Comparison of Corrosion Performance of AA7075 and AA2070 in Various Test Environments

Age-hardenable Al alloys are vital to aerospace applications due to their high specific strength; as such, their corrosion resistance is critical to ensuring long-term sustainable use.¹  To date, the aerospace industry has focused on legacy Al-Cu-Mg (AA2xxx) and Al-Zn-Mg-Cu (AA7xxx) alloys. In recent years, third-generation Al-Li alloys like AA2099, AA2199, AA2098, AA2050, AA2060, and AA2070 have garnered interest due to their decrease in density and enhanced corrosion resistance when compared to incumbent alloys.^2-3  For the legacy/incumbent aerospace alloys, it is well established that individual microstructural features, such as constituent particles and dispersoids, establish local anodes and cathodes that lead to galvanic corrosion on the metallurgical scale, which controls alloy corrosion resistance.^4-6  Currently, the aerospace industry predicts in-service corrosion performance through testing and evaluation of materials exposed to laboratory-accelerated corrosion cabinet testing and/or established outdoor test sites. However, some currently established cabinet tests are not always indicative of the corrosion morphology exhibited from long-term outdoor exposure.^7-11  Therefore, a comparison of these tests is necessary to ensure proper life prediction and alloy selection with regards to corrosion performance. This is particularly true for AA2XXX Al-Li alloys, which research has shown exhibit different exfoliation susceptibility in various accelerated tests and in comparison to Point Judith, RI seacoast exposure.^9-11  Historically, the indoor laboratory-accelerated cabinet style corrosion tests utilized to characterize the corrosion performance of age-hardenable Al alloys in various environments over a wide range of exposure times include ASTM B117,¹²  ASTM G34 (EXCO),¹³  ASTM G110,¹⁴  and ASTM G85.¹⁵ 

Through the thermo-mechanical processing (TMP) of age-hardenable Al alloys, several different types of secondary-phase intermetallic particles (IMPs) are formed: precipitate particles, dispersoids, and constituent particles. Previous researchers have established these IMPs can serve as local anodes or cathodes in comparison to the matrix and drive localized corrosion of age-hardenable Al alloys.^4,16-25  Several studies by Birbilis and Buchheit have established that the formation of these secondary phases, particularly large constituent particles, in 7xxx alloys can drive pitting and detrimentally impact the overall corrosion properties of an age-hardenable Al alloy due to galvanic interactions between the IMPs and the matrix.^4-6,16-17,26  A common method utilized to characterize the electrochemical properties of individual metallurgical phases is the microelectrochemical cell (microcell) method.^5-6,25-31 

While there are integrated computational materials engineering (ICME)-based approaches for predicting mechanical properties and performance based on alloy metallurgy from thermo-mechanical processing, the same does not exist for corrosion resistance. As such, the development of an ICME approach to predict corrosion performance of aerospace Al alloys would enable significant enhancement in alloy design, life prediction, and risk assessment.

In order to develop an ICME approach to corrosion-related alloy design, data must be generated for both input parameters and validation and verification sets. Input data could include the TMP path, metallurgy, and electrochemical properties of individual microstructural elements. Validation and verification data would come from accelerated and/or outdoor corrosion testing or observations from in-service material. Due to the wealth of corrosion data on AA7075-T6, both from microcell and accelerated and outdoor corrosion testing, this alloy is a viable starting point for the development and verification of an ICME approach aimed at the prediction of corrosion performance.

The overall objective of this work is to provide insight into the evolution of corrosion, both in morphology and depth of attack, over time as a function of temper for AA2070-T8 in comparison with the more conventional age-hardenable Al alloy AA7075-T6 to provide input, verification, and validation to facilitate an ICME approach to alloy design that incorporates the effect of processing (i.e., temper) on corrosion morphology. Specifically, the aims of this work were to quantify corrosion morphology and depth of attack for AA7076-T6 and AA2070 in various tempers after various exposure times in two different accelerated cabinet style corrosion tests and two different outdoor exposure environments to understand corrosion type, evaluation, and the ability to utilize accelerated methods to predict outdoor performance. Related work to develop electrochemical input data on the discrete microstructural elements found in AA2070, as AA7075 is largely available in the literature;^{4-6,16-17,23,25-26}  utilizing microcell will be in a subsequent publication.

