Study on Tensile Creep Behavior of 12Cr1MoV Alloy Steel under High-Temperature Alkali Metal Salt Environment for Solar Thermal Power Generation

The heat exchange tubes of solar thermal power generation work in molten salt environment with periodic temperature change. In order to reveal the tensile creep behavior of 12Cr1MoV pipeline steel under high-temperature alkali metal salt environment, the tensile creep behavior of 12Cr1MoV alloy under different applied load and reaction temperature in high-temperature alkali metal chloride salt environment was studied. The results show that the deformation of 12Cr1MoV alloy in 600°C, NaCl-35%KCl mixed salt environment is mainly controlled by diffusion creep; with the increase of stress, the creep life of 12Cr1MoV alloy decreases. The creep fracture mechanism of 12Cr1MoV alloy in 600°C, NaCl-35%KCl mixed salt environment is intergranular ductile fracture; the increase of temperature will enhance the activation and oxidation of the chlorine atoms, thereby accelerating the corrosion of the base metal and increasing the spheroidization speed of the pearlite matrix, and the creep deformation rate of the alloy increases with increasing temperature.


Introduction
12Cr1MoV as an alloy material with high strength and good heat resistance [1][2][3] has been widely used in solar thermal power generation heat exchange tubes. Compared with coal and other fossil fuels, solar thermal power generation has the advantages of high power generation efficiency and strong controllability [4]. However, the heat storage material in solar thermal power generation contains a certain amount of potassium and chlorine, which causes certain corrosion to the heat exchange tubes in the solar thermal power station in the molten state, thereby affecting the service life of the entire system [5][6][7]. At the same time, heat pipes will be deformed by the high-temperature stress and steam pressure [8]. In actual service conditions, heat pipes will experience both hot corrosion and creep attack, and such combined effects were studied recently.
The creep properties of salt-coated alloy are available in literatures. Most studies mainly focus on the influence of a single saline environment on creep rupture life, deformation mechanisms of alloy, and creep properties [9]. Various fac-tors that reduce the creep life of alloys by thermal corrosion, such as the generation and propagation of surface cracks, have been confirmed by previous studies. The creep fracture behavior of Nimonic-90 alloy at high-temperature and hot corrosion environment was studied by Gosh [10]. The results show that the minimum creep rate will increase and the creep life will decrease under the corrosive condition. A large number of cracks and defect structures are produced in the surface of the material, then sprout and expand into the interior, due to the oxidation, sulfidation, and chlorination reactions occurring simultaneously. There is a dispute over whether different alkali metal salts affect the creep behavior of materials. Suryanarayanan et al. [11] indicated that the creep life of the material is reduced in different corrosion conditions, especially in the NaCl-Na 2 SO 4 mixed salt. Bagui et al. [12] show that the creep progress will be accelerated in mixed salt and high-temperature environment, but no significant effect on creep progress with different proportion of NaCl-Na 2 SO 4 mixed salt. In the current study, on He and Xiong's point of view [13], it is interesting to note that creep deformation of microstructures in NaCl-40%Na 2 SO 4 was the weakest, indicating that sulfates can suppress corrosion of molten chlorides. However, the specimens in chloride mixture show relatively high compression creep strain and steady-state creep rate compared with the bare specimens, and this effect accelerates with the increase of temperature.
A considerable research has been done to determine the creep properties and creep crack behavior through the tensile creep test under high-temperature corrosion conditions. However, the interaction between creep and high-temperature corrosion has not been well verified. In this paper, the effect of hot corrosion damages on compression creep properties of 12Cr1MoV alloy was investigated at different range of temperature and tensile stresses in NaCl-35%KCl mixture. This study is aimed at investigating the influence of interaction between hot corrosion and creep properties.

