Accessible Published by De Gruyter August 18, 2021

Development of an accelerated test for pitting corrosion of CUSTOM 450 by rDHM

Omid Pedram, Esmaeil Poursaeidi, Ramin Khamedi, Hassan Shayani-jam and Yousef Mollapour

Abstract

A key application of CUSTOM 450 alloy is in the construction of gas turbine compressor blades. The study of pitting corrosion can prevent the failure of many gas turbine compressor blades. In this study, a reflective digital holography microscopy method was employed to investigate the growth of pitting corrosion in depth. To this end, a constant potential of 350 mVSCE in a 3.5 wt.% NaCl solution was applied to the specimen. The generated pits were simulated in three dimensions, and it was indicated that pitting corrosion rate was decreased as time passed. Comparing the obtained experimental data with the data gathered from the real industrial environment surrounding a compressor installation, an accelerated test was proposed. By the proposed accelerated test, it is possible to produce a pit similar to the one that will be initiated and propagated at any time in the future in real conditions.

1 Introduction

Corrosion, pitting, and stress corrosion cracking (SCC) frequently happen in chloride-containing environments. Moreover, chloride (Cl) affects the corrosion of structural materials in the petrochemical industry and marine environment [1, 2, 3].

Numerical [4, 5, 6, 7] and experimental methods, such as eddy current [8], optical microscopy, scanning electron microscopy (SEM), X-ray diffraction [9, 10, 11], electrochemical atomic force microscopy (EC-AFM) [12], confocal laser scanning microscopy (CLSM) [13], digital holography [14, 15], and electrochemical measurements [16], can provide useful information about pitting corrosion.

Poursaeidi et al. [17] investigated corrosion pits created by erosion and corrosion on the surface of a CUSTOM 450 compressor blade using experimental methods such as SEM and energy dispersive X-ray spectroscopy (EDX). Anantha et al. [18] studied the corrosion behavior of tempered AISI 420 martensitic stainless steel in 0.1 M NaCl. Volta potential mapping measured by scanning Kelvin probe force microscopy (SKPFM) indicated higher electrochemical (practical) nobility of the carbides with respect to the martensitic matrix whereas regions adjacent to carbides showed lower nobilities due to chromium depletion. Open circuit potential and cyclic potentiodynamic polarization measurements showed metastable corrosion activities associated with a weak passive behavior and a risk of localized corrosion along certain carbide boundaries. In-situ AFM measurements revealed selective dissolution of certain carbide interphases and martensitic inter-lath regions indicating higher propensity to localized corrosion. Many studies have used electrochemical measurements to investigate pitting corrosion behavior in chloride environment for high strength stainless steels [19, 20, 21]. Stainless steels, like other passivable metals and alloys, are prone to pitting in halide ion-containing electrolytes. Pitting manifests as a breakdown of the passive film at local areas giving raise to the formation of microgalvanic cells in which the anode is constituted by the pits and the cathode is the large surrounding unaltered passive surface [22]. The pit nucleation potential (Enp), is usually defined as the lowest potential from which pitting nucleates and develops on a passive surface. Similarly, the pit protection potential (Epp), is the potential below which pits, once grown, cannot further propagate and hence repassivate. In between Enp and Epp, existing pits can grow but no new pits can nucleate [22]. Frangini and De Cristofaro [22] used the cyclic potentiodynamic polarization technique as an accelerated method to find Enp and Epp of alloys 304, 316, 405, 430, and 436. They used deaerated NaCl and mixtures of NaCl + Na2SO4 solutions as electrolyte.

Habib [23] investigated pitting corrosion and crevice corrosion of titanium alloy, aluminum, copper, aluminum–brass alloy, 304 and 316 stainless steel in seawater, using 3D-holographic interferometry. Wang et al. [24] observed the dynamic pitting processes of iron in 0.5 mol dm–3 H2SO4 originated from chloride ions by the help of digital holography. Li et al. [25] studied in situ with the in-line digital holography of the general corrosion and pitting dynamical processes of the iron electrode in KCl solution. Li et al. [26] reconstructed three-dimensional images numerically to study pitting dynamic processes of the X70 carbon steel in NaCl solution. Following the same procedure as [26], Yuan et al. [27] conducted a similar experiment with copper. Klages et al. [28] integrated transmission DHM with optical and ellipsomicroscopy to follow in-line pitting corrosion of 316 stainless steel in 0.9 wt.% NaCl solution. Yuan et al. [29] monitored pitting corrosion of a 304 stainless steel rod in 0.1 mol dm–3 FeCl3, with on-line reflective digital holographic surface imaging. As reviewed and presented above, many investigations have been done into different alloys and environments, nevertheless, the results are not applicable to the pitting corrosion in the CUSTOM 450 alloy.

