This study aimed to determine the optimal heat treatment and build orientation to minimize the susceptibility of additively manufactured (AM) Alloy 625 to crevice corrosion. To accomplish this, metal-to-metal and acrylic-to-metal remote crevice assembly (RCA) experiments were performed for as-made (NT) AM, stress-relieved (SR) AM, solution-annealed AM, and solution plus stabilization-annealed AM Alloy 625 in two different build orientations. Current vs. time data from metal-to-metal RCA experiments were analyzed using commercially available statistical analysis software used to perform analysis of variance. While there was a lack of statistical evidence that build orientation affects crevice corrosion susceptibility, there was strong evidence heat treatment affects crevice corrosion susceptibility. Specifically, according to Tukey’s Multiple Comparison, alloys that were heat treated had a statistically significant lower charge passed as compared to the NT specimens. This finding was consistent with measured penetration depth where NT AM specimens had the largest maximum penetration depth. In contrast, acrylic-to-metal RCAs were used to calculate crevice corrosion current density (rate) and repassivation potential. While current densities for the AM materials were comparable, the forward motion of the active crevice corrosion front on the NT and SR specimens was found to be slow, resulting in high damage accumulation locally. Both metal-to-metal and acrylic-to-metal RCA results are discussed within the context of nonhomogenized microstructures associated with AM.

Crevice Corrosion of Alloy 625

Alloy 625 is a nickel-based superalloy with mechanical and corrosion resistance properties, making it attractive for high-temperature, marine, and nuclear applications.1-2  However, challenges associated with Alloy 625 include susceptibility to crevice corrosion and difficulty of machining via conventional methods owing to its high strength and thermal resistance. While additive manufacturing (AM) is a method for overcoming machining challenges, little is still known about the corrosion performance of AM structures in environments they will be exposed to.

The nominal composition of Alloy 625 consists of Ni (58 mass% min.), Cr (20 mass% to 23 mass%), Fe (5 mass% max.), Mo (8 mass% to 10 mass%), Nb (3.15 mass% to 4.15 mass%), and trace amounts of C (0.1 mass% max.), Mn (0.5 mass% max.), Si (0.5 mass% max.), Al (0.4 mass% max.), and Ti (0.4 mass% max.).1  Alloy 625 owes its corrosion resistance to the presence of chromium (Cr), molybdenum (Mo), and niobium (Nb). Specifically, Cr tends to form a passive film, a thin oxide layer in the order of nanometers in thickness, that protects the surface. Mo is believed to increase the protectiveness of the passive film3-4  while Nb has been reported to enhance corrosion resistance.5-7  While it is considered a corrosion-resistant alloy (CRA), Alloy 625 is known to be prone to crevice corrosion.8-10 

Miller and Lillard11  investigated crevice corrosion of Alloy 625 in ASTM artificial seawater where they studied current vs. time data for remote crevice assemblies. In those experiments, an artificial crevice was constructed using two Alloy 625 washers and an Alloy 625 bolt. The assembly was placed in a solution of ASTM substitute ocean water and the current response to an applied potential was recorded. After an initial period characterized by decreasing current values with time (on the order of an hour), the current began to increase reaching a steady-state value after a few hours. Optical images from time-lapsed experiments revealed that crevice corrosion initiated at a location farthest away from the crevice mouth and progressed towards it during propagation. Further, it was concluded that crevice corrosion initiation in Alloy 625 was associated with transpassive dissolution.12  They also found that the onset of the steady-state current during propogation coincided with a cessation of crevice corrosion movement toward the mouth. That is, forward motion associated with the active front at early times stopped and additional propagation was associated with increased damage normal to the surface (increasing depth) at this location. The distance maintained between the active front and the crevice mouth was called xcrit and was found to be a function of applied potential with smaller distance associated with higher applied potentials.

In a series of subsequent publications,13-15  Lillard, et al., further investigated crevice corrosion initiation and propagation of Alloy 625 in ASTM substitute seawater using an acrylic crevice former and having a transparent window allowed for real-time crevice corrosion imaging, and determination of the active surface area and location of the active front at all times. These parameters allowed for calculating the current density and the IR corrected potential at the active front, i.e., the wall potential. The authors reported that as the active front propagates toward the crevice mouth, the crevice current density increases to values on the order of 0.1 A/cm2 at xcrit. Concerning the wall potential calculations, at early times the voltage dropped down the length of the crevice was small. Correspondingly, the potential at the tip of the crevice (farthest from the mouth) was close to the applied potential supporting the previous finding that initiation was associated with transpassive dissolution. In addition, a mechanism of crevice corrosion was proposed based on the generally accepted mechanism of pitting corrosion in a one-dimensional (1D) pit (pencil electrode experiment).16-23  It was proposed that crevice corrosion acted as a “1D pit in reverse.” Specifically, the limiting current density for crevice propagation was inversely proportional to distance from the mouth as it is in a 1D pit. Further, the value of xcrit marks the distance at which the crevice would transition from diffusion control to activation/IR control if it continued its forward motion.

Crevice Corrosion in Additively Manufactured Alloy 625

An alternative to traditional extrusion methods for producing Alloy 625 is AM.1,24-27  Its low footprint and the near net-shaped structures that result are of particular interest in locations where it may be difficult or untimely to obtain replacement parts. However, AM results in high residual stress and, minimally, thermal stress relief is required before a part can be used.

Cabrini, et al.,28  investigated crevice corrosion susceptibility of AM Alloy 625 produced by Laser Powder Bed Fusion (LPBF) in neutral 0.6 M NaCl (40°C) after heat treatment. By running potentiostatic tests, the authors concluded that the as-built AM Alloy 625 had higher crevice corrosion resistance than conventionally wrought Alloy 625. In addition, they found that heat treating AM Alloy 625 after what amounted to a short annealing, 980°C for 32 min followed by quenching, did not significantly affect crevice corrosion susceptibility. In another study, Cabrini, et al.,29  investigated corrosion of as-built AM Alloy 625, also via LPBF, in boiling ferric sulfate/sulfuric acid solution. The authors reported that AM Alloy 625 had a corrosion rate about half that of wrought Alloy 625. In addition, a short anneal at 980°C for 32 min further reduced the corrosion rate. Moreover, the authors confirmed their results by studying the damage profile stating that corrosion attacks were the most penetrating for wrought Alloy 625 and the least penetrating for heat-treated AM Alloy 625.

In comparison with the Cabrini studies, a study by Jung, et al.,30  showed that AM Alloy 625, produced by directed energy deposition (DED) heat treated at 1,200°C for 3 h (solutionization without quenching), had inferior corrosion properties as compared to wrought Alloy 625. The polarization curve, obtained in 3.5% NaCl solution, of AM Alloy 625 was less noble with a lower breakdown potential than that of the wrought. Moreover, the authors investigated the effect of heat treatment at 1,065°C, 1,130°C, and 1,200°C on the polarization curve of AM Alloy 625 though it is not clear how these temperatures (and times) were decided. They concluded that heat treatment helped in raising the breakdown potential with the optimum heat treatment temperature being 1,200°C.

