Effect of Laser Shock Processing and Aluminizing on Microstructure and High-Temperature Creep Properties of 321 Stainless Steel for Solar Thermal Power Generation

Key Laboratory of Efficient & Clean Energy Utilization, School of Energy and Power Engineering, Changsha University of Science & Technology, Changsha 410114, China Hunan Province 2011 Collaborative Innovation Center of Clean Energy, and Smart Grid, Changsha 410114, China College of Engineering and Designing, Hunan Normal University, Changsha 410006, China Japan Electric Power Central Research Institute, Tokyo 240-0196, Japan


Introduction
To meet increasing global energy demand in an environmentally sustainable manner, greater emphasis is being given to the development and dissemination of renewable energy technologies. Solar energy is an important renewable energy source that is expected to play a significant role in the future energy supply mix. Concentrated solar power (CSP) technology is an important option for acquiring solar energy which draws increasing attention during past several decades [1,2]. It is internationally believed that for every 1 percentage point increase in the annual average power generation efficiency of solar thermal power generation systems, the average power generation cost will be reduced by 5% to 8%. Therefore, it is vitally significant to improve the power generation efficiency, reduce the cost, and increase service life of the solar thermal power generation system. However, as the key component of CSP, the heat exchange tube should be paid more attention [3][4][5][6].
Austenitic stainless steels (ASS) are commonly characterized by favorable ductility and excellent corrosion resistance [7][8][9]. AISI321, an austenitic stainless steel stabilized with titanium, is a promising material for load-bearing applications in solar thermal power generation, nuclear power reactors, boilers, pressure vessels, expansion bellows, and stack liners [10,11]. The heat exchange pipes made of AISI321 are prone to have creep damage under long-term hightemperature conditions, coupled with the combined effects of corrosive media and pressure within the pipe, to further accelerate pipeline failure and bursting.
Aluminizing can significantly improve the resistance of stainless steel to high-temperature oxidation, friction, and corrosive medium like atmosphere, hydrogen sulfide, seawater, and liquid metals [12,13]. Different techniques as chemical vapor deposition (CVD), hot dip aluminizing [14], pack aluminizing [15,16], electrolytic deposition of Al from ionic liquids [17], and Ar-plasma deposition are applied for depositing Al on the surface of the steel. Among them, the powder pack aluminizing technique is widely used due to its advantages of producing a uniform dense layer on different shapes and sizes of specimens. Ref. [18][19][20][21] analyzed the phase composition of the aluminized layer and found the 321 stainless steel with a thicker layer showed a better resistance of oxidation and corrosion. On the other hand, the coating formed with the normal method of application is relatively porous with a multitude of defects and impurities. In particular, the Al 2 O 3 film or Fe-Al phase may crack, promoting microcrack propagation at a high temperature and stress, which causes the aluminized steel exhibiting a decrease in the creep life in high temperature. Moreover, the creep behavior of aluminized steel exhibits a higher creep rate and shorter creep rupture life than those of uncoated specimens [22,23]. Ref. [24] also pointed out that brittle coatings with poor bearing properties were detrimental to the creep properties of the aluminized steel.
Laser shock peening (LSP) is an advanced surface enhancement method for metallic materials to enhance the resistance of creep, fatigue, and corrosion, which can delay the crack nucleation and expansion [25,26]. In the LSP process, high-amplitude shock waves are generated through rapid expansion of high-temperature plasma induced by the contact between high-power density laser pulse and the material surface. While the shock wave propagates into the material, plastic deformation occurs to a depth where the stress no longer exceeds the elastic limit of the material, which induces residual stresses throughout the affected depth [27,28]. Yella et al. reported that using an absorbent tape as a sacrificial layer was a good way to perform LSP on the stainless steel and it did not cause surface damage [29]. Vázquez Jiménez et al. pointed out that the best LSP path for extending the high-cycle fatigue life was perpendicular to the rolling direction [30]. Lu et al. found that the treated sample displayed better tensile properties such as stronger flow stress and higher ultimate tensile strength with an increasing LSP impact time [31]. To improve the creep resistance and anti-corrosion performance simultaneously, technology of aluminizing with LSP has been already investigated partially. Lai demonstrated that the fatigue life of material with aluminizing and LSP was increased visibly [32]. Ref. [33][34][35][36] also testified that the mechanical property of the aluminized steel was greatly improved by LSP.
In view of the above considerations, many researches mainly focused on the anticorrosion capability, LSP parameters, and the influence on fatigue behaviors. Nevertheless, the studies about high-temperature creep properties of the stainless steel with aluminizing and LSP is rarely reported. Therefore, in order to ensure the stable operation of stainless steel heat exchanger under actual working conditions, the effect of surface modification processes, aluminizing with LSP, on the microstructure and high-temperature creep properties of 321 stainless steel was investigated in this paper.