Reflective digital holography microscopy (rDHM) is a useful method to examine and identify metallographic microstructures [30, 31, 32, 33]. It takes out topographic information in an adjustable field of view with a lateral resolution like conventional microscopy. The lateral resolution of rDHM, however, can be enhanced by similar super-resolution techniques of conventional microscopy and transmission DHM [34, 35]. In comparison to other 3D metallography methods, an rDHM set up is markedly low-cost and has a straightforward sample preparation procedure.

In this work, a two-point bending CUSTOM 450 was subjected to a constant potential of 350 mVSCE in a 3.5 wt.% NaCl solution for three different time intervals. Accordingly, corrosion pits were initiated and propagated on the surface of the sample. Three pits which were grown for different time spans, were randomly picked and reconstructed three-dimensionally by rDHM, and the depth of pits was measured. The relation between time and depth was obtained from the experiment and compared with the relation that directs the real industrial environment. Based on these relations, the running time of the accelerated test could be determined. This time was considerably less than the actual amount of time required for the propagation of the pits in original industrial conditions.

2 Sample preparation, applying potential, and rDHM requirements

2.1 Sample preparation for electrochemical test

The CUSTOM 450 sample with chemical composition as listed in Table 1, was in fact taken from the first stage of a compressor blade from such a turbine installed in a seaside power plant (Fig. 1a).

Fig. 1 (a) Fractured blade, (b) Two point bending sample.

Fig. 1

(a) Fractured blade, (b) Two point bending sample.

Table 1

Chemical composition of CUSTOM 450 alloy.

Element Ni Cr Mo C Mn Cu W Si P Co V Fe
Wt.% 6.49 14.72 0.706 0.0243 0.714 1. 4 0.0193 0.273 0.0209 0.0494 0.104 Balance

Since the compressor blade is initially without defect before service, it was decided to use smooth samples. Taking the dimensions of the fractured blade into account, and noting that the blade works in the elastic region [6], a two-point bending sample was built, employing the ASTM G39 [36] standard. A wire-cut electrical discharge machine was utilized to prepare the test samples with dimensions of 0.5 × 74 × 5 mm3 (Fig. 1b). Corrosion characteristics of the alloy were described in the literature for instance in [6, 7]. Several wet silicon carbide sandpapers with discrete grit sizes of # 100, # 220, # 400, # 600, # 800, # 1000, # 2000, and # 3000 were used to mechanically abrade the samples. Subsequently, using a 2.5 µm alumina solution, we polished the sample to achieve a mirror-like brightness. Ultimately, it was cleaned with alcohol.

According to the equation presented in ASTM G39 standard (Eq. (1)), and by considering the stress value (σ) as 0.8 the yield stress at the maximum bending region [37, 38], the parting support could be obtained 68 mm.

(1) L=ktE/σ sin 1 (Hσ/ktE)

where the geometrical parameters are described in Fig. 2. The yield strength and modulus of elasticity (E) are 1 060 MPa and 200 GPa, respectively [39, 40].

Fig. 2 The geometrical parameters in the two-point bending test.

Fig. 2

The geometrical parameters in the two-point bending test.

2.2 Sample preparation for metallography

To investigate the microstructure of the sample, the hot mount method was used. Then the sample was etched for 30 s by Ralf’s solution which consists of 100 cc of water, 200 cc of methanol, 100 cc of HCl, 2 g of CuCl2, 7 g of FeCl2, and 5 cc of HNO3.

2.3 Equipment required for applying potential

Martensitic alloys are sensitive to pitting corrosion in chloride environments [41]. Furthermore, due to sodium and chloride presence in the spectrum analysis of the EDX of the fractured blade [6] a 3.5 wt.% NaCl solution was used as the electrolyte, according to the ASTM G44 [42] standard.