Objective of this Investigation

The objective of this study is to determine the optimal processing parameters to minimize the susceptibility of LPBF AM Alloy 625 to crevice corrosion. The parameters studied include build direction and heat treatment (solution annealing, solution annealing plus stabilization annealing, and stress relieving). To accomplish this, data were collected from cyclic potentiodynamic polarization (CPP) curves in ASTM substitute seawater as well as current vs. time data from remote crevice assemblies (RCA) including metal-to-metal RCA and acrylic-to-metal RCA. The metal-metal crevice assembly data were analyzed using analysis of variance (ANOVA) to find the level of statistical significance of heat treatments and build orientation on corrosion susceptibility. In comparison, the acrylic-to-metal crevice assembly was used to calculate crevice corrosion current density (rate) and repassivation potential. The results from the two methods are contrasted.

Additively Manufacturing Specimens

AM 625 specimens were prepared using the LPBF method and a 3D Systems ProX DMP 320 machine. The certification for the powder used in this process was presented in Table 1. It is worth noting that the solute content of the AM 625 powder is close to that of wrought 625 except for the Fe content which is significantly lower in the AM 625 powder (0.40 mass%) as compared to the ASTM specification for the wrought material (5 mass% max., ASTM B44331 ). As shown in Figures 1(a) through (c), three types of specimens were produced: lollipop-shaped specimens, built at a 50° angle with respect to the bed, and washers built at both 50° and 90° angles. As a result, 90° built structures had identical surfaces on both sides while 50° built structures had a rough surface on one side (bed side) and a smoother surface on the other side (laser side). Following fabrication, specimens were heat treated and the parameters used in each heat treatment were summarized in Table 2: solution annealed (SA), solution plus stabilization annealed (SSA), and stress relieved (SR). Crevice corrosion data from these heat treatments were compared to the as-made/notreatment (NT) condition. For comparison with the heat treatments used here, ASM recommends a solution annealing temperature range of 1,095°C to 1,205°C while stress relieving can be performed below 650°C to avoid carbide precipitation.32 

Table 1.

Comparison Between Wrought 625 and AM 625 Compositions Measured by EDS Along with AM 625 Powder Certification

Comparison Between Wrought 625 and AM 625 Compositions Measured by EDS Along with AM 625 Powder Certification
Comparison Between Wrought 625 and AM 625 Compositions Measured by EDS Along with AM 625 Powder Certification
FIGURE 1.

Three specimen shapes were used: (a) 50°-built lollipops, (b) 50°-built washers, and (c) 90°-built washers.

FIGURE 1.

Three specimen shapes were used: (a) 50°-built lollipops, (b) 50°-built washers, and (c) 90°-built washers.

Close modal
Table 2.

The Parameters Used in Heat Treatment of AM Alloy 625

The Parameters Used in Heat Treatment of AM Alloy 625
The Parameters Used in Heat Treatment of AM Alloy 625

For comparison with AM specimens, conventionally wrought 625 specimens were also tested. This alloy was tested in the mill annealed condition.

To eliminate the effect of surface finish, all specimens were ground to 1200 grit SiC paper and degreased with acetone, ethanol, and deionized (DI) water before testing unless noted in the text.

Electrochemical Testing

Electrochemical experiments included CPP tests as well as metal-to-metal and acrylic-to-metal RCA potentiostatic and potentiodynamic tests. All tests in this study were performed at room temperature (approximately 21°C) in open-to-air ASTM D1141 substitute ocean water prepared using a commercially available salt that meets the standard (Lake Products).33  The solution was prepared by mixing the salt in the proper concentration using DI water and adjusting the pH of this solution to 8.2 using 0.1 M HCl and 0.1 M NaOH solutions. Experiments were conducted in a variety of three-electrode cells using a saturated calomel reference electrode (SCE), a platinized niobium mesh counter electrode, and a working electrode that depends on the type of experiment as explained below. Tests were performed using commercially available potentiostats and software.

CPP tests were performed using lollipop-shaped specimens (Figure 1[a]) immersed in solution such that the stick of the lollipop could be used for electrical contact. This configuration avoided the complication of specimen crevicing however, for the 50° build, it did not allow for tests to be conducted on the two distinct sides of the specimens separately. As such, the current density measurement is an average of the two sides. Experiments were begun by monitoring the open-circuit potential (OCP) for 1 h after immersion. After this period, scans were conducted from a cathodic potential of −0.1 VOCP to an apex potential of 1.5 VOCP or an apex current density of 0.01 A/cm2, whichever was achieved first. This was followed by reversing the scan direction and finishing at a final potential of 0.1 VOCP. The scan rate used in the CPP tests was 0.1667 mV/s.

Crevice tests using a metal-to-metal RCA setup were run for each heat treatment-orientation combination, for example, SA heat-treated specimen 50° built-laser side. The setup, described by Miller and Lillard,11  consisted of an Alloy 625 bolt, an Alloy 625 nut, a polytetrafluoroethylene (PTFE) washer, an cetal washer, and two Alloy 625 washers as illustrated in Figure 2(a). In this setup, two Alloy 625 washers of the same heat treatment and build orientation were placed on an Alloy 625 bolt to form the crevice, such that the walls of the crevice were formed by material of the same build orientation. The assembly of the RCA was performed while the parts were immersed in ASTM substitute ocean water. This ensured the crevice was fully wet before testing. To assemble the RCA, the Alloy 625 washers were placed on the bolt followed by the PTFE washer. Finally, an acetal washer was placed on top of the PTFE just below the nut to uniformly distribute the torque while the PTFE helps to prevent crevice corrosion from occurring in this region. The Alloy 625 nut bolt assembly was tightened using a torque wrench set to 15 ft/lb (20.34 N/m). This setup was used as the working electrode. After monitoring the OCP for 1 h, a potentiostatic hold at 0.3 VSCE (via ramping at a scan rate of 5 mV/s to 0.3 V) was performed for 60 h, and current vs. time data was recorded. A potential of 0.3 VSCE was chosen to place Alloy 625 in the passive region of the polarization curve.
FIGURE 2.

Two types of RCA were constructed from the Alloy 625 washers presented in Figure 1 to be used as the working electrode in electrochemical testing: (a) metal-to-metal RCA and (b) acrylic-to-metal RCA.

FIGURE 2.

Two types of RCA were constructed from the Alloy 625 washers presented in Figure 1 to be used as the working electrode in electrochemical testing: (a) metal-to-metal RCA and (b) acrylic-to-metal RCA.

Close modal

In the acrylic-to-metal RCA setup, shown in Figure 2(b), a single Alloy 625 washer was used along with a transparent acrylic washer to provide a window for real-time imaging.13  While this RCA was also assembled in solution, a lower torque of 10 ft/lb (13.56 N/m) was used to minimize the deformation of the acrylic. In addition to potentiostatic tests, repassivation was studied using the acrylic-to-metal RCA setup in a modified ASTM G19234  electrochemical test (THE method) similar to that used by Lillard and Mehrazi.15  The test consisted of 10 min at the OCP, followed by ramping up of the potential at a scan rate of 5 mV/s to 0.3 VSCE. The potential was then held at 0.3 VSCE for 20 ks followed by repassivation. Unlike Lillard and Mehrazi who approached repassivation by stepping down the applied potential by 25 mV every 45 min, repassivation in this study was achieved via a backward scan of 0.01667 mV/s to the OCP. During the experiment, the current vs. time data were recorded at 1 Hz. Correspondingly, images of crevice corrosion damage were captured using an Olympus SZX16 stereoscope, coupled with a camera, at a magnification of 0.7×. Though about 40% of the crevice fell within the fields of view, 25% of the total crevice area was analyzed and radial symmetry was assumed. In experiments where 10% or more of the damage was not symmetric, the results were discarded. Images were captured at a rate of 1 image/min, which falls within the optimal rate range shown by Lillard, et al.13 

Surface Characterization Techniques

After corrosion testing, the corroded surface of the test specimens was studied using several surface characterization techniques including inverted metallurgical optical microscopy (OM), optical profilometry, scanning electron microscopy (SEM), and energy dispersive x-ray spectrometry (EDS).