Materials and Methods
2.1. 321 Stainless Steel. The austenitic steel AISI 321 was procured commercially for this investigation. The result of elemental composition analysis is presented in Table 1.

Powder Pack
Aluminizing. The pack powder mixture consisted of 68 wt% Fe-Al powder, 30 wt% Al 2 O 3 filler, and 2 wt% NH 4 Cl activator. The samples were ground to #2000 grit-sized SiC paper finish and cleaned thoroughly before placing inside the pack, and then, samples were kept inside the pack in an alumina crucible, which was subsequently covered with a lid. Pack aluminizing experiments were carried out in a KSL-1400X box-type sintering furnace, and samples cooled to room temperature after being heated at 950°C for 12 h. The cross-sectional image of the coating layer is shown in Figure 1. The EDS and SEM results indicate that aluminized coating is mainly composed of an Al 2 O 3 outer layer, the transition layer of the Fe-Al phase (such as Fe 3 Al, FeAl, FeAl 2 , and a small amount of AlN), and the diffusion layer with the AlFe(Ni) phase and AlCrFe phase. In the case of aluminizing processes, there was a formation of surface layers of a new composition and structure. The formation of such a local transition zone in the form of individual areas or intermetallic compounds directly in aluminizing led to the fact that high-stress microcracks appear in this area and significant local overstresses cause loss of strength of the solid coating system in thermal deformation action on it.
2.3. LSP Treatment. The LSP was performed with the LAMBER-08 pulsed laser which is operating at a 1064 nm wavelength and delivering 7 J pulse energy in 20 ns, with a 2 Hz repetition rate. The diameter of laser spot is 3 mm, the overlapping rate is 50%, and the laser power density is about 4.95 GW/cm 2 . In order to protect the surface of the metal material from direct laser burning and couple with the laser energy better, the black adhesive tape is selected as the impact protection layer. The constraining layer is water, to constrain the high-temperature plasma generated by the laser irradiation and increase the shock wave pressure [37]. 2.6. XRD and Microstructure. The surface phases of the aluminized steel with different times of LSP impact were analyzed by a TD3000 X-ray diffractometer (XRD). The Quanta 2000 environment scanning electron microscope (SEM) and the attached energy dispersive spectrometer (EDS) were used to examine the creep fracture.