The 3.5 wt.% NaCl electrolyte was produced from reagents of analytical grade and twice-distilled water. All electrochemical measurements were performed by means of the OrigaFlex multi-Channel system at room temperature under non-deaerated condition. Using a three-electrode electrochemical cell with a saturated calomel electrode (SCE) as a reference, platinum as an auxiliary electrode, and a CUSTOM 450 stainless steel strip with an area of 4 mm2 as the working electrode, experiments were conducted.

2.4 Required supplies for rDHM depth measurements

As shown in Fig. 3, the collimated beam was produced by a He–Ne laser (MEOS, 633 nm, 5 mW). This beam traveled toward a beam splitter (BS) which divided it into two perpendicular beams. Each beam headed to a pre-defined route. Having gone through a microscope objective (MO1) (Olympus, 10 × , NA 0.25), one of the beams focused on a flat mirror (M). Finally, to build the reference beam, it reflected back on the camera (DCC1545M, Thorlabs, 8-bit dynamic range, 5.2 µm pixel pitch). The information from the pitting region was carried by the other beam. This beam reached the surface of the specimen after going through MO2 (Olympus, 5 × , NA 0.25). Part of a surface not affected by pitting plays the role of a reference to holograms.

Fig. 3 Schematic setup of rDHM; SF: spatial filter, L1: collimating lens, BS: beam splitter, NDF: neutral density filter, MO: microscope objective, L2: tube lens, M1: mirror.

Fig. 3

Schematic setup of rDHM; SF: spatial filter, L1: collimating lens, BS: beam splitter, NDF: neutral density filter, MO: microscope objective, L2: tube lens, M1: mirror.

By a combination of propagated and reference beams and analyzing it in the Fourier domain, the recorded digital hologram could be reconstructed. The intensity I(x, y) and the phase Ø(x, y) of the sample were obtained from the complex amplitude E(x, y), as in Eq. (2):

(2) I(x,y)=|E(x,y) | 2 (x,y)=arctan( I[E(x,y)] Re[E(x,y)] )

Different optical paths originate from the different depth of pits, which, according to Eq. (2), brought about distinct changes in phase.

3 Experimental procedure

3.1 Electrochemical test

Cyclic potentiodynamic polarization was performed by reverse scans back to the starting potential. The potential range –500 to 400 mVSCE was scanned for the test. All the conditions for this test were described in Section 2.3. The potential applied in the rDHM experimental procedure was established from this electrochemical test.

3.2 rDHM experimental procedure

An experiment was designed to determine the pitting growth rate at the maximum bending region. To minimize the effect of side conditions, such as surface perpetration on pitting, one sample was used for three tests. To achieve that goal, the whole sample was insulated except the maximum bending section and the upper part, which had to be connected to the working electrode socket of the OrigaFlex system. As shown in Fig. 4, three sections of the maximum bending location were not insulated. The reason for this is explained below, according to the test procedure.

Fig. 4 Schematic of the sample.

Fig. 4

Schematic of the sample.

  1. First of all, a constant potential was applied to all three sections for 10 min. Next, the sample was cleaned according to the procedure described in ASTM G1 [43] to remove corrosion products from the surface. Then, the sample was observed under the ZSM-1001-3E Esfahan optical microscope and using an 18-megapixel Canon EOS 1200D camera, 3 sections were photographed. Thus, parts 1 to 3 were not insulated and corroded.

  2. Secondly, section 1 was insulated; thus, it seemed it was not present in the second experiment. Sections 2 and 3 were subjected to the potential for 10 min, and the sections (2 and 3) were photographed. While section 1 was not corroded due to insulation, the other two sections were corroded owing to exposure.

  3. In the third stage, sections 1 and 2 were insulated; hence, they were not present in the third experiment. Section 3 was exposed to the potential for 10 min and was photographed. Thus, part 3 was not insulated and corroded, while parts 1 and 2 were insulated and not corroded in this experiment.

According to this procedure, sections 1, 2, and 3 were at the potential for 10, 20, and 30 min, respectively. Thus, in the current stage, the sample was prepared for rDHM. According to the arrangement of Fig. 3, the depths of pits produced in the three sections described in Fig. 4, were measured by rDHM. Therefore, the average depths were obtained for 10, 20, and 30 min, meaning that over a period of 30 min, the average pit growth rate could be achieved.