Cyclic Potentiodynamic Polarization

Before testing, the approximate composition of the specimens was measured using EDS. The results are compared with the AM Alloy 625 powder certification in Table 1. As seen in this table, the EDS results from the AM 625 material closely approximate the powder certification.

Typical CPP curves for the four AM specimens, and the wrought specimen, in ASTM substitute ocean water are shown in Figure 3. Four critical values were taken from the CPP curves and summarized in Table 3: the corrosion potential (Ecorr), corrosion current density (icorr), passive current density (ipass), and breakdown potential (Ebd). It is important to note that these values were obtained graphically where Ecorr and icorr were obtained from the intersection of the Tafel slopes, ipass was an average value in the region above Ecorr where there was little change in current density with increasing potential, and Ebd was obtained from the intersection of a straight line fitted to the passive region and another fitted to the transpassive region. The CPP tests revealed that, like wrought Alloy 625 in ASTM substitute ocean water, AM Alloy 625 specimens were spontaneously passive (lacking an active-to-passive transition) with negative hysteresis. Negative hysteresis occurs when the current density in the reverse scan is lower than that in the forward scan. This indicates that the passive film rapidly repairs itself and pitting does not initiate.35-36  As seen in Figure 3 and Table 3, heat treatment did not significantly affect the CPP results of AM Alloy 625.
FIGURE 3.

CPP curves for SA AM, SSA AM, SR AM, NT AM, and Wrought 625 in ASTM artificial seawater.

FIGURE 3.

CPP curves for SA AM, SSA AM, SR AM, NT AM, and Wrought 625 in ASTM artificial seawater.

Close modal
Table 3.

Data Obtained from the CPP Curves Shown in Figure 3

Data Obtained from the CPP Curves Shown in Figure 3
Data Obtained from the CPP Curves Shown in Figure 3
After the tests and the corresponding polarization into the transpassive region, the wrought and AM specimen surfaces were characterized using SEM/EDS, and the results are presented in Figures 4 and 5 for wrought and SA AM Alloy 625, respectively. While only the results for the SA AM condition are presented, they are typical of the other AM Alloy 625 specimens. As seen in these figures, the surface of the specimen after the CPP experiment was covered by a surface film. This film had a “mud-cracked” appearance and was rich in Mo, Nb, and O, presumably a mixed Mo-Nb oxide. In addition, while each specimen had a similar mud-cracked appearance there was some variation in the size of the cracks as shown in Figures 4 and 5 with the AM materials having larger cracks as compared to the wrought material. This mud-cracked surface film has also been observed in Alloy 625 crevices and is believed to be associated solely with transpassive dissolution and re-deposition of highly insoluble Mo and Nb oxides.
FIGURE 4.

SEM image and the corresponding EDS map for a wrought Alloy 625 lollipop specimen after the CPP test. Those images of corroded Wrought Alloy 625 indicate the presence of some layer of Mo, Nb, and O over the surface.

FIGURE 4.

SEM image and the corresponding EDS map for a wrought Alloy 625 lollipop specimen after the CPP test. Those images of corroded Wrought Alloy 625 indicate the presence of some layer of Mo, Nb, and O over the surface.

Close modal
FIGURE 5.

SEM image and the corresponding EDS map for an SA AM Alloy 625 lollipop specimen after the CPP test. The presence of the layer of Mo, Nb, and O over the surface appears more prominently.

FIGURE 5.

SEM image and the corresponding EDS map for an SA AM Alloy 625 lollipop specimen after the CPP test. The presence of the layer of Mo, Nb, and O over the surface appears more prominently.

Close modal

Statistical Comparison of Current vs. Time Data

Current vs. time data from RCA experiments are frequently used to assess crevice corrosion susceptibility, however, there is no generally agreed-upon parameter from these data that is an indicator of crevice corrosion susceptibility/resistance. As such, it was decided to statistically compare critical values from a plot of these data to determine if the data could be used to evaluate the effect of heat treatment and build orientation on crevice corrosion in AM 625.

Figure 6 shows the current vs. time curves for the four heat-treatment conditions for the crevice formed by the laser side of two 50° built AM Alloy 625 washers. In addition, the current vs. time curve for the crevice formed by wrought washers is presented on the same plot for reference. As has been previously shown,11  crevice corrosion begins with an incubation period marked by a drop in current with time, followed by a rapid increase in current marking the end of the incubation period and the onset of crevice corrosion. After initiation, the current increase associated with radial crevice corrosion propagation continues to occur before reaching a peak steady-state current value. Hence, to evaluate the effect of heat treatment and build orientation on crevice corrosion susceptibility, five critical values were chosen off the current vs. time curves: (1) initiation time, (2) initiation current, (3) time-to-peak current (including initiation time), (4) peak current magnitude, and (5) charge passed (shaded area under the curve) and each of these parameters are illustrated in Figure 7.
FIGURE 6.

Current vs. time data for AM Alloy 625 specimens built at a 50° angle, laser side. Wrought Alloy 625 data is shown for comparison.

FIGURE 6.

Current vs. time data for AM Alloy 625 specimens built at a 50° angle, laser side. Wrought Alloy 625 data is shown for comparison.

Close modal
FIGURE 7.

The four values taken from a current vs. time curve are: initiation time, initiation current, time-to-peak, peak current, and charge passed (shaded area under the curve).

FIGURE 7.

The four values taken from a current vs. time curve are: initiation time, initiation current, time-to-peak, peak current, and charge passed (shaded area under the curve).

Close modal

To determine the statistical significance of any trends in these parameters with heat treatment or build orientation, a commercially available statistical analysis software, Minitab, was used to perform an ANOVA of the data. The experimental design implemented was randomized complete block design (RCBD) where heat treatment (treatment factor) was randomly assigned to washers of each build orientation (Blocking factor). It is a “complete” design because data were collected for all heat treatment-build orientation combinations. Although this design allows for the evaluation of the effects of two factors, heat treatment and build orientation, it falls short of allowing the evaluation of the interactive effect between the two factors.37  Hence, the main assumption is negligible or no interaction between heat treatment and build orientation.

To evaluate the effects of heat treatment and build orientation on a quantifiable response, the data were fit to the statistical model:
formula
Here, Yij is a measured response for the heat treatment i and build orientation j combination, μ is the constant representing the overall response mean, τi is the deviation due to heat treatment i, βj is the deviation due to build orientation j, εij is the deviation due to random error for heat treatment i and build orientation j combination which is confounded with the two-factor interaction effect, and a and b are the number of heat treatments (four in this case) and build orientations (three in this case).