Results and Discussion
3.1. Surface Roughness and Microhardness. Surface roughness plays a dominant role in affecting the creep performance and reflecting the collection characteristics of the material surface. Surface roughness is evaluated by three index, including contour arithmetic mean deviation (R a ), contour maximum height (R z ), and contour unit average width (R ms ), where R a can simultaneously reflect the microscopic geometrical features and the height of the convex peak [38]. If these rough peaks are equivalently approximated as microgaps, the stress concentration of the rough peaks can be calculated using the notch stress formula [39]: where k is the material constant and m is the ratio of R z to R ms . The larger the value of m is, the greater the stress concentration is. Table 2 presents the surface roughness of specimens with different times of LSP impact. As can be seen, the roughness of the aluminized sample is 0.952 μm, and the roughness increases rapidly with the increase of impacts' times. The more the stress concentration is, the easier to generate microcracks, resulting in internal damage of the material. Nevertheless, the roughness is related to the laser absorption, namely, the overly smooth metal surface is not conducive to the absorption of laser [40]. The optical photograph and 3D surface profile for different conditions are shown in Figure 2. It can be seen from Figures 2(a) and 2(d) that the surface of aluminized samples is smooth and flat with regular dimples, and the average pit depth is 7.174 μm. The morphology of the aluminized sample with single LSP impact (Figures 2(b) and 2(e)) displays the existence of obvious pits and steps, whose shapes are square or circle. The shape of the pits depends on the contact pressure between the protective layer and the sample [41]. The average pit depth is 10.813 μm with an increase of 50.7% and the deformation amount is 3.639 μm. The morphology of the sample after three impacts is shown in Figures 2(c) and 2(f). The plastic deformation was aggravated by the plastic loading wave. The depth of the pit is 20.718 μm with an increase of 188.8%, and the deformation amount is 13.544 μm. As the LSP time increases, the roughness and the degree of plastic deformation are alleviated, which attributes to the generation of surface pits predominately. Figure 2(g) is the pulse sequence applied to treat the specimens. It consists of firing pulses in a zigzag-type scanning pattern, covering a total area of 24 × 8 mm in the central zone of the specimens. A laser impact region with a diameter of 1 mm and a depth of 21.5 μm can be observed on the aluminized samples after three LSP impacts, and the plastic deformation is one of the important indexes for evaluating the enhancement effect of LSP. Figure 3 presents the microhardness of aluminized samples before and after LSP. The microhardness of the aluminized sample gradually increases with increasing distance from the penetrating layer to the surface, which could be imputed to intermetallic compounds formed in the penetrating layer mainly. In addition, the grain grows and the material softening phenomenon occurred during the hightemperature aluminizing process. Therefore, the hardness was found to be lower when the depth of the layer was measured to 180 μm. However, subjected to radial shock waves released by LSP, the surface of the aluminized steel would    International Journal of Photoenergy generate severe plastic deformation, advancing the surface hardness of materials significantly. It is well known that plastic deformation will cause work-hardening effect and grain refinement, thus increases the hardness of the work piece. Similar conclusions have been drawn from Ref. [42,43]. As the number of impacts increases, the surface microhardness increases from 395 HV of the nonimpacted sample to 625 HV and 718.4 HV. Figure 4 is the XRD pattern of    International Journal of Photoenergy aluminized samples before and after LSP. There are mainly three phases, FeAl, FeAl 2 , and Fe 3 Al, and none of the new phase is observed after LSP. The position of the crystal surface is hardly changed, while the intensity of the diffraction peak is obviously reduced, and the full-width half maxima (FWHM) become larger. It illustrates that surface grains were refined after LSP, and the grain refinement phenomenon of three-time impact samples was more obvious than single-impact samples, which is consistent with the conclusion of the previous hardness enhancement mechanism.
3.2. Creep Curves. Figure 5 shows the high-temperature creep curves and creep rate curves of the aluminized steels with different times of LSP impact under a series of loading stress. It is observed that all the specimens have similar curve shapes to characterize three stages of typical creep behaviors, namely, the deceleration creep stage, stationary creep stage, and accelerated creep stage. The first stage of creep is shorter while the second stage of creep turns longer. As shown in  (Figures 5(c) and 5(f)) LSP are similar, which implies that the three states of materials were remarkably affected by the external loads. Figure 6 represents the relationship between the applied stress and creep rupture time of the austenitic stainless steel treated with different surface modification processes and all results show that the creep rupture time is inversely correlated to the stress at the creep experimental temperature. At the same time, we found that the steady-state creep rate decreases and creep rupture time extends with the increase of impact times. The steady-state creep rate of specimens with single and three times decreases by an order of magnitude, and their creep rupture lives are increased by 42.8% and 232.1%, respectively, when the applied stress is 210 MPa, which clearly demonstrates that the high-temperature creep resistance of aluminized samples is significantly improved by LSP technology.
In the stationary stage, the effective dislocation density remains unchanged to keep a balance between the material recovering and hardening due to the constant external stress and high-temperature environment. Consequently, the stationary creep rate with the smallest value and the simplest deformation mechanism is the key parameter to reflect the creep behavior of the material. Figure 7 displays that the applied stress is positively relevant to the minimum creep rate of the steady-state creep region. The minimum creep rate and applied stress are well-known to be related by a power law as presented in where _ ε is the stationary creep strain rate, A x is a complex constant correlated to the material structure, σ is the applied stress, n is the creep stress index, Q c is the creep apparent activation energy, R is the molar gas constant (R = 8:314 KJ/mol), and T is the absolute temperature. Equation (2) on both sides conducts the logarithmic processing to obtain Equation (3). When the creep temperature T is constant, lnε and lnσ have the linear relationship of slope n (Equation (4)) as follows: ð4Þ Figure 8 exhibits the logarithmic relationship between the steady-state creep rate and the applied stress. It can be concluded that the n values of aluminizing and aluminizing composite single and three-time LSP samples are 13.73, 19.71 and 23.13, respectively. The matrix alloy is 321 austenite stainless steel, and the values 3 and 5 of stress exponent stand for two important creep types of solid solution. When n ≤ 3, the stress exponent has a typical value of 3, this signifies that the creep behavior is very different from that of a pure metal, which is called the first class of alloy-type creep; when n = 4 − 7, its typical value is 5, which indicates that the creep behavior is similar to that of pure metals, which is called the second pure metal creep; when n ≥ 8, the steady-state creep behavior can no longer be described by power-law equations [44]. Ref. [45] points out that if a metallic had a larger apparent stress exponent, the metal may have a higher creep stress threshold, which can predicate that the aluminizing composite LSP material has a higher creep stress threshold and better creep resistance.