4 Results and discussion

4.1 Metallography with optical and scanning electron microscopy

Custom 450 is martensitic precipitation alloy. Figure 5 illustrates microscopic image of the sample. Lath martensite microstructure in a ferrite matrix is visible in Fig. 5.

Fig. 5 Microstructure of the sample.

Fig. 5

Microstructure of the sample.

To precisely studying the microstructure, SEM of the sample is illustrated in Fig. 6. In this figure, white precipitates can be clearly seen adjacent to the martensite lath. Dimensions of these precipitations (L1–L5, for instance) from age hardening manufacturing of the alloy are approximately 0.4 micron.

Fig. 6 Scanning electron micrograph of the microstructure, L1 to L5 are diameters of random sphere-like precipitations.

Fig. 6

Scanning electron micrograph of the microstructure, L1 to L5 are diameters of random sphere-like precipitations.

Chemical compositions of the alloy were studied with EDXS. Table 2 shows chemical compositions of the precipitations adjacent to the martensite region. Table 3 indicates chemical compositions of the ferrite matrix with the precipitations around it.

Table 2

Chemical compositions of the precipitations adjacent to martensite region.

Name Series wt.%
C K 36.65
Cr K 2.86
Fe K 8.62
Nb L 51.87
Table 3

Chemical compositions of the ferrite matrix with the precipitations around it.

Name Series wt.%
Si K 0.81
Cr K 14.75
Mg K 1.00
Fe K 73.35
Ni K 4.62
Cu K 1.88
Nb L 3.27
Mo L 0.23

Initiation and growth of a stable pit due to passive layer breakdown in stainless steel has a different mechanismfrommicro-galvanic corrosion and is more destructive and faster. The pitting corrosion of stainless steel is more expansive which can transit to a crack. Chloride attack to a passive layer and its failure leads to pitting corrosion. Thus, it has no relation with microstructure. It is possible to have pit initiation due to microstructural factors, but the stable pits initiatied due to passive layer breakdown, contain microstructure pits, as they grow.

4.2 Results of cyclic potentiodynamic test

The potential–logarithmic current density behavior of the stressed sample in 3.5 wt.% NaCl solution of the cyclic potentiodynamic test is shown in Fig. 7. It seems the pitting corrosion occurred during the polarization.

Fig. 7 Potentiodynamic polarization curve of CUSTOM 450 sample stressed in 3.5 wt.% NaCl solution for the cyclic potentiodynamic test.

Fig. 7

Potentiodynamic polarization curve of CUSTOM 450 sample stressed in 3.5 wt.% NaCl solution for the cyclic potentiodynamic test.

According to Fig. 7, the pitting potential was 170 mVSCE, and the protection potential was determined to be 0 mVSCE, which was less than the pitting potential. This means, after initiation and growth of the stable pit, the potential reduced to zero as crater depth increases, which led to slow pit growth. Hence, taking Fig. 7 into account, it was decided to use 350 mVSCE, to have stable pit growth in the rDHM progress.

4.3 Results of rDHM

Figure 8 shows the pits initiated and propagated in part of section 3 of each test described in the previous section. Considering the same pits over time reveals that the more time elapsed, the more pits initiated, the bigger the mouth of pits grew, and the deeper the pits became. Figure 8 is taken after surface cleaning at each stage. As shown in Fig. 8, the dimensions of the pits changed in the second ten minutes less than the first ten minutes. Approximately, there was little evidence of change in the mouth dimensions of the pits after 20 min. The number of the pits on the surface are 3 after 10 min, 7 after 20 min, and 12 after 30 min, in the 4 mm2 area.

Fig. 8 The pits initiated and propagated on the surface of the region in section 3 of Fig. 4 at the potential of 350 mVSCE, (a) after 10 min, (b) after 20 min, (c) after 30 min.

Fig. 8

The pits initiated and propagated on the surface of the region in section 3 of Fig. 4 at the potential of 350 mVSCE, (a) after 10 min, (b) after 20 min, (c) after 30 min.