In the ANOVA analysis, p value was calculated to determine the probability of no correlation (null hypothesis). As such, a higher p value indicates a lower probability of a correlation:38-40 

  1. For p value > 0.1, there is not enough evidence a given factor affects response.

  2. For 0.05 < p value < 0.1, there is some evidence a given factor affects response.

  3. For 0.01 < p value < 0.05, there is strong evidence a given factor affects response.

  4. For p value < 0.01, there is very strong evidence a given treatment affects response.

For p value < 0.1, Tukey’s multiple comparison was used to find which heat treatment(s) and orientation(s) yield significantly different responses.41 

It is important to note that build orientation had a p value greater than 0.1 for all of the measured responses. Hence, there is not enough evidence that build orientation affects crevice corrosion susceptibility. Therefore, build orientation is ignored in the presented analysis unless noted otherwise. It is worth mentioning that Sander, et al.,42  had a similar finding regarding the effect of build orientation on the susceptibility of as-built AM 316L (UNS S31603(1)) to pitting corrosion. By investigating specimens built at a 0°, 45°, and 90° angle, the authors concluded that pitting corrosion performance did not differ significantly among the three build orientations.42 

A standard assumption to perform the ANOVA is the homogeneity of variance. Hence, before performing the ANOVA, Levene’s test was used to check if the groups have the same variances and the Levene’s p value for the effect of heat treatment on initiation time, initiation current, time-to-peak, and the peak current was 0.677, 0.120, 0.592, and 0.293, respectively. With the p values all above a significance level of 0.1, the homogeneity of variance was not violated, and performing the ANOVA was justified.

Time to Initiation and Inititation Current

Crevice corrosion initiation time was plotted against heat treatment for each of the three different build orientations of the AM alloy as shown in Figure 8. The initiation time for the wrought alloy is also shown on the same plot (purple dot) for reference. The mean initiation time for each heat treatment is also presented in Figure 8 (black plot). The mean initiation time for each heat treatment arranged from lowest to highest follows the order SSA < SA < NT < SR. The p value for heat treatment was found to be 0.041 indicating strong evidence heat treatment affects initiation time. Also presented is Tukey’s Multiple Comparison Analysis (significance level, α = 0.05) where heat treatments that do not share a letter in the “grouping” in Figure 8 are significantly different. Thus, SSA has a significantly lower initiation time than SR. That is, the onset of crevice corrosion is significantly delayed by the SSA thermal processing.
FIGURE 8.

Initiation time data for all heat treatment-build orientation combinations. The mean initiation time is shown for all heat treatments (black plot). According to Tukey’s Multiple Comparison at significance level α = 0.05, means fell into two groups: group A (shown in blue) and group B (shown in yellow). The groups overlap in the green area which means lying in different groups are significantly different. Initiation time for wrought 625 (purple point) is shown for reference.

FIGURE 8.

Initiation time data for all heat treatment-build orientation combinations. The mean initiation time is shown for all heat treatments (black plot). According to Tukey’s Multiple Comparison at significance level α = 0.05, means fell into two groups: group A (shown in blue) and group B (shown in yellow). The groups overlap in the green area which means lying in different groups are significantly different. Initiation time for wrought 625 (purple point) is shown for reference.

Close modal

Crevice initiation current data for each heat treatment-orientation combination was also analyzed using the ANOVA. However, the p value for heat treatment was found to be 0.112 (>0.1) indicating a lack of sufficient evidence heat treatment affects initiation current.

Time to Peak and Peak Current

Crevice corrosion propagation was quantified using time-to-peak (time to steady-state propagation) and peak current. The time-to-peak and peak current data were analyzed for all heat treatment-orientation combinations using the ANOVA. However, the p value of the heat treatment effect on time-to-peak and peak current were found to be 0.516 and 0.237, respectively. As both p values were greater than 0.1, it was concluded that there is a lack of sufficient evidence heat treatment affects either response.

Total Corrosion Damage: Charge Passed

The charge passed is directly proportional to mass lost due to corrosion via Faraday’s law. Hence, to compare total corrosion damage for each heat treatment and orientation combination, the charge passed due to crevice corrosion was calculated from the current vs. time curves. The calculation was performed by finding the area under the current vs. time curve from the onset of crevice corrosion (initiation time) to the end of the test as illustrated in Figure 7 by the yellow-shaded region. The charge passed results are plotted against heat treatment for each of the three different build orientations in Figure 9. Again, the wrought condition is shown for reference. In addition, the mean charge passed was also plotted. As seen in this figure, the charge passed for NT was greater than the heat-treated alloys SR, SSA, and SA with SA being a factor of four less than NT. Having a Levene’s p-value of 0.504, performing the ANOVA was possible, resulting in a p-value for heat treatment of 0.08. Thus, there is evidence that charge passed is affected by heat treatment. Hence, Tukey’s Multiple Comparison Analysis was performed at the significance level, α = 0.1. As before, heat treatments that do not fall in the same group, as visually illustrated by the letters in Figure 9, are significantly different. In this case, the lack of thermal processing after manufacturing (NT) is statistically more likely to sustain crevice corrosion damage as compared with AM 625 which has been solutionized (SA). Interestingly, as seen in Figure 9, the charge passed for wrought 625 was 85.0 C (coulombs), making it the second most susceptible after NT.
FIGURE 9.

Charge passed data for all heat treatment-build orientation combinations. The mean charge passed is shown for all heat treatments (black plot). According to Tukey’s Multiple Comparison at significance level α = 0.1, means fell into two groups: group A (shown in blue) and group B (shown in yellow). The groups overlap in the green area and means lying in different groups are significantly different. Charge passed for wrought 625 (purple point) is shown for reference.

FIGURE 9.

Charge passed data for all heat treatment-build orientation combinations. The mean charge passed is shown for all heat treatments (black plot). According to Tukey’s Multiple Comparison at significance level α = 0.1, means fell into two groups: group A (shown in blue) and group B (shown in yellow). The groups overlap in the green area and means lying in different groups are significantly different. Charge passed for wrought 625 (purple point) is shown for reference.

Close modal
To benchmark the charge passed calculations, surface profile images were taken for each heat treatment-orientation combination. Figures 10 through 12 show optical images of each pair of AM 625 washers after a metal-to-metal RCA test and a line scan showing the damage profile. Looking at the optical images, it is worth mentioning that the damaged area on each top washer was a mirror image of that on its corresponding bottom washer. However, for consistency, line scans were only obtained for the bottom washers. In addition, for a given washer, multiple line scan measurements were taken at different locations to ensure the presented line scans were representative. Moreover, for the presented damage profiles, baseline calculations were done using commercially available software, Origin, with the flat uncorroded surface taken as depth zero.
FIGURE 10.

Optical images of AM 625 washers, built at a 90° angle, after undergoing a metal-to-metal RCA test and line scans showing the damage profile. The maximum penetration depth recorded is indicated on each plot.

FIGURE 10.

Optical images of AM 625 washers, built at a 90° angle, after undergoing a metal-to-metal RCA test and line scans showing the damage profile. The maximum penetration depth recorded is indicated on each plot.

Close modal
FIGURE 11.

Optical images of AM 625 washers, built at a 50° angle (laser side), after undergoing a metal-to-metal RCA test and line scans showing the damage profile. The maximum penetration depth recorded is indicated on each plot.

FIGURE 11.

Optical images of AM 625 washers, built at a 50° angle (laser side), after undergoing a metal-to-metal RCA test and line scans showing the damage profile. The maximum penetration depth recorded is indicated on each plot.