Two aluminum alloys, AA7075 in the T6 temper and AA2070 in the T3, T8, and two overaged tempers, were utilized in this study. The registered composition limits for each alloy are listed in Table 1. AA2070 was received from Arconic Inc. as a 6.35 cm thick plate in the T3 and T8 tempers. To create the overaged tempers, T8 material was aged for 48 h at 150°C, which will be referred to as T8+1, and for 120 h at 150°C, which will be referred to as T8+2. Vickers hardness was used to check the aging practice. AA7075 was produced by a Kaiser Aluminum Llc.as 7.62 cm thick plate. Specimens for metallurgical and corrosion testing were taken from the rolled plate at the t/4 position, where t is the thickness of the plate. Grain size, determined via the line intercept method, is contained in the Results section.

Corrosion coupons, measuring 15.24 cm by 7.62 cm, were machined to expose the surface containing the L (longitudinal aka rolling direction) by T (long transverse aka plate width direction) plate directions. All coupons for accelerated cabinet and outdoor exposure testing had the surface oxide and machined altered surface layers removed prior to exposure by first cleaning in acetone with ultrasonic agitation for 5 min and air drying then exposing to an alkaline etch consisting of 15% NaOH at 180°F for 2 min and rinsing with deionized (DI) water. This was followed by a desmutting procedure consisting of immersion in 1:1 HNO3:DI water for 30 s, after which the specimens were rinsed with DI water and left to air dry. Prior to any corrosion testing or outdoor exposure, the starting topography of the specimen surface was characterized using white light interferometry (WLI).

ASTM G34,¹³  the standard test for exfoliation corrosion susceptibility of AA7XXX series aluminum alloys (EXCO), was performed on AA7075-T6 for 3 h to 96 h. The solution pH was maintained at ∼0.4 in a 4 M NaCl + 0.5 M KNO₃ + 0.1 M HNO₃ immersion solution. The specimens were set flat on the bottom of the container with the back and edges concealed with stop off lacquer to ensure that no exfoliation corrosion flaked off the sample surface thereby retaining the full corrosion morphology. This test was not performed on AA2070, as the literature establishes that EXCO does not correlate with seacoast exposure estimations of corrosion for third-generation Al-Li alloys.^9-10  The specimens were ranked based on the severity of exfoliation corrosion, as determined by a blind survey of ten individuals. The specimens were ranked as P (pitting), EA (superficial), EB (moderate), EC (severe), and ED (very severe) by comparing the corrosion morphology of laboratory specimens to the examples given in ASTM G34.

ASTM B117¹²  was utilized to determine the corrosion performance in a pH neutral accelerated cabinet test. Samples were set at a 30° exposure angle for the full exposure time in an atomized 5 wt% NaCl solution (pH ∼6.5–7.2). Exposure times for all specimens ranged from 24 h to 168 h.

ASTM G85,¹⁵  Annex 2, or the modified ASTM acetic acid salt intermittent spray (MASTMAASIS), was performed from 6 h to 4 weeks. This test has been shown to correlate with outdoor exposure for third-generation Al-Li alloys.^9-10  Specimens were set at a 30° exposure angle for the full exposure time in an atomized 5 wt% NaCl solution. The sample chamber was kept at 49°C. This test utilizes an acidified salt spray test (pH ∼2.8–3.0) with an intermittent air purge on a 6 h cycle consisting of ¾ h spray, 2 h dry-air purge, and 3 ¼ h soak at high relative humidity.

To compare a marine atmospheric exposure to a more industrial exposure location, outdoor laboratory testing was performed at two locations. The Naval Research Laboratory in Key West provided a marine exposure environment while the Arconic exposure site in Carson, CA was determined to be an appropriate industrial testing location. Specimens were exposed from 3 months to 1 y at each location at a 30° exposure angle for the full exposure time.¹³ 

After corrosion testing, all specimens were sectioned in two halves. One half was utilized for cross-sectional analysis and the other for WLI.

