both decreased compared to their asrolled conditions, as shown in figure 6(a); however, the annealing has no significant impact on the ductility of the α/β-rolling and the β-rolling samples, as shown in figure 6(b). The impact toughness of both the samples after annealing significantly increased, approximately 50%, as shown in figure 6(c). The grain refinement in both  samples after annealing does not have a contribution to a noticeable strengthening but results in improved impact toughness.
Regarding the effect of annealing temperature, the UTS of the α/β-rolling and β-rolling samples decreased with increasing annealing temperature, however, the YS of the samples did not change significantly with increasing temperature, as shown in figure 6(a). The ductility of the β-rolling sample gradually increased with increasing temperature, but the ductility of the α/β-rolling sample maintained an Els of 17%−18%, as shown in figure 6(b). Both the two samples exhibited an increased toughness with increasing annealing temperature, from 55 to 100 J cm −2 for the β-rolling sample, and from 72 to 140 J cm −2 , as shown in figure 6(c). Note that the α/βrolling samples always had much higher impact toughness than those of the β-rolling samples.

Recommendations and future works
Considering the actual environment of offshore applications, further study needs to investigate the corrosion resistance of the new alloy after different annealing under high-temperature and high-pressure conditions and hydrogen sulfide, carbon dioxide, high concentrations of chloride ions, and even elemental sulfur environments.

Conclusions
The effect of annealing on the microstructure and mechanical properties of an α/β-rolled and β-rolled Ti-6Al-4V-0.5Ni-0.5Nb alloy developed for oil application was investigated. The conclusions are as follows: (1)The α/β-rolled alloy exhibited a bimodal structure and had moderate strength and high impact toughness, while the β-rolled alloy exhibited a nearly lamellar structure and had high strength but low impact toughness.
(2)Annealing introduced a significant grain refinement in both the α/β-rolled and β-rolled samples, however, the grain refinement does not contribute to grain-refinement strengthening but gives rise to improved impact toughness.
(3)The β-rolled alloy after annealing always exhibited higher strength but lower impact toughness than that of the α/β-rolled alloy after the same annealing.
(4)Further study needs to investigate the corrosion resistance of this alloy in hostile service environments, especially for high-pressure and high-temperature well applications.