Close modal
FIGURE 12.

Optical images of AM 625 washers, built at a 50° angle (powder bed side), after undergoing a metal-to-metal RCA test and line scans showing the damage profile. The maximum penetration depth recorded is indicated on each plot.

FIGURE 12.

Optical images of AM 625 washers, built at a 50° angle (powder bed side), after undergoing a metal-to-metal RCA test and line scans showing the damage profile. The maximum penetration depth recorded is indicated on each plot.

Close modal
The maximum penetration depth for each heat treatment-orientation combination is shown in Figure 13, along with that for the wrought alloy (purple dot) for reference. In addition, the mean maximum penetration depth for each heat treatment (black plot) is plotted on the same graph. With a Levene’s p-value of 0.265, the ANOVA was performed resulting in a p-value of 0.001 indicating very strong evidence heat treatment affects maximum penetration depth. Hence, Tukey’s Multiple Comparison Analysis was possible at a significance level, α = 0.01. As before, the grouping is visually illustrated in Figure 13, and heat treatments that do not lie in the same group yield significantly different responses. Consistent with the results from charge passed calculations, the lack of thermal processing after manufacturing (NT) is statistically more likely to sustain crevice corrosion damage as compared with AM 625 which has been solutionized (SA).
FIGURE 13.

Maximum penetration recorded data for all heat treatment-build orientation combinations. The mean maximum penetration recorded is shown for all heat treatments (black plot). According to Tukey’s Multiple Comparison at significance level α = 0.01, means fell into two groups: group A (shown in blue) and group B (shown in yellow). Means lying in different groups are significantly different. The maximum penetration recorded for wrought 625 (purple point) is shown for reference.

FIGURE 13.

Maximum penetration recorded data for all heat treatment-build orientation combinations. The mean maximum penetration recorded is shown for all heat treatments (black plot). According to Tukey’s Multiple Comparison at significance level α = 0.01, means fell into two groups: group A (shown in blue) and group B (shown in yellow). Means lying in different groups are significantly different. The maximum penetration recorded for wrought 625 (purple point) is shown for reference.

Close modal

In comparison to these results, the Cabrini, et al.,29  study of AM Alloy 625 corrosion in boiling ferric sulfate/sulfuric acid solution, also concluded that heat-treated AM specimens had smaller mass loss and shallower penetration depths than those that were untreated. While the results here, for seawater, are consistent with those of Cabrini, et al., contrary to the Cabrini study, in this study, the performance of the untreated AM Alloy 625 was not remarkably different than that of the wrought Alloy 625.

Comparison of Crevice Current Density and Repassivation

The current vs. time data analyzed thus far is limited as, in and of itself, it does not provide corrosion rate data. This owes to the inability to calculate the active area and its location from the crevice mouth at any given time using the metal-to-metal RCA setup. By replacing one of the metal washers with an acrylic washer and imaging the active crevice, the corrosion area and its location could be determined as functions of time, enabling current density and IR drop calculations. Knowing the IR drop and the location of the active front, it is also possible to calculate the crevice repassivation potential. Those data are presented in this section and the statistical analysis is omitted.

Current Density

The first step in determining the surface current density at the active front is to correct the experimental current (imeas) for the passive current that emanates from the boldly exposed portion of the RCA (ipass). This is done by fitting a curve to the measured current at early times and extrapolating it to the theoretical passive current for Alloy 625 in ASTM artificial ocean water calculated from CPP curves. Having corrected the experimental current for passive current, the second step in determining current density is to determine the active area. Here, real-time imaging of the surface during propagation and a MatLab image processing routine are used to calculate the active area (ΔA), the number of “new pixels” in consecutive images being due to active corrosion. The surface current density at the active front is then calculated, (imeas-ipass)/ΔA, as a function of distance traveled. Further information on the calculations and assumptions is available in the publication by Lillard, et al.13 

Figure 14 shows the current, active corrosion area, current density, and applied potential on the same plot against time for the SA AM Alloy 625. On this plot, the four vertical dashed lines correspond to: (1) 2,400 s: the point of crevice corrosion initiation, (2) 20,100 s: beginning of the potential ramp down, (3) 28,980 s: the point at which current density starts to decrease with time (the intersection of the two orange straight lines in Figure 14), and (4) 38,880 s: repassivation, defined as the point at which current changes polarity. As seen in Figure 14, the potential ramp-down is associated with a decrease in both the total current and active corrosion area, however, current density maintains its increasing trend for an extended period. This occurs because the active corrosion area decreases at a rate that is higher than the decreasing rate of the current.
FIGURE 14.

Current, Δ corrosion area, and current density for SA AM Alloy 625 repassivation test. Four critical points are indicated: initiation, ramp down, transition, and repassivation.

FIGURE 14.

Current, Δ corrosion area, and current density for SA AM Alloy 625 repassivation test. Four critical points are indicated: initiation, ramp down, transition, and repassivation.

Close modal
Figures 15(a) and (b) show current density as a function of time and position, respectively, for the four different AM conditions as well as that for the wrought alloy. As seen in these plots, the current density increases as the active front moves toward the crevice mouth until it reaches a critical distance from the crevice mouth, xcrit as observed elsewhere.13,15  While there is no apparent trend in current density with heat treatment in these figures, this does not mean that there is no relationship between crevice corrosion damage and heat treatment as the surface damage varies with current density and the velocity of the active front. It can be shown that damage depth, z, is equal
formula
where n is the number of equivalents of Alloy 625 and equal to 2.12 equiv., F is Faraday’s constant (96,485 C/equiv.), ρ is alloy 625 density (8.44 g/cm3), M is the molecular mass of Alloy 625 given by 59.90 g/mol, i(x,t) is current density at location, x, and time, t, and t1 and t2 represent the initial and final times.
FIGURE 15.

(a) Current density vs. time; and (b) current density vs. distance from the crevice mouth.

FIGURE 15.

(a) Current density vs. time; and (b) current density vs. distance from the crevice mouth.

Close modal
The corresponding calculated penetration depth from Equation (2) as a function of location from the crevice mouth is shown in Figure 16 for the AM specimens and the wrought specimen. As seen in this figure, the current density calculations predict that NT AM should have the largest maximum penetration depth followed by SR AM, SSA AM, and SA AM. The wrought specimen has a maximum penetration depth similar to that of SR AM. To benchmark these predictions, surface profile results taken at the end of the electrochemical test are shown in Figure 17. As seen in this figure, the locations and depths of the calculated damage agree well with the predicted values, that is, the current density/velocity measurements appear to be a relatively accurate method for calculating crevice corrosion damage. As it relates to the performance of the AM alloys, it is clear that the as-built alloy, NT, had greater penetration depths as compared to the wrought material. In addition, while stress relief, SR, improves AM 625 performance, a solution anneal, SA, can improve the crevice corrosion performance beyond that of the wrought material. From an engineering perspective, this is an important finding as no additional/specially developed processing for the AM material beyond the established heat treatments for Alloy 625 is necessary to mitigate crevice corrosion in AM Alloy 625.
FIGURE 16.

Calculated depth as a function of distance from the crevice mouth obtained using Equation (2). The maximum penetration depth is indicated.

FIGURE 16.

Calculated depth as a function of distance from the crevice mouth obtained using Equation (2). The maximum penetration depth is indicated.