As shown in Fig. 8, after cleaning the corrosion products from the surface of the sample and the pits, it is observed that inside the pits are dark, so the laser light cannot reflect. Inside the pit, therefore, is diffusive and cannot be photographed by rDHM. Accordingly, by physical vapor deposition, the surface of the sample was coated with aluminum (25 µm thickness) in order to make it mirror-like and reflective. A DST3-A Nanostructured Coating Co. machine was used. The sample was coated by aluminum in thermal evaporation mode at a starting pressure of 4 × 10–4 (Torr).

The pits grew, had the same appearance based on the time they spent under the exposed potential. Here, a pit was selected randomly from each section, and rDHMwas measured.

A pit that was grown for 10 min, from section 1 of Fig. 4 (a pit similar to Fig. 8a) was randomly selected, and its depth was measured by rDHM (Fig. 9a). Figure 9b shows the freezes (When light emits from two or more coherent sources, interference happens and creates dark and light zones which is called the freezing area.) formed on this pit. Freeze reconstruction was done in a fuzzy and intense manner concerning the reference surface, which was the sample surface. Figure 9c and d represent the two and three-dimensional reconstruction of the pit. The yellow parts are more profound than the blue parts. By cross-sectioning of each part of the two-dimensional reconstruction, the depth of the pit in that region can be obtained. Figure 9e shows the depth variation of the deep part of the pit cross-sectioned in Fig. 9c. As is clear in Fig. 9e, the maximum depth of the pit after 10 min in a 3.5 wt.% NaCl solution at 350 mVSCE is 250 µm.

Fig. 9 rDHM reconstruction of a pit at a potential of 350 mVSCE for 10 min in 3.5 wt.% NaCl solution: (a) a randomly selected pit from section 1 of the sample, (b) freezes mapped on the pit, (c) two-dimensional reconstruction, (d) three-dimensional reconstruction, and (e) depth measurement at region cross-sectioned in Fig. 9c.

Fig. 9

rDHM reconstruction of a pit at a potential of 350 mVSCE for 10 min in 3.5 wt.% NaCl solution: (a) a randomly selected pit from section 1 of the sample, (b) freezes mapped on the pit, (c) two-dimensional reconstruction, (d) three-dimensional reconstruction, and (e) depth measurement at region cross-sectioned in Fig. 9c.

The same procedure goes for the 20-minute and 30-minutes pits, and the depths are illustrated in Fig. 10e and 11e, respectively. To put it concisely, the maximum depth of the pit after 20 and 30 min is 350 µm and 400 µm, respectively.

Fig. 10 rDHM reconstruction of a pit at a potential of 350 mVSCE for 20 min in 3.5 wt.% NaCl solution: (a) a randomly selected pit from section 2 of the sample, (b) freezes mapped on the pit, (c) two-dimensional reconstruction, (d) three-dimensional reconstruction, and (e) depth measurement at region cross-sectioned in Fig. 10c.

Fig. 10

rDHM reconstruction of a pit at a potential of 350 mVSCE for 20 min in 3.5 wt.% NaCl solution: (a) a randomly selected pit from section 2 of the sample, (b) freezes mapped on the pit, (c) two-dimensional reconstruction, (d) three-dimensional reconstruction, and (e) depth measurement at region cross-sectioned in Fig. 10c.

Fig. 11 rDHM reconstruction of a pit at a potential of 350 mVSCE for 30 min in 3.5 wt.% NaCl solution: (a) a randomly selected pit from section 3 of the sample, (b) freezes mapped on the pit, (c) two-dimensional reconstruction, (d) three-dimensional reconstruction, and (e) depth measurement at region cross-sectioned in Fig. 11c.

Fig. 11

rDHM reconstruction of a pit at a potential of 350 mVSCE for 30 min in 3.5 wt.% NaCl solution: (a) a randomly selected pit from section 3 of the sample, (b) freezes mapped on the pit, (c) two-dimensional reconstruction, (d) three-dimensional reconstruction, and (e) depth measurement at region cross-sectioned in Fig. 11c.

Considering the rDHM results, growth in depth in the final ten minutes is less than the initial ten minutes. This means the growth rate as a function of depth decreases as time passes.

Table 4 illustrates the time and depth of pits obtained from rDHM in Figs. 911. At first, after the failure of the passive layer, pitting corrosion initiated, propagated, and became deeper as time passed. Gradually, either corrosion products covered the wall of the pit or part of it repassivated. As the corrosion rate decreased, the pit growth rate decreased in depth as well. According to Table 4, the pit grew 250 µm in the first ten minutes, while at the last ten minutes it grew just 50 µm in depth.