Materials and Methods
The heat-treated 12Cr1MoV heat pipe steel is chosen as the raw material in this experiment. The chemical components are shown in Table 1.
Alkali metal chloride is the main cause of corrosion on the surface of heat exchange tubes during solar thermal power generation [14]. NaCl-35%KCl is selected as hot corrosion environment in our experiments. The study of the relationship between tensile creep property and stress of solar thermal power exchange tube under high-temperature alkali metal conditions has important engineering significance. Therefore, the tensile creep tests under different stress conditions at high-temperature alkali metal environment were designed, as shown in Table 2.
Similarly, temperature has an important effect on the corrosion rate of the solar thermal power exchange tube. Therefore, the experimental scheme of tensile creep of the sample in the high-temperature alkali metal salt at different temperatures is designed, as shown in Table 3.
The experimental equipment is an electronic creep relaxation tester RDL100 manufactured by Changchun Mechanical Science Research Institute Co., Ltd., equipped with an EDC222 controller manufactured by German DOLI company and an AI-808 temperature controller. The surface of the sample was coated with a layer of salt film. All specimens are insulated at experimental temperature for 30 min before experiment to ensure uniform heating.

Results and Discussion
3.1. Creep Curves. Figure 1 shows the relationship between the creep strain and time (creep curves) of 12Cr1MoV in different conditions. It can be seen from Figure 1(a) that there are three typical stages in the creep process. With the increase of pressure, the time of the steady-state creep stage is obviously reduced, and the creep stage enters the accelerated creep stage from the deceleration creep stage, which leads to the fracture failure of the specimen. The creep strains of the three specimens were 0.149, 0.195, and 0.229, respectively, and the steady-state creep rates were 0:913 × 10 −3 /h, 1:12 × 10 −3 /h, and 1:91 × 10 −3 /h, respectively. The results show that the tensile stress has a significant influence on the creep deformation, and the strain value decreases with the increase of the applied load at the hot corrosion specimen.
As shown in Figure 1(b), tensile creep tests of three groups of specimens at different temperatures were carried out in 41.8 h. At 600°C, the sample is broken and the tensile creep strain is about 0.149. However, other two samples at 560°C and 580°C are still in the steady creep stage, there is no obvious instability phenomenon that occurred, and the tensile creep strains of them are only 0.00402 and 0.0136, respectively. By calculating, at the temperature of 560°C, 580°C, and 600°C, the steady creep rates of three specimens are 0:5467 × 10 −4 /h, 2:385 × 10 −4 /h, and 9:125 × 10 −4 /h, respectively. Figure 1(c) is the tensile creep curve of the sample in a mixed salt and salt-free environment at a high temperature of 600°C. Under the same experimental time, there are three complete creep stages in the specimens at the mixed salt environment, while the specimens in salt-free environment are still in the steady-state creep stage. It can be seen from the creep curve that during the first 15 h tensile test, the effect of the presence or absence of salt on the creep is not significant. With the progress of the experiment, the salt-coated samples quickly transitioned from the steady-state creep phase to the accelerated creep phase, until they failed and fractured, the salt-free sample maintains a stable creep strain rate and is still in the steady-state creep stage during the experiment.
where _ ε is the steady-state creep rate, A is a constant related to material characteristics, σ is the creep stress applied, Q C is the apparent creep activation energy (kJ/mol), and R is the molar gas constant with a value of 8.314 J/(mol·K), T is the thermodynamic temperature. When the temperature or stress is constant, the stress exponent and apparent creep activation energy can be expressed as  3 International Journal of Photoenergy Based on the above discussion, it can be found that the creep stress exponent of 200-230 MPa is 0.12. A large number of studies have shown that the stress exponent is directly related to the creep deformation mechanism. The material deformation is controlled by the diffusion creep mechanism at n = 1, the material deformation is a dislocation slip mechanism at n = 3, the material deformation is dislocation climb mechanism in high-temperature environment at n = 5, and the material deformation is the dislocation climb mechanism at low temperature at n = 7 [9]. The stress exponent of this experiment is close to 1 in the above creep deformation mechanism. Therefore, the tensile creep deformation of 12Cr1MoV alloy is controlled by diffusion creep under the condition of 600°C, NaCl-35%KCl. The ln _ ε and 1/T relationship curve is shown in Figure 2; the apparent creep activation energy is 425.6 kJ/mol by calculation.
3.3. Fracture Appearance. Figure 3 shows creep fracture morphology at high temperature in alkali metal salt environment under different stress conditions. Necking phenomenon can be observed in these three figures. The central region of the fracture perpendicular to the tensile stress is the fiber region, which is the starting point of the fracture; the nucleation occurs, grows, and aggregates in this region; thus, a deep hole can be observed, and the cracks radiate from the center to the periphery with flower shape. When the cracks extend to the surface, all specimens occur obvious shear lips which form certain angle with the tensile stress direction. The morphology shows a typical cup cone fracture, which can be judged as ductile fracture. It can be seen that with the increase of tensile stress, the proportion of the shear lip and fiber zone decreases while the proportion of radiation zone increases. Combined with the creep curve of Figure 1(a), it is found that greater stress results in larger radiation zone and shorter steady-state creep time. Figure 4 is the fracture morphology of the samples in 200 MPa and 230 MPa, respectively. It can be seen that the tensile creep rupture fracture mainly distributes dimples which have different shapes and sizes under different stress conditions at high temperature. The size and depth of the dimples are related to the number of voids and plasticity of the void core when the material breaks. If the dimple has a lot of nucleation sites or the material is poorly plasticized, the dimples formed at the time of the fracture are smaller and shallower. Or the dimples are bigger and deeper [10].
Comparing Figures 4(a) and 4(b), the colony boundary is indicated by a red dotted line, dimple size of specimen 1# (200 MPa) is larger than specimen 3# (230 MPa), the area of the colony boundary is smaller, and the cracks primarily propagate along the colony boundary. The stress has a direct influence on the plastic deformation ability of the material. The lower the stress, the better the plasticity and the better the toughness.