3.3. Fracture Morphology. The creep macroscopic fracture appears the necking phenomenon while the crack-like plastic deformation occurred near the typical ductile fracture. The fracture existence of clear cross-section grains with strong sense of third dimension testifies that it belongs to intergranular fracture. The area shrinkage of impacted specimens is higher than that of nonimpacted specimens, implying that the impacted specimens have a better plasticity. It can be seen from Figure 1 that there are some defects on the coating layer, such as coarse second-phase particles (AlN) and pores. During the high-temperature creep process, the abundant intracrystal cross slips glided to the pore position or grain boundary lattice defects caused by the residual tensile stress, resulting in plenty of micropores. These micropores continued to converge into cracks which caused the expansion and separation of grain boundaries under the tensile stress. According to the SEM morphology of fracture surfaces and the EDS analysis of Figure 9, it can be distinguished that the fracture surface of aluminized steels can be roughly divided into three layers: the outer layer (FeAl), the transition layer (mainly composed of FeAl 2 and Fe 3 Al), and the 6 International Journal of Photoenergy 7 International Journal of Photoenergy austenitic stainless-steel matrix layer. The second phase generated from aluminizing had a large separation tendency from the matrix, which is easily affected by high-temperature environment and the tensile stress. Then, the tendency led to the generation of macropores and eventually fracture. On the one hand, grains of the aluminized layer were refined by the LSP treatment (Figure 9(b)). The rough second phase was evenly distributed on the substrate and the pores' size was narrowed, so that microcracks were difficult to generate. In addition, the laser shock-induced residual compressive stress prevents further creep slips and the separation between crystals. On the other hand, under the impact of high-power lasers, strenuous plastic deformation (Figure 9(c)) accelerates the generation of crystal substructures, changes the direction of grains, and increases the dislocation density. High-density dislocations undergo disorder and discretional slips in all directions at high temperatures to annihilate dislocations with particularly slip orientations. This kind of interaction consumes most time and heat to extend the anticreep life of samples [46]. At the same time, high-density dislocations induce significant differences in orientation amongst grain boundaries, resulting in higher internal stresses impeding dislocations, suppressing grain boundary slippage, and delaying the growth of grains to effectively retard crack initiation [47]. The advance in impact times brings out more obvious grain refinement and plastic deformation; therefore, the creep life of the laser-shocked sample with multiple impacts is vitally improved.
The mechanism of creep damage is the exfoliation, formation, and development of voids followed by the occurrence of failure, which attributes to the plastic deformation incompatibility between intragranular inclusions or second phases and matrix radically. The voids are circular or polygonal, dispersing at the intersection of three grain boundaries or on the grain boundary mainly, where the austenite stainless steel voids preferentially nucleate on the inclined grain boundary [48]. Through connected to one another, some hollows turn into a chain eventually, indicating that the voids formed independently intersect with their growth mutually. In particular, as nucleation centers, inclusion particles are contained in some voids. As can be seen from Figures 10(a) and 10(b), dimples and intergranular cracks are visible in the propagation region. Besides, there are several fracture edges' dimple accumulation, secondary cracks, and parallel stripe structures around. After initiating along grain boundaries, the cracks stretch rapidly and converge into larger cracks during the creep process, leading to the separation among grains. Although creep fractures occur at the grain boundaries predominantly, the creep rupture is influenced by the state, structure, and precipitates of the grain boundary seriously. As described in Figures 10(c)       9 International Journal of Photoenergy concentrations around particles to cause breakage. Nevertheless, the grain size of the second-phase particles (or inclusions) in aluminized samples with LSP impacts is significantly refined, and the degree of stress concentration is reduced to enhance the creep resistance. In addition, the austenite stainless steel is subjected to soften during the high-temperature aluminizing process, leading to grain coarsening and plasticity deterioration. But grain refinement increased plasticity after LSP, the plasticity of materials is closely related to creep behaviors. Therefore, the impact of laser on the grain refinement of the aluminized steel is the microscopic reason for the improvement in creep performance.

Conclusion
In this study, the effect of laser shock processing and aluminizing on microstructure and high-temperature creep properties of the 321 stainless steel were investigated. The main conclusions are described as follows: (1) A tightly bounded dense iron-aluminum coating was obtained on the surface of the 321 austenitic stainless steel by powder-embedded aluminizing. The roughness, surface plastic deformation, and surface microhardness of the aluminized specimens were significantly increased after LSP treatment, and the strengthening effect of aluminizing specimens with LSP three times is most obvious (3) The dimple size, void size, and density of impacted samples became smaller than those of nonimpacted samples. As the number of LSP increases, the dislocation density and the plasticity of material increase because of the more obvious grain refinement and plastic deformation on the surface, which comprehensively improves the creep performance of the aluminized steel

Data Availability
All data used to support the findings of this study are included within the article.

Disclosure
An earlier version of this paper was partly presented in terms of an invited report at the 25 th Congress of International Federation for Heat Treatment and Surface Engineering (http://www.25ifhtse.csp.escience.cn/dct/page/70051).

Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this paper.