Table 4

Depth of the pit and its corresponding time at 350 mVSCE potential.

Time (minutes) Depth of pit (micrometers)
0 0
10 250
20 350
30 400

Figure 12a was achieved from Table 4, which indicates decreasing in the slope of the graph at the last interval. Equation (3) was achieved from curve fitting the points of Table 4, which made it possible with a good approximation to obtain the depth of pits at any time at the potential of 350 mVSCE for the sample.

Fig. 12 (a) Depth of the pit and its corresponding time at 350 mVSCE potential, (b) The difference between the practical measurement and the proposed formula.

Fig. 12

(a) Depth of the pit and its corresponding time at 350 mVSCE potential, (b) The difference between the practical measurement and the proposed formula.

(3) D=92.916× t 0.434

in which D is depth in micrometers, and t is the time in minutes. Figure 12b shows the error in percentage between Eq. (3) and Table 4.

From Eq. (3), the time required for pitting corrosion initiation and growth to reach the desired depth can be calculated for the specimen in conditions of the accelerated test.

According to [7, 44], and Eq. (4), the time needed for pitting corrosion initiation and growth to reach the desired depth, under the actual conditions of the compressor blade, can be obtained.

(4) D=37.66× t 0.47 43.72

In which D is depth in micrometers, and t is the time in months.

Table 5 suggests that the proposed accelerated test can model pitting corrosion, which will be initiated and grow in the actual conditions in the future, as long as the requirements presented in the right column of the table are met. Equations (3) and (4) establish the relation of these two states. In other words, by performing the proposed accelerated test, the gas turbine operator can reach a pit that will be seen on the compressor blade in real working conditions before it occurs on the site. Therefore, predictions are made by applying the related confidence coefficients before any failure happens.

Table 5

Occurrence of pitting corrosion in actual working condition and accelerated test.

Real condition (Eq. (4)) Accelerated test (Eq. (3))
1- Components in the air in the power plant site 1- 3.5 wt.% NaCl solution
2- 200 MPa tensile stress [7] 2- Appliance of 0.8 yield stress
3- No potential 3- Appliance of 350 mVSCE potential

5 Conclusions

  1. The pits were reconstructed in 3D through the rDHM method. The pits initiated and propagated at 350 mVSCE for 10, 20, and 30 min with 250, 350, and 400 µm depth, respectively. The depth values also indicated a decrease in pit growth over time.

  2. The relation between time and depth was achieved from the experiment and compared with the relation that governs the real industrial environment to propose the performance time of the accelerated test.

  3. The achieved time was significantly less than the time needed for the blades of the compressor to suffer from defects due to the pitting corrosion in the industrial environment.

  4. The operators of the power plants could reap the benefit of the proposed accelerated test since it makes the modeling of actual pits technically feasible. This allows the operator to predict future blade defects to prevent adverse economic consequences.


Ramin Khamedi Department of Mechanical Engineering University of Zanjan PO Box: 45371-38791 Zanjan Iran Department of Mechanical Engineering School of Engineering and Applied Science Khazar University PO Box: AZ1096 Baku Azerbaijan Tel.: +989122465282
Tel.: +989127435973

References

[1] T. Laitinen: Corros. Sci. 42 (2000) 421 –441. DOI:10.1016/S0010-938X(99)00072-4 Search in Google Scholar

[2] N.A. Mariano, D. Spinelli: Mater. Sci. Eng. A. 385 (2004) 212 – 219. DOI:10.1016/j.msea.2004.06.041 Search in Google Scholar

[3] N.N. Khobragade, M.I. Khan, A.P. Patil: Trans. Indian Inst. Met. 67 (2014) 263–273. DOI:10.1007/s12666-013-0345-° Search in Google Scholar

[4] E.D. Cendales, F.A. Orjuela, O. Chamarraví: J. Phys. Conf. Ser. 687 (2016) 012067. DOI:10.1088/1742-6596/687/1/012067 Search in Google Scholar

[5] A. Turnbull, L. Wright, L. Crocker: Corros. Sci. 52 (2010) 1492 – 1498. DOI:10.1016/j.corsci.2009.12.004 Search in Google Scholar