Microstructural Change
3.4.1. Microstructure of Section Areas. Figure 5 reveals microstructure characteristics of cross sections where corrosion depth and internal attack can be reflected with mixed salt or salt-free environments at different temperatures. Temperature has significant effect on the microstructure of the section. The load stress is set as 200 MPa. The samples are corroded by alkali metal salt at the high temperature, and the corrosion layer presents a hierarchical structure. The crack direction of corrosion layer is consistent with the direction of loading stress. The outer layer and the second outer corrosion are porous oxide which are easy to fall off. The inner layer, as the boundary layer between oxide and substrate organization, is not completely corroded. The thickness of the corrosion layer at 600°C is about 600 μm, while it is only 30 μm at 560°C. It is shown that with the increase of temperature, more metal matrix is corroded by alkali metal salt, and the interface between the matrix and the corrosion layer will gradually advance to the center of the matrix. In salt-free environment, the oxide layer formed by thermal oxidation is relatively dense and fuzzy and suggests a protective role. Energy-dispersive X-ray spectroscopy (EDS) analytical technique was employed to perform the elemental analysis of the inner and outer corrosion layers; the equipment used was called field emission scanning electron microscope, and the results are presented in Table 4.
According to Table 4, the corrosion outer layer mainly contains Fe, O, and a small amount of Cr, Cl, and Si. Meanwhile, 2.78% of Cr and 1.05% of Cl in the corrosion inner layer were detected. The results show that the Cr 2 O 3 and SiO 2 protective oxide layers formed between the corrosion inner layer and the base layer are destroyed by hightemperature chloride corrosion, resulting in cracking, crushing, and peeling. In contrast to the study on chloride corrosion of alkali metal [15], the destruction mechanism of 12Cr1MoV in high-temperature chlorine salt mixture is the activation oxidation behavior of Cl. Activated oxidation is the process of gradual destruction of the chloride from the surface to the base layer at high temperature. As long as the chloride salt is sufficient and has a certain oxygen partial pressure, the oxidation reaction in the material is equivalent Fe + Cl 2 ⟷ FeCl 2 ð5Þ At the beginning of the corrosion reaction, chlorine (Cl 2 ) is produced by the interaction between the chlorine salt and the metal oxides at high temperature. FeCl 2 is formed by  . Activated oxidation of chlorine in a high-temperature chloride salt corrosive environment is a process of gradual destruction from the surface of the material to the substrate. As long as chlorine salt is sufficient and oxygen pressure is maintained, the oxidation reaction inside the material is equivalent to a cyclic process. Therefore, the continuous activation of oxidation makes the thermal corrosion more destructive to the oxidation of specific heats.