Cross-sectional metallurgical characterization of all material was performed after exposure on both the surface containing the S (short transverse aka plate thickness direction) by T plate directions and the L X S face to determine the type and depth of corrosion attack. It is important to note that all figures contained herein are for the S X T face, but results are consistent between the two faces examined. Before cross sectioning, all specimens were cleaned with nitric acid for 10 min according to ASTM G1³²  to remove any corrosion product. The 10 min time was selected as it had been found in previous research to remove sufficient corrosion product while not imparting any surface corrosion on a pristine sample. Material were sectioned and mounted in EpoThin™^† two-part epoxy and polished with successive grades of SiC paper, 1 μm diamond paste, and finally, colloidal silica. All mounted specimen were etched with Keller’s reagent (190 mL H₂O, 5 mL HNO₃, 3 mL HCl, and 2 mL HF) for 30 s to expose the metal microstructure. Micrographs were taken before and after etching to ensure etching did not cause further corrosion damage.

Pit depth measurements were determined in cross section as both the maximum and average measurements. The maximum was taken as the average of the five largest pits seen in all cross sections. At least three cross sections per sample were analyzed. Additionally, for AA7075-T6, two separate samples per exposure time were tested and evaluated. Due to material limitations, only one sample per temper and exposure time was tested and evaluated for AA2070. Intergranular corrosion (IGC) depth was taken as the maximum if and when seen in any cross section.

WLI is a noncontact, three-dimensional optical imaging technique used to qualitatively measure the surface topography of a material. Following exposure, a (11.5 mm × 11.5 mm) region of the L X T exposed surface was imaged with a Zygo NewView 7300^† interferometer using a 5X SLWD lens (8.836 μm lateral resolution), 0.5X image zoom, normal FDA noise threshold, and 320 × 240 at 210 Hz camera mode. This was used to determine the corrosion morphology and depth of attack along the exposed surface. The topographic data file produced by the interferometric scan (an ASCII-type. xyz file) was postprocessed using a MATLAB^† routine developed at UTRC. The code converts the three-dimensional data file into a binary image file that Si processed to include only the pitting features below the nominal surface. Image analysis was performed in situ using the MATLAB Image Processing Toolbox^† to identify discrete corrosion pits and to tabulate their centroid coordinates, mean diameters, and average and extreme depths for each region. Cross-sectional metallurgical characterization of all material was performed in the S X T face to confirm and/or supplement the WLI information as WLI is limited by line of sight. Before analysis, all material was cleaned with nitric acid for 10 min according to ASTM G-1 to remove any corrosion product.

Figure 1 shows a 3D isopleth of the microstructure of AA7075-T6 and AA2070-T8 after etching using Keller’s reagent. In AA7075, the average grains size measures 850 μm in L, 225 μm in T, and 50 μm in S. It is similarly noted that the grains for the AA2070-T8 alloy measure 550 μm in L, 275 μm in T, and 60 μm in S. Figure 2 shows the hardness value as a function of aging time where AA2070-T3 was aged from the natural condition to approximately the T8 condition. The 60 h data point represents the T8 condition as received from Arconic. Data points after 60 h were aged from the T8 starting condition and the x-axis location indicates the total time aged assuming the T8 condition already received 60 h. For example, the T8+1 temper would be located at 108 h (48 h OSU aging + 60 h assumed aging by Arconic). The hardness values shown in Figure 2 agree with expected trends for an age-hardenable Al alloy, i.e., hardness initially increases to a peak value and then decreases with additional aging. The commercial T8 alloy is slightly below the peak hardness value as seen in Figure 2, but it should be noted that hardness is only being used a proxy to assess proper times for overaging. T8+1 and T8+2 were chosen to ensure they were aged past the peak position and do not correspond with any commercial temper available for this alloy.