Close modal
FIGURE 17.

Optical images showing crevice corrosion damage at the end of the acrylic-to-metal RCA test and the corresponding surface profile. The maximum penetration depth is indicated.

FIGURE 17.

Optical images showing crevice corrosion damage at the end of the acrylic-to-metal RCA test and the corresponding surface profile. The maximum penetration depth is indicated.

Close modal

Crevice Repassivation

To determine the crevice repassivation potential, the potential of the wall as a function of distance is needed. Wall potential calculations are possible from the crevice current, with knowledge of the crevice resistance, Rcrev. The crevice resistance is a function of the location of the active front and a relationship for it was derived by Lillard, et al.13  Figure 18 shows the applied potential, potential drop, and wall potential throughout the repassivation test for an SA AM specimen. The previously mentioned four critical points of initiation, ramp down, transition, and repassivation are represented by four dashed lines on the plot. It is important to note that at initiation and repassivation, the potential drop (IR) is negligible, and wall potential approaches the applied potential. This is due to the low current values recorded at these two points. However, between initiation and ramp down, potential drop increases with time causing the wall potential to deviate from the applied potential reaching a maximum at ramp down. Once the potential ramp down is initiated, potential drop decreases with time and wall potential starts to approach the applied potential once again. It is worth mentioning that, like current density, wall potential calculations were not possible before initiation or after repassivation because no active corrosion area was detected.
FIGURE 18.

Applied potential, IR-potential drop, and wall potential for SA AM Alloy 625 repassivation test. Four critical points are indicated: initiation, ramp down, transition, and repassivation.

FIGURE 18.

Applied potential, IR-potential drop, and wall potential for SA AM Alloy 625 repassivation test. Four critical points are indicated: initiation, ramp down, transition, and repassivation.

Close modal

Table 4 presents the wall potentials at transition (ET) and repassivation (ERp) and their corresponding times for the AM specimens as well as the wrought specimen. All of the transition potentials lay between 0.107 VSCE and 0.128 VSCE while all of the repassivation potentials lay between −0.014 VSCE and −0.041 VSCE. Hence, transition potential and repassivation potentials for the studied specimens lay within a small range of 0.021 V and 0.027 V, respectively. While there is no clear trend with heat treatment, it is noted that SA had the most positive repassivation potential while the wrought material had the most negative transition potential. More negative repassivation and transition potentials are associated with more stable crevice corrosion.

Table 4.

Transition Potential and Repassivation Potential and Corresponding Time

Transition Potential and Repassivation Potential and Corresponding Time
Transition Potential and Repassivation Potential and Corresponding Time

Results presented here show clear trends in crevice corrosion susceptibility with heat treatment. Specifically, statistical analysis of current vs. time data found that heat treatment had a strong effect on time to initiation, as was shown in Figure 8. In addition, there was also a statistical correlation between the mean charge passed at a fixed experimental time for each heat treatment, from largest to smallest: NT>SR>SSA>SA. This finding was consistent with measured penetration depth where NT AM specimens had the largest maximum penetration depth followed by SR AM, SSA AM, and SA AM. Analysis of crevice corrosion kinetics found similar trends, with NT and SR specimens being associated with large damage depths while SA and SSA specimens had relatively little damage during equal propagation times. Possible rationales for these relationships with heat treatment include the formation of second-phase particles, differences in residual stress owing to high cooling rates (1 × 106°C/s), and differences in microstructure.43 

Concerning differences in microstructure, optical images as a function of heat treatment taken after crevice experiments clearly showed the damage was microstructure-specific as seen in Figure 19. For example, corrosion damage in NT and SR AM specimens revealed the melt pools formed during fabrication.43-44  It is also clear from these images that the temperature associated with stress relieving (and used in other studies) is insufficient to homogenize the microstructure.45  To further pursue this, an SEM investigation of the microstructure was undertaken.
FIGURE 19.

Optical images taken at the end of acrylic-to-metal tests. Melt pools appear clearly on the corroded surface of stress relieved and As Made AM Alloy 625 as indicated with the red arrows.

FIGURE 19.

Optical images taken at the end of acrylic-to-metal tests. Melt pools appear clearly on the corroded surface of stress relieved and As Made AM Alloy 625 as indicated with the red arrows.

Close modal
Figure 20(a) is an SEM micrograph of the crevice corrosion damage in the as-built material (NT AM) after exposure. Crevice corrosion damage revealed the melt pools associated with AM LPBF fabrication. The microstructure in these melt pools consists of “grains” characterized by columnar dendrites with the same growth direction. This dendritic structure is associated with the exceedingly high cooling rates inherent in AM LPBF fabrication.46-47  They can appear as long columns (Figure 20[c]) or cells (Figure 20[d]) depending on the cross section and growth direction. The interdendritic region (cell walls) is rich in Nb and Mo.43-46  Correspondingly, the cell interior is depleted in Mo and Nb and enriched in Ni and Cr.43  It is this Mo and Nb depleted region that appears to be the most susceptible to the oxidizing conditions within the crevice solution leaving behind the Mo and Nb rich interdedritic regions.
FIGURE 20.

SEM images of the corroded NT AM specimen showing (a) melt pools highlighted in dashed lines and columnar grains running parallel to the build direction. (b) Higher magnification SEM images showing dendritic (1) and cellular (2) structures. (c) A magnified view of the dendritic structures and (d) the cellular structures.

FIGURE 20.

SEM images of the corroded NT AM specimen showing (a) melt pools highlighted in dashed lines and columnar grains running parallel to the build direction. (b) Higher magnification SEM images showing dendritic (1) and cellular (2) structures. (c) A magnified view of the dendritic structures and (d) the cellular structures.

Close modal
Figure 21 shows the SEM images for an SR AM specimen after exposure in a remote crevice experiment. As can be seen in this figure, the stress relief at the given conditions results in a microstructure that is similar to the as-built, with melt pools (Figure 21[a]) and remnants of the interdendritic regions (Figure 21[b]) still present. It is proposed that this microstructure accounts for the lower crevice corrosion resistance of the SR AM specimens as compared to the solution-annealed specimens. Correspondingly, Figure 22 shows the SEM images for an SSA AM specimen. Like the SA AM and wrought specimens (not shown), the SSA alloy exhibits a homogenized microstructure with no melt pools or dendritic structures, as shown in Figure 22(a). Additionally, the microstructure exhibited submicrometric carbides scattered all over the microstructure, as indicated by arrows in Figure 22(b) consistent with the literature.1,46  One might argue that the similarity between the SA, SSA, and wrought microstructures is evidence that microstructure plays no role in crevice corrosion susceptibility (given the relatively poor performance of the wrought specimens).
FIGURE 21.

SEM images of the corroded SR AM specimen showing (a) melt pools highlighted in dashed lines and columnar grains running parallel to the build direction. (b) Higher magnification SEM images showing remnants of the dendritic structures with the interdendritic regions adorned with precipitants of platelet shaped δ phase (Ni3Nb), and spherical-shaped Nb-rich metal carbides (MC). In addition, elongated Cr-rich M23C6 metal carbides have been spotted on the grain boundaries.

FIGURE 21.