[6] E. Poursaeidi, O. Pedram: Int. J. Eng. Trans. B. 27 (2013) 785 – 792. DOI:10.5829/idosi.ije.2014.27.05b.15 Search in Google Scholar

[7] O. Pedram, E. Poursaeidi: J. Fail. Anal. Prev. 18 (2018) 423–434. DOI:10.1007/s11668-018-0417-5 Search in Google Scholar

[8] O. Pedram, Y. Mollapour, H. Shayani-Jam, E. Poursaeidi, R. Khamedi: Met. Mater. Int. (2020). DOI:10.1007/s12540-020-00640-w Search in Google Scholar

[9] C. Vasilescu, S.I. Drob, P. Osiceanu, P. Drob, J.M.C. Moreno, S. Preda, S. Ivanescu, E. Vasilescu: Met. Mater. Int. 21 (2015) 242–250. DOI:10.1007/s12540-015-4074-x Search in Google Scholar

[10] E. Poursaeidi, A.M. Niaei, M. Arablu, A. Salarvand: Int. J. Surf. Sci. Eng. 11 (2017) 85 –99. DOI:10.1504/IJSURFSE.2017.084663 Search in Google Scholar

[11] S.W. Baek, J.K. Lee, J.J. Kim, K.J. Kim: Met. Mater. Int. 21 (2015) 479–484. DOI:10.1007/s12540-015-4521-° Search in Google Scholar

[12] H. Chen, Z. Qin, M. He, Y. Liu, Z. Wu: Materials 13 (2020) 668. DOI:10.3390/ma13030668 Search in Google Scholar

[13] M.F. Zuo, Y.L. Chen, Z.L. Mi, H.T. Jiang: J. Iron Steel Res. Int. 26 (2019) 1000–1010. DOI:10.1007/s42243-019-00250-w Search in Google Scholar

[14] L. Li, C. Wang, S. Chen, X. Yang, B. Yuan, H. Jia: Electrochim. Acta 53 (2008) 3109 –3119. DOI:10.1016/j.electacta.2007.11.040 Search in Google Scholar

[15] L. Li, C. Wang, B. Yuan, S. Chen: Electrochem. Commun. 10 (2008) 103–107. DOI:10.1016/j.elecom.2007.11.004 Search in Google Scholar

[16] Z. Zhang, Z.F. Hu, L. He, X.B. Zhang, X.X. Fang, B.S. Zhang, Z.X. Ba: J. Iron Steel Res. Int. 27 (2020) 719 –731. DOI:10.1007/s42243-019-00345-4 Search in Google Scholar

[17] E. Poursaeidi, A.M. Niaei, M. Lashgari, K. Torkashvand: Appl. Phys. A Mater. Sci. Process. 124 (2018) 629. DOI:10.1007/s00339-018-2007-5 Search in Google Scholar

[18] K.H. Anantha, C. Örnek, S. Ejnermark, A. Medvedeva, J. Sjöström, J. Pan: J. Electrochem. Soc. 164 (2017) C85. DOI:10.1149/2.0531704jes Search in Google Scholar

[19] W. Tian, S. Li, N. Du, S. Chen, Q. Wu: Corros. Sci. 93 (2015) 242–255. DOI:10.1016/j.corsci.2015.01.034 Search in Google Scholar

[20] W. Tian, N. Du, S. Li, S. Chen, Q. Wu: Corros. Sci. 85 (2014) 372–379. DOI:10.1016/j.corsci.2014.04.033 Search in Google Scholar

[21] S. Tokuda, I. Muto, Y. Sugawara, N. Hara: ECS Transactions 80 (2017) 1407. DOI:10.1149/08010.1407ecst Search in Google Scholar

[22] S. Frangini, N. De Cristofaro: Corros. Sci. 45 (2003) 2769 –2786. DOI:10.1016/S0010-938X(03)00102-1 Search in Google Scholar

[23] K. Habib: Desalination 166 (2004) 171 –190. DOI:10.1016/j.desal.2004.06.072 Search in Google Scholar

[24] C. Wang, S. Chen, X. Yang, L. Li: Electrochem. Commun. 6 (2004) 1009–1015. DOI:10.1016/j.elecom.2004.07.020 Search in Google Scholar