Figures 3 through 5 establish that exfoliation occurs on AA7075-T6 when exposed to ASTM G34 (EXCO) for sufficient time. A macroimage of each specimen is shown in Figure 3 and the ASTM G34 ratings after visual inspection are shown in Figure 4. The 3 h exposure was ranked as “pitted” as was also observed in the top down optical in Figure 3. The 6 h and 12 h AA7075 exhibited blistering and the beginnings of exfoliation damage. These were ranked as EA and EB, respectively. Material immersed for 24 h and longer showed severe exfoliation and were ranked as EC or ED. Specimens immersed for 48 h to 96 h exhibited very severe exfoliation (an EXCO rating of ED). While the 6 h exposure was ranked as EA, “superficial exfoliation,” it showed IGC and only slight evidence of the beginnings of the grain lift off, which are characteristic of exfoliation corrosion, upon cross sectioning as shown in Figure 5(a). The corrosion damage found after 6 h exposure was approximately 75 μm in depth. After 48 h of exposure, exfoliation via grain lift off was evident from surface examination and in cross section, as shown in the cross-sectional view in Figure 5(b). The depth of attack from the cross-sectional view after 48 h was approximately 100 μm. After 96 h, dramatic lifting of the surface occurred, as shown in Figure 5(c). This morphology is typical of exfoliation often seen for alloys and tempers where grain boundaries preferentially corrode and voluminous corrosion product lifts grains off the surface.¹⁹ 

WLI was performed after 3 h of EXCO exposure and the results agree with optical observations that pitting has occurred on the sample surface. Figures 6(a) and (b) show WLI results for the 3 h exposure and confirm the P rating. After 6 h of exposure, shown in Figure 6(c), small blisters formed on the sample surface. Additional WLI for the longer exposure times was performed; however, due to the nature of exfoliation corrosion that produces grain lift off the surface and WLI’s line of sight limitation, WLI was not useful in depth of attack evaluations after EXCO testing. Presumably, the most heavily corroded areas would exhibit the largest amount of grain lift off and therefore register at heights above the uncorroded surface and veil the depth of the attack below.

The corrosion morphology and depth of corrosion attack after ASTM B117 is shown in cross section and from above in Figures 7 thorugh 9, respectively. The results show that pitting was the main corrosion morphology for all alloys, temper, and exposure times, as seen in Figures 7 and 8 for cross-sectional analysis and WLI, respectively. The WLI scans of AA7075-T6 (shown in Figure 8[a]) reveal a relatively uniform distribution of pits across the exposed surface while a lesser number of pits are unevenly dispersed for all AA2070 tempers, as shown in Figures 8(b) thorugh (e). Figure 9 shows the depth of attack measurements for all alloys and tempers after 168 h of exposure, the longest exposure time examined for ASTM B117 testing in this study. The maximum pit depths measured from WLI analysis are greater than that of the cross-sectional analysis, yet the average depths for WLI are less than that from cross section. These discrepancies are most likely due to the ability of WLI to measure the entire surface and incorporate a far larger number of pits than was able to be done via manual cross-sectional analysis. Depth of attack as a function of exposure time will be discussed in a subsequent section.

The results from ASTM G85-A2 testing are shown in Figures 10 through 12. Pitting is seen for AA7075-T6, AA2070-T8, AA2070-T8+1, and AA2070-T8+2, as cross-sectional analysis shows in Figure 10 and WLI shows in Figure 11. After 4 weeks of exposure, IGC fissures and pitting along stringers (i.e., corrosion of larger secondary phase particles aligned and broken up in the rolling direction) appear in the AA7075 alloy, as seen in Figures 10(a) and (f). This pitting along stringers, shown more clearly in cross section of the S X L orientation in Figure 10(f), is likely due to the presence of large, hard, cathodic particles such as Al₇Cu₂Fe which control localized corrosion. The pitting along stringers is not able to be captured using a characterization technique such as WLI. Severe IGC and grain lifting was observed for AA2070-T3, as shown in Figure 10(b). This specimen was unable to be imaged using WLI due to the exfoliation on the material surface. The commercially peak aged and the overaged AA2070 tempers show pitting that undercuts the surface in cross section. Figure 11 shows the results from WLI for peak and overaged AA2070, which suggests general corrosion with very little of the original surface remaining.

The maximum and average pit depth and IGC depth, as determined by cross-sectional analysis, are shown in Figure 12 for 4 weeks of exposure to ASTM G85-A2. All alloys/tempers except AA2070-T3 exhibited approximately the same average and maximum pit depth. It is important to note that only 2070-T3 showed appreciable IGC and, as such, is the only alloy/temper where IGC depth was measured and reported. While AA2070-T3 had the lowest average and maximum pit depth, this alloy was susceptible to IGC attack, which was measured as deep as 400 μm. Depth of attack as a function of exposure time will be discussed in a subsequent section.