SEM images of the corroded SR AM specimen showing (a) melt pools highlighted in dashed lines and columnar grains running parallel to the build direction. (b) Higher magnification SEM images showing remnants of the dendritic structures with the interdendritic regions adorned with precipitants of platelet shaped δ phase (Ni3Nb), and spherical-shaped Nb-rich metal carbides (MC). In addition, elongated Cr-rich M23C6 metal carbides have been spotted on the grain boundaries.

Close modal
FIGURE 22.

SEM images of the corroded SSA AM specimen showing (a) a homogenized microstructure. (b) Higher magnification SEM images showing submicrometric carbides.

FIGURE 22.

SEM images of the corroded SSA AM specimen showing (a) a homogenized microstructure. (b) Higher magnification SEM images showing submicrometric carbides.

Close modal

It is also believed that the low Fe content in the AM powder and correspondingly the alloy is responsible for improved crevice corrosion performance of the SA and SSA AM alloy as compared to the wrought alloy. Similarly, the Cabrini, et al.,29  specimens, which also had an improved crevice corrosion resistance as compared to wrought, also had a low Fe content (0.45 mass%). Thus, even though they did not fully solutionize their specimens, and columnar grains were still present in their annealed specimens, the lower Fe content likely also contributed to the improved crevice corrosion performance they observed. This also explains the findings of the Jung, et al., study which investigated the directed energy deposition specimens heat treated at 1,200°C for 3 h (solutionization).30  They concluded, “Corrosion properties of the 3D printed alloys are generally inferior to the casting alloys, and this is due to the microstructure.” However, the powder material used for fabrication had an Fe content of 3.65 mass%. In comparison, the resulting AM 625 specimens had an Fe content of 3.59 mass% whereas their wrought had an Fe content of 2.75 mass%.

Crevice corrosion of AM Alloy 625 produced by LPBF was investigated in ASTM artificial seawater and the conclusions are summarized below:

  • It was concluded that crevice corrosion propagation during potentiostatic holds of AM Alloy 625 is consistent with transpassive dissolution as evidenced by the corrosion deposit found inside the crevice which had the same “mud-cracked” appearance and chemical composition as that found in noncreviced specimens that were potentiodynamically polarized above the transpassive potential.

  • Statistical analysis of current vs. time data found that there was strong evidence heat treatment affects crevice corrosion susceptibility. For example, nonheat-treated AM specimens had the largest charge passed during crevice propagation, in comparison with solutionized specimens, agreeing with the maximum penetration depth results from profilometry. Alternately, there was not sufficient statistical evidence that build orientation affects crevice corrosion susceptibility.

  • Predicted depth profiles from kinetic studies (current density) were in good agreement with profilometry data and provided further evidence to support the conclusion that solutionization reduces crevice corrosion susceptibility of AM 625 material. In that work, it was shown that solutionized specimens had less damage (depth and radial propagation) as compared to stress-relieved or nonheat-treated specimens. However, as it relates to the critical potentials calculated from these studies, it was concluded that there is little to no effect of heat treatment on either the transition or repassivation potentials.

  • While stress relief improves AM 625 performance, a solution anneal can improve the crevice corrosion performance beyond that of the wrought material. From an engineering perspective, this is an important finding as no additional/specially developed processing for the AM material beyond the established heat treatments for Alloy 625 are necessary to mitigate crevice corrosion in AM Alloy 625.

  • It was concluded that microstructure was responsible for crevice corrosion susceptibility in heat-treated AM specimens. Specifically, insufficient or no heat treatment resulted in a nonhomogenized microstructure with clearly visible melt pools and dendritic structures that were susceptible to crevice attack. In comparison, fully solutionized specimens resulted in homogenized microstructures that were more resistant to crevice attacks.

  • It was concluded that the increased crevice corrosion susceptibility of wrought material, in comparison with the solutionized AM 625, was owed to the significantly higher Fe content in the wrought alloy.

(1)

UNS numbers are listed in Metals & Alloys in the Unified Numbering System, published by the Society of Automotive Engineers (SAE International) and cosponsored by ASTM International.

Trade name.

The authors gratefully acknowledge funding from America Makes Advanced Tools for Rapid Qualification (ATRQ) cooperative agreement FA8650-16-5700 sub-recipient agreement 20190054 and project managers Rick Fowler and Jason Thomas at the National Center for Defense Manufacturing & Machining (NCDMM). We would also like to thank our other team members for their valuable input throughout the project: David Ervin (Northrup Grumman), John Rails and Kyle Wade (HII Newport News Shipbuilding) and Ryan Overdorff (3D Systems). Approved for public release: Distribution is unlimited, AFRL-2023-6026, 28 Nov 2023.