[25] L. Li, C. Wang, S. Chen, X. Yang, B. Yuan, H. Jia: Electrochim. Acta 53 (2008) 3109 –3119. DOI:10.1016/j.electacta.2007.11.040 Search in Google Scholar

[26] L. Li, C. Wang, B. Yuan, S. Chen: Electrochem. Commun. 10 (2008) 103–107. DOI:10.1016/j.elecom.2007.11.004 Search in Google Scholar

[27] B. Yuan, C. Wang, L. Li, S. Chen: Electrochem. Commun. 11 (2009) 1373–1376. DOI:10.1016/j.elecom.2009.05.008 Search in Google Scholar

[28] P.E. Klages, M.K. Rotermund, H.H. Rotermund: Corros. Sci. 65 (2012) 128–135. DOI:10.1016/j.corsci.2012.08.023 Search in Google Scholar

[29] B. Yuan, Z. Li, S. Tong, L. Li, C. Wang: J. Electrochem. Soc. 166 (2019) C3039-C3047. DOI:10.1149/2.0061911jes Search in Google Scholar

[30] B.P. Thiesing, C.J. Mann, S. Dryepondt: Appl. Opt. 52 (2013) 4426 –4432. DOI:10.1364/AO.52.004426 Search in Google Scholar

[31] Y. Pourvais, P. Asgari, P. Abdollahi, R. Khamedi, A.R. Moradi: J. Opt. Soc. Am. B 34 (2017) B36-B41. DOI:10.1364/JOSAB.34.000B36 Search in Google Scholar

[32] P. Asgari, Y. Pourvais, P. Abdollahi, A.R. Moradi, R. Khamedi, A. Darudi: Mater Des. 125 (2017) 109–115. DOI:10.1016/j.matdes.2017.03.085 Search in Google Scholar

[33] V.F. Rad, R. Khamedi, A.R. Moradi: Mater. Lett. 239 (2019) 21–23. DOI:10.1016/j.matlet.2018.12.020 Search in Google Scholar

[34] Z. Zalevsky, D. Mendlovic: Optical Superresolution, Springer, Berlin (2004). DOI:10.1007/978-0-387-34715-° Search in Google Scholar

[35] M. Aakhte, V. Abbasian, E.A. Akhlaghi, A.R. Moradi, A. Anand, B. Javidi: Appl. Opt. 56 (2017) Δ°-Δ13. DOI:10.1364/AO.56.0000Δ° Search in Google Scholar

[36] ASTM G39, Standard Practice for Preparation and Use of Bent-Beam Stress-Corrosion Test Specimens, ASTM International (1999). Search in Google Scholar

[37] E. Poursaiedi, A. Salarvand: J. Mater. Eng. Perform. 25 (2016) 3448 –3455. DOI:10.1007/s11665-016-2166-5 Search in Google Scholar

[38] Technical datasheet, CUSTOM 450 Stainless, CARPENTER (2009) 1–12. Search in Google Scholar

[39] M. Kamaya, T. Haruna: Corros. Sci. 49 (2007) 3303–3324. DOI:10.1016/j.corsci.2007.01.011 Search in Google Scholar

[40] A. Turnbull: Proc. R. Soc. A 470 (2014) 20140254. DOI:10.1098/rspa.2014.0254 Search in Google Scholar

[41] W. Dietzel, A. Turnbull: Stress Corrosion Cracking (No. GKSS– 2007 –15) GKSS-Forschungszentrum Geesthacht GmbH, Germany (2007). DOI:10.1016/B978-008043749-1/00325-5 Search in Google Scholar

[42] ASTM G44, Standard Practice for Exposure of Metals and Alloys by Alternate Immersion in Neutral 3.5% Sodium Chloride Solution, ASTM International (2013). Search in Google Scholar

[43] ASTM Committee G1 on Corrosion of Metals, Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens, ASTM International (2017). Search in Google Scholar

[44] D. Linden: Long term operating experience with corrosion control in industrial axial flow compressors, In Proceedings of 14th Turbomachinery Symposium, (2011) 12–15. DOI:10.21423/R1X079 Search in Google Scholar

Received: 2020-07-20
Accepted: 2021-05-04
Published Online: 2021-08-18
Published in Print: 2021-08-31

© 2021 Walter de Gruyter GmbH, Berlin/Boston, Germany