An example of the corrosion morphology after a year at seacoast for all alloys and tempers is shown in Figure 13. While pitting is present on all alloys and tempers, clear IGC attack was seen on AA7075-T6 and AA2070-T3. For AA2070-T3, IGC was the primary mode of corrosion attack. For the overaged tempers of AA2070, intersubgranular attack (IsGC), or attack of the dislocation substructure within a grain, was also apparent, as shown in Figure 14. In a rare instance, IGC was found in the overaged tempers of AA2070, but confirmed IGC was only found once. AA2070-T8 was found to have only pitting. The pit depth was measured all alloys and tempers and is shown in Figure 15 for 1 y of exposure. After 1 y of exposure at the seacoast, all alloys and tempers had a maximum depth of pitting attack of approximately 60 μm. While AA2070-T3 had the lowest maximum and average pitting depth at approximately 40 μm and 20 μm, respectively, IGC was measured to have a maximum depth of approximately 110 μm. Due to the shallow nature of the pits for up to 1 y of seacoast exposure, WLI imaging was not able to accurately measure pit depth.

Corrosion results after a year of exposure in an industrial atmosphere are contained in Figures 16 and 17. The results are similar to the seacoast exposure. AA7075-T6 and AA2070-T3 both exhibited pitting and IGC, with IGC being the primary corrosion morphology for AA2070-T3. The overaged tempers of AA2070 again showed IsGC and very rare instances of possible IGC. AA2070-T8 only showed pitting. From the measurements of the depth of attack contained in Figure 17, AA2070-T3 had the lowest depth of pitting attack, but IGC was measured to depths of 80 μm after 1 y of exposure. The pit depth measured for all alloys and tempers was on the order of ∼30 μm to 50 μm. Again, due to the shallow nature of the pits for up to 1 y of industrial exposure, WLI imaging was not able to accurately measure pit depth. This technique may be reevaluated after longer exposure times for samples still in test. Depth of attack as a function of exposure time will be discussed in a subsequent section.

The progression in the corrosion attack as a function of time is shown in Figure 18 for both accelerated cabinet tests. In the ASTM B117 chamber, each specimen had an increasing depth of attack with time. At the shortest time points (24 h and 48 h), all alloys and tempers had a depth of attack of approximately 10 μm to 20 μm with little statistical difference between all values reported. Upon further exposure, the depth of attack does slightly vary with alloy and temper, particularly at the 168 h exposure time, the longest conducted in this study. Specifically, the 2070-T3 specimen had a depth of attack of 40 μm while the AA7075-T6 specimen had a depth of attack of 70 μm. Stated another way, at the 168 h exposure time, the AA7075 had the highest and AA2070-T3 had the lowest corrosion attack depth. However, upon consideration of the error bars reported, only a slight statistical difference is seen between any of the measurements. No IGC was observed on this test, and, therefore, was not measured.

In the ASTM G85-A2 accelerated cabinet test, the depth of attack was approximately 20 μm after 6 h and increased to approximately 200 μm after 4 weeks for all samples. Little statistical difference was observed as a function of alloy or temper for pitting. The depth of IGC on AA2070-T3 was measured to be deeper than pitting for all times but 6 h. The IGC attack was measured to be approximately 20 μm after 6 h and 400 μm after 4 weeks. The IGC growth rate was faster than that for pits as shown by the increasing difference between the IGC depth and pit depth as a function of time.