3.4.2.
Microstructure of Substrate Areas. The 12Cr1MoV matrix structure is usually composed of ferrite and pearlite, and a large amount of cementite is distributed on pearlite.
With the continuous increase of temperature in the hot corrosion environment, the corresponding changes in the structure of the material matrix have also occurred. Figure 6 shows the metallographic diagrams in the matrix under different temperature environments, and Figure 7 shows SEM morphology and the corresponding EDS energy spectrum of pearlite. It can be seen from Figure 7 that the ferrite and cementite in pearlite are sandwiched with each other, with a large number and compact arrangement, and the precipitate phase appears at the grain boundary. According to Figures 6(a)-6(c), with the increase of temperature, the atomic force increases, and the long strip cementite accelerates to a point transition, partly changes into a bead shape, then accumulates into pellets and disappears from pearlite, which gradually exposes the ferrite substrate. Combined with 7 International Journal of Photoenergy Figure 1(b), within the same experimental time, the sample at 600°C occurs fracture failure, while another samples are still in the steady-state creep stage at 560°C and 580°C. Meanwhile, the creep deformation rate of the specimen 5# (580°C) is obviously larger than that of the specimen 4# (560°C). As the strength of the material increases with the increase of pearlite number, the spheroidization can weaken the yield strength, compressive strength, and creep limit of the material and accelerate its creep deformation. On the basis of EDS analysis, the results show that the precipitates are mainly composed of 94.15% of Fe, 3.15% of C, 1.34% of Cr, and small amounts of O and Si. And by comparing the relevant research [16], the particles are M 23 C 6 -type carbides, in which M represents Fe and Mo elements. The results show that the number of irregular point-like precipitates gradually increases, the size becomes larger, and the surface becomes rougher. The grain boundary binding force becomes weakened, and the impact toughness and mechanical properties of the material are severely degraded. Therefore, the precipitation behavior of carbides becomes the main reason for the tensile strength of the materials under increasing temperature.

Conclusion
(1) Tensile creep deformation of 12Cr1MoV steel is controlled by diffusion creep at NaCl-35%KCl salt environment with 600°C (2) Creep loading temperature has a great influence on the hot corrosion behavior of high-temperature alkali metal chloride salt. Under the same loading stress, as the temperature increases, the degree of hot corrosion deepens. Its corrosion mechanism is activated oxidation of chlorine (3) Creep loading stress has a direct influence on the fracture failure of 12Cr1MoV steel. The nucleation grows and polymerizes to form micropores, and the micropores increase and grow in the longitudinal and lateral directions and polymerize into microcracks; the greater the stress, the faster the crack propagation, the weaker tensile performance, the shorter the steady creep phase time, and the faster the transition to the third stage of creep, resulting in rapid fracture failure (4) The main cracks are found at the colony boundary, and the cracks primarily propagate along the colony boundary after creep testing (5) Creep has a great influence on the matrix structure of 12Cr1MoV steel in high temperature. Increasing the temperature from 560°C to 600°C, the phenomenon of pearlite and cementite spheroidization becomes obvious and the toughness and tensile properties of the material are weakened, leading to shorter steadystate creep time and accelerating the fracture failure

Data Availability
No data were used to support this study.

Conflicts of Interest
The authors declare that they have no conflicts of interest.