1.
Floreen
S.
,
Fuchs
G.E.
,
Yang
W.J.
, “
The Metallurgy of Alloy 625
,”
Superalloys 718, 625, 706 and Various Derivatives
(
Pittsburgh, PA
:
The Minerals, Metals & Materials Society
,
1994
),
p
.
13
37
.
2.
Eiselstein
H.L.
,
Tillack
D.J.
, “
The Invention and Definition of Alloy 625
,”
Superalloys 718, 625 and Various Derivatives
(
Pittsburgh, PA
:
The Minerals, Metals & Materials Society
,
1991
),
p
.
1
14
.
3.
Lutton Cwalina
K.
,
Demarest
C.R.
,
Gerard
A.Y.
,
Scully
J.R.
,
Curr. Opin. Solid State Mater. Sci.
23
,
3
(
2019
):
p
.
129
141
.
4.
Wanklyn
J.N.
,
Corros. Sci.
21
,
3
(
1981
):
p
.
211
225
.
5.
Davies-Smith
L.R.
,
Lane
J.D.
,
Riley
T.
,
Br. Corros. J.
22
,
2
(
1987
):
p
.
90
94
.
6.
Hamdy
A.S.
,
El-Shenawy
E.
,
El-Bitar
T.
,
Int. J. Electrochem. Sci.
1
,
4
(
2006
):
p
.
171
180
.
7.
Mohammadi Zahrani
E.
,
Alfantazi
A.M.
,
Corros. Sci.
65
(
2012
):
p
.
340
359
.
8.
Shaw
B.A.
,
Moran
P.J.
,
Gartland
P.O.
,
Corros. Sci.
32
,
7
(
1991
):
p
.
707
719
.
9.
Martin
F.J.
,
Natishan
P.M.
,
Lucas
K.E.
,
Hogan
E.A.
,
Grolleau
A.M.
,
Thomas
E.D.
,
Corrosion
59
,
6
(
2003
):
p
.
498
504
.
10.
Lillard
R.S.
,
Jurinski
M.P.
,
Scully
J.R.
,
Corrosion
50
,
4
(
1994
):
p
.
251
265
.
11.
Miller
D.M.
,
Lillard
R.S.
,
J. Electrochem. Soc.
166
,
11
(
2019
):
p
.
C3431
C3442
.
12.
Salgado
D.M.
,
Lillard
R.S.
,
J. Electrochem. Soc.
164
,
13
(
2017
):
p
.
C801
C809
.
13.
Lillard
R.S.
,
Mehrazi
S.
,
Miller
D.M.
,
J. Electrochem. Soc.
167
,
2
(
2020
):
p
.
021511
.
14.
Lillard
R.S.
,
Miller
D.M.
,
Clemons
C.B.
,
Kreider
K.L.
,
Young
G.W.
,
J. Electrochem. Soc.
167
,
14
(
2020
):
p
.
141503
.
15.
Lillard
R.S.
,
Mehrazi
S.
,
J. Electrochem. Soc.
168
,
2
(
2021
):
p
.
021511
.
16.
Li
T.
,
Scully
J.R.
,
Frankel
G.S.
,
J. Electrochem. Soc.
166
,
11
(
2019
):
p
.
C3341
C3354
.
17.
Li
T.
,
Wu
J.
,
Frankel
G.S.
,
Corros. Sci.
182
(
2021
):
p
.
109277
.
18.
Srinivasan
J.
,
Kelly
R.G.
,
J. Electrochem. Soc.
163
,
13
(
2016
):
p
.
C759
C767
.
19.
Srinivasan
J.
,
Kelly
R.G.
,
J. Electrochem. Soc.
163
,
13
(
2016
):
p
.
C768
C777
.
20.
Gaudet
G.T.
,
Mo
W.
,
Hatton
T.
,
Tester
J.
,
Tilly
J.
,
Isaacs
H.
,
Newman
R.
,
AIChE J.
32
,
6
(
1986
):
p
.
949
958
.
21.
Beck
T.R.
,
J. Electrochem. Soc.
120
,
10
(
1973
):
p
.
1317
.
22.
Tester
J.W.
,
Isaacs
H.S.
,
J. Electrochem. Soc.
122
,
11
(
1975
):
p
.
1438
1445
.
23.
Newman
R.C.
,
Isaacs
H.S.
,
J. Electrochem. Soc.
130
,
7
(
1983
):
p
.
1621
1624
.
24.
Gola
K.
,
Dubiel
B.
,
Kalemba-Rec
I.
,
J. Mater. Eng. Perform.
29
,
3
(
2020
):
p
.
1528
1534
.
25.
Carrozza
A.
,
Lorenzi
S.
,
Carugo
F.
,
Fest-Santini
S.
,
Santini
M.
,
Marchese
G.
,
Barbieri
G.
,
Cognini
F.
,
Cabrini
M.
,
Pastore
T.
,
Mater. Des.
225
(
2023
):
p
.
111545
.
26.
DebRoy
T.
,
Wei
H.
,
Zuback
J.
,
Mukherjee
T.
,
Elmer
J.
,
Milewski
J.
,
Beese
A.
,
Wilson-Heid
A.
,
De
A.
,
Zhang
W.
,
Prog. Mater. Sci.
92
(
2018
):
p
.
112
224
.
27.
Criales
L.E.
,
Arısoy
Y.M.
,
Lane
B.
,
Moylan
S.
,
Donmez
A.
,
Özel
T.
,
Int. J. Mach. Tools Manufac.
121
(
2017
):
p
.
22
36
.
28.
Cabrini
M.
,
Lorenzi
S.
,
Testa
C.
,
Pastore
T.
,
Brevi
F.
,
Biamino
S.
,
Fino
P.
,
Manfredi
D.
,
Marchese
G.
,
Calignano
F.
,
Scenini
F.
,
J. Electrochem. Soc.
166
,
11
(
2019
):
p
.
C3399
C3408
.
29.
Cabrini
M.
,
Lorenzi
S.
,
Testa
C.
,
Brevi
F.
,
Biamino
S.
,
Fino
P.
,
Manfredi
D.
,
Marchese
G.
,
Calignano
F.
,
Pastore
T.
,
Materials
12
,
11
(
2019
):
p
.
1742
.
30.
Jung
G.-S.
,
Park
Y.-H.
,
Kim
D.-J.
,
Lim
C.-S.
,
Corros. Sci. Technol.
18
,
6
(
2019
):
p
.
258
266
.
31.
ASTM Standard B443-19
, “
Standard Specification for Nickel-Chromium-Molybdenum-Columbium Alloy and Nickel-Chromium-Molybdenum-Silicon Alloy Plate, Sheet, and Strip
” (
West Conshohocken, PA
:
ASTM International
,
2019
).
32.
Heat Treatment, ASM International Handbook Committee
,”
ASM handbook
,
vol.
4
(
Materials Park
:
ASM International, Ohio
,
1991
).
33.
ASTM Standard D1141
, “
Standard Practice for the Preparation of Substitute Ocean Water
” (
West Conshohocken, PA
:
ASTM International
,
2013
).
34.
ASTM Standard G192-08(2020)e1
, “
Standard Test Method for Determining the Crevice Repassivation Potential of Corrosion-Resistant Alloys Using a Potentiodynamic-Galvanostatic-Potentiostatic Technique
” (
West Conshohocken, PA
:
ASTM International
,
2020
).
35.
Esmailzadeh
S.
,
Aliofkhazraei
M.
,
Sarlak
H.
,
Prot. Met. Phys. Chem. Surf.
54
,
5
(
2018
):
p
.
976
989
.
36.
Lameche
S.
,
Nedjar
R.
,
Rebbah
H.
,
Adjeb
A.
,
Asian J. Chem.
20
,
4
(
2008
):
p
.
2545
2550
.
37.
Casler
M.D.
,
Agron. J.
107
,
2
(
2015
):
p
.
692
705
.
38.
Ganesh
S.
,
Cave
V.
,
N Z Vet. J.
66
,
2
(
2018
):
p
.
55
56
.
39.
Goodman
S.
,
Semin Hematol
45
,
3
(
2008
):
p
.
135
140
.
40.
Dahiru
T.
,
Ann. Ib Postgrad. Med.
6
,
1
(
2011
),
41.
Abdi
H.
,
Williams
L.
,
Encycl. Res. Des.
3
,
10
(
2010
):
p
.
1
5
.
42.
Sander
G.
,
Babu
A.P.
,
Gao
X.
,
Jiang
D.
,
Birbilis
N.
,
Corros. Sci.
179
(
2021
):
p
.
109149
.
43.
Lass
E.A.
,
Stoudt
M.
,
Williams
M.
,
Katz
M.
,
Levine
L.
,
Phan
T.
,
Gnaeupel-Herold
T.
,
Ng
D.
,
Metall. Mater. Trans. A
48
,
11
(
2017
):
p
.
5547
5558
.
44.
Zhang
F.
,
Ilavsky
J.
,
Lindwall
G.
,
Stoudt
M.R.
,
Levine
L.E.
,
Allen
A.J.
,
Appl. Sci.
11
,
18
(
2021
):
p
.
8643
.
45.
Marchese
G.
,
Parizia
S.
,
Rashidi
M.
,
Saboori
A.
,
Manfredi
D.
,
Ugues
D.
,
Lombardi
M.
,
Hryha
E.
,
Biamino
S.
,
Mater. Sci. Eng. A
769
(
2020
):
p
.
138500
.
46.
Marchese
G.
,
Lorusso
M.
,
Parizia
S.
,
Bassini
E.
,
Lee
J.
,
Calignano
F.
,
Manfredi
D.
,
Terner
M.
,
Hong
H.
,
Ugues
D.
,
Lombardi
M.
,
Biamino
S.
,
Mater. Sci. Eng. A
729
(
2018
):
p
.
64
75
.
47.
Chlebus
E.
,
Gruber
K.
,
Kuźnicka
B.
,
Kurzac
J.
,
Kurzynowski
T.
,
Mater. Sci. Eng. A
639
(
2015
):
p
.
647
655
.