The progression in the corrosion attack as a function of time is contained in Figure 18 for both the seacoast and the industrial outdoor exposures. The pitting depth is shown for each alloy as a closed symbol, while the IGC depth of attack for AA2070-T3 is shown as an open square. It was observed that the depth of all corrosion increased as a function of time. After 3 months, the pit depth of attack was approximately 20 μm to 40 μm at both locations. The depth of IGC for AA2070-T3 was approximately 50 μm at seacoast and 30 μm in the industrial location. Interestingly, the IGC depth of attack was notably higher in the seacoast exposure, as was the difference between pitting and IGC. After 1 y at these locations, the pitting depth of attack at the seacoast had increased to approximately 60 μm to 70 μm for all alloys and tempers, while the industrial exposure depth of attack was 40 μm to 60 μm. The IGC depth of attack on AA2070-T3 after 1 y was approximately 120 μm at seacoast and 80 μm in the industrial setting. In general, pitting depth of attack for all alloys and tempers overlaps when error bars are considered. For the industrial setting, AA2070-T8+2 appears to develop a statistically significant deeper pitting attack than the other alloys for the last two exposure times, but longer exposure times should be evaluated to determine this definitively.

A direct comparison between the two outdoor exposure locations is shown in Figure 19 for all alloys as a function of exposure time. For AA7075-T6, AA2070-T3, and AA2070-T8 (Figures 19[a] through [c]), the seacoast exposure site initially had a lower depth of attack, but after 1 y developed deeper pits than the industrial setting. It should be said that this trend is only statistically significant for AA7075-T6 at 1 y. Longer times should be considered for AA2070-T3 and T8. For AA2070-T8+1, pits are always deeper at the seacoast site, and, again, there is not always a statistically significant separation in the measurements. For AA2070-T8+2, the depths of attack at the seacoast site is initially deeper, but the industrial site transitions to the deepest after 9 months. Again, this is not statistically significant when error bars are considered. For IGC of AA2070-T3, a statistically significant trend can be identified. The depth of IGC is consistently greater under seacoast exposures, and the difference between the seacoast and industrial sites increases with time.

ASTM B117 showed pitting only for both alloys and all tempers. On the other hand, ASTM G85-A2 revealed pitting and IGC for AA7075-T6 and AA2070-T3, which is consistent with that seen in outdoor exposures. Additionally, ASTM G85-A2 showed pitting only for AA2070-T8, AA2070-T8+1, and AA2070-T8+2. The environmental conditions in these accelerated tests are extremely different. While ASTM B117 has a relatively neutral pH, ASTM G85-A2 is acidic through the addition of acetic acid. Also, ASTM B117 has a continuous salt spray and keeps the samples in a hot, humid atmosphere while ASTM G85-A2 is a cyclic test (intermittent air purge on a 6 h cycle consisting of ¾ h spray, 2 h dry-air purge, and 3¼ h soak at high-relative humidity) which varies the relative humidity in the cabinet. While both tests are utilized to understand the corrosion resistance of aerospace Al alloys, it is important to compare the corrosion morphology to the expected service exposure environment. Should the accelerated test not match with outdoor exposure testing, it should not be utilized in the creation of an ICME approach to corrosion-resistant alloy design.

The corrosion morphology in both the seacoast and industrial exposure locations was characterized by IGC on the AA7075-T6 and AA2070-T3 with pitting occurring solely on AA7075-T6. For the overaged tempers of AA2070, IsGC occurred in addition to pitting. Only pitting was seen on the AA2070-T8. By comparison of results in accelerated cabinet tests and outdoor exposure, it is shown that the ASTM G85-A2 corrosion morphology is more representative of the expected corrosion attack seen at both the seacoast exposure location and the industrial location. Both outdoor locations would have cyclic wetting and drying (in the form of rain and sun) which is replicated in ASTM G85-A2.

The corrosion morphology of AA2070 was shown to vary as a function of temper. For the underaged temper, the alloy was susceptible to IGC and grain lifting in both the ASTM G85-A2 accelerated cabinet test and at the long-term outdoor exposure locations. In ASTM B117 testing, no change in corrosion morphology as a function of temper was found. This variation in corrosion susceptibility with heat treatment has previously been reported.^33-35  The dominant corrosion morphology may have been impacted by the presence of Cu-containing IMPs at the grain boundaries which are cathodic to the surrounding matrix.³⁶  In the underaged condition, a Cu-depleted region may be formed at the grain boundaries due to the existence of Cu-containing IMPs directly on the grain boundary.^33,36  This can lead to preferential corrosion of the Cu-depleted near grain boundary areas and IGC attack in the T3 temper as the corrosion potential has been well established to increase with solid solution Cu content.¹⁸  As aging produces Cu-containing strengthening precipitates in the matrix, the difference in corrosion potential between the near grain boundary region and matrix would lessen as the Cu content homogenizes, thereby healing the susceptibility to IGC and creating pitting as the dominate corrosion morphology in the aged tempers. This explanation for change in predominant corrosion morphology has been suggested by other researchers when a similar trend was found in AA2050. ^34,37 

However, this homogenization impact on the corrosion resistance in the peak aged condition could be reversed due to the coarsening of the Cu-containing particles upon further overaging.³⁸  As has been reported in the literature where extremely aged AA2050 has shown increased propensity for IGC attack.³⁴  IGC was only seen on rare occasion in the T8+1 and T8+2 tempers investigated here. Longer aging time would need to be studied to determine if and when extremely long aging in AA2070 produces prolific IGC.

WLI is a useful technique which can supply large amounts of statistical data over a short period of time for a wide variety of corrosion morphologies.³⁹  However, proper use of this technique requires understanding both its capabilities and its limitations. For instance, WLI will not give an accurate measurement of pit depth for samples where the pitting undercuts the sample surface because a large portion of the pit will be out of the line-of-sight of the instrument. As such, undercutting corrosion morphology like undercutting pits (as seen in AA2070 T8, T8+1, and T8+2) or IGC and exfoliation (as seen in AA7075-T6 and AA2070-T3) has to be analyzed in a cross-sectional view.²²  Also, for exfoliated samples such as AA7075 in the EXCO solution and AA2070-T3 in ASTM G85-A2, the exfoliated morphology was not able to be measured using WLI due to line of sight issues. An example of this is shown in Figure 20 where the WLI instrument was not able to fully resolve the exfoliated topology in the 7075-T6 sample. Also, the different angles from the exfoliated layer cannot be captured in the WLI field-of-view.

Furthermore, WLI has the ability to interrogate a greater number of pits and, therefore, it provides a complementary analysis to cross-sectional metallography with the additional benefit of measuring average and extreme pit depths for a given area. As such, the maximum pit depths that are reported from WLI analysis are greater than that of the cross-sectional analysis. However, the average pit depth as determined by WLI is much smaller than that measured by cross-sectional analysis because WLI is able to capture a larger number of pits/corrosion features.

Various accelerated cabinet tests were performed to determine the corrosion morphology of two different aluminum alloys with various temper designations.

• •

  It was shown that the observed corrosion morphology varied by alloy type and temper with AA7075-T6 and AA2070-T3 showing exfoliation and pitting and AA7075-T6 and AA2070 in the T8, T8+1, and T8+2 tempers showing pitting.

• •

  It was shown that the observed corrosion morphology varied by testing environment with the ASTM G85-A2 showing better agreement with outdoor exposures than ASTM B117.

• •

  Comparing the accelerated tests to the outdoor exposure locations, ASTM G85-A2 provides a more accurate representation of the corrosion morphology which would be expected during a long-term exposure.

• •

  AA2070 in the overaged tempers exhibits IsGC whereby the dislocation substructures (i.e., subgrains) undergo preferential attack. This type of attack was seen most prolifically in the outdoor exposures.

• •

  Seacoast exposures were found to produce a more aggressive IGC attack in AA2070-T3 than industrial exposures, while the pitting depth of attack was consistent between outdoor exposure locations.

• •

  The WLI instrument was able to characterize a larger population of pit depths as well as extreme pit depths which might be missed in cross-sectional analysis.

The authors wish to thank Dr. Lynne Karabin and Ms. Marsha Egbert at the Arconic Technical Center for their assistance in performing outdoor exposure at the Carson, CA test site. Additional acknowledgment to Dr. Lynne Karabin for providing AA2070. The authors also wish to thank Dr. James Moran for his assistance in connecting with the NRL outdoor exposure site and Dr. Christine Sanders for her help in performing the NRL outdoor exposures. The work by Mia Sethi and Raghav Sahai is gratefully acknowledged. This material is based on research sponsored by Office of Naval Research under Agreement No. N00014-14-2-0002. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the Office of Naval Research or the U.S. Government.