A Study of the Effects of Microstructure on the Seawater Jet Impingement Resistance of Nickel Aluminum Bronze

Results (cont'd)

Mass Loss

Because results might be unduly influences by the localized corrosion, corrosion rates were not calculated. However percent mass loss were determined for all eleven specimens. The percent mass losses results, after 30 days of jet impingement testing at a testing at 15 fps and 30 fps and after cleaning to remove bulk corrosion products, are shown in Table 7. The mass loss is associated with material removed by the water jetting upon the impingement area, and other forms of corrosion identified in Table 6. Some of the mass loss, particularly in the 30 fps test is also related to corrosion which occurred on those portions of the specimen surfaces not exposed to direct impingement during testing.

Table 7. Percent Mass Loss after Thirty-Day Natural Seawater Jet Impingement Testing
Heat Treatment CodePercent Mass Loss
15 fps30 fps
After CleaningFirst CleaningSecond CleaningThird Cleaning
Alloy A (9.4% Al)
OQ 0.48 1.17 1.19 1.20
OQ+T 0.34 0.86 0.88 0.89
AC 0.39 0.93 1.05 1.08
AC+T 0.32 0.80 0.87 0.89
FC 0.42 0.85 0.94 0.95
Alloy B (8.1% Al)
OQ 0.21 1.45 1.49 1.53
OQ+T 0.32 0.95 0.98 0.99
AC 0.27 1.18 1.49 1.55
AC+T 0.35 1.00 1.10 1.13
FC 0.36 0.66 0.73 0.75
Alloy C (8.8% Al)
As-Cast (Control) 0.38 0.83 0.94 0.96

After the 15 fps test, percent mass loss ranged from 0.32 % to 0.48 % in Alloy A and from 0.21 % to 0.36 % in Alloy B. The mass loss in the control, Alloy C was 0.38 %. Even when each alloy in each heat treatment combination is examined, no clear trend is apparent, except that the average mass loss for Alloy A is 0.39 % while it drops to 0.30 % for Alloy B. In summary, the average mass loss observed in the 15 fps (~4.6 mps) jet impingement test is quite low, and no correlation with alloy composition, heat treatment or microstructure is apparent.

As previously mentioned, it was necessary to clean the 30 fps specimens three times to remove the corrosion products. The percent mass loss, after each cleaning is shown in Table 7. The results indicate that the second cleaning removed almost all of the corrosion products. After the second cleaning, the percent mass loss ranged from 0.88 % to 1.19 % in Alloy A and from 0.73 % to 1.49 % in Alloy B. As was the case at 15 fps, the average mass loss realized in the 30 fps (~9.1 mps) test is quite low, and averages 0.99 % in alloy A, 1.16% % in Alloy B. It was 0.94 % for the control, Alloy C. Four specimens, Alloys A and B in the FC condition, and Alloy A in both the OQ+T and AC+T conditions, exhibited mass losses lower than the control, Alloy C. Overall, no correlation with composition, heat treatment, or microstructure is apparent.

The data in Table 7 indicates that in most case that the percent mass loss doubled or tripled upon going from 15 to 30 fps. The only exceptions are Alloy B in both the OQ and the AC conditions, which exhibited a six- to seven-fold increase.

A comparison of the percent mass loss, based upon the quiescent six-month exposures in seawater ( 14) to the two 30-day jet impingement test, can be seen in Table 8. Despite the longer duration of six months versus 30 days, and the presence of significant localized attack associated with the quiescent test, most of the 15 fps and all of the 30 fps jet impingement mass losses were greater. Although it is clear that Alloy A exhibited a lower percent mass loss than Alloy B in the quiescent test, the two alloys performed about the same in both jet impingement tests. This is in contrast to Shalaby et al ( 15), who compared rate of mass loss rate under quiescent conditions, with those developed after ultrasonically induced corrosion cavitation. Their laboratory test results showed that the cavitation, under cathodic protection, was ten times greater, and under freely corroding conditions, was 186 times greater than the quiescent result. However, their short term mass loss rate results under ultrasonic conditions results can not be readily compared to the jet impingement data shown in Table 8.

Table 8. Comparison of Six month Quiescent Seawater Test Results with Thirty Day Seawater Jet Impingement Result
Heat Treatment CodePercent Mass Loss
Quiescent TestJet Impingement Tests
15 fps30 fps*
Alloy A (9.4% Al)
OQ 0.12 0.48 1.19
OQ+T 0.23 0.34 0.88
AC 0.14 0.39 1.05
AC+T 0.16 0.32 0.87
FC 0.17 0.42 0.94
Alloy B (8.1% Al)
OQ 0.32 0.21 1.49
OQ+T 0.27 0.32 0.98
AC 0.41 0.27 1.49
AC+T 0.65 0.35 1.10
FC 0.58 0.36 0.73
* after second cleaning

The large mass loss observed in Alloy B in the quiescent seawater corrosion test ( 14) was associated with severe localized attack under corrosion deposits which emanated from the interface between the specimen positioning hole and glass rod from which the specimens were suspended in the seawater. These were loose crevice sites. At that time, it was surmised that these deposits might not form in actual service, or would be swept away under more dynamic conditions. While some crevice corrosion was noted after the 30 fps tests, the resulting corrosion products did not appear to affect the boldly exposed surfaces as they did in the earlier quiescent test.

In summary, the percent mass loss results for both Alloys A and B, at each of the two velocities in the jet impingement tests, shows no clear correlation with composition, heat treatments or microstructures. It is noted, however, that the mass loss in the tempered conditions, OQ+T and AC+T, for both Alloys A and B were less than their respective OQ and AC counterparts.

Maximum Depth of Attack

Only three specimens developed visible impingement in the 15 fps test, as previously noted in Table 5. In contrast, eight specimens showed such damage after the 30 fps test. The maximum depths of attack, which is related to erosion-corrosion damage, are shown in Table 9. From the 15 fps data, it can be seen that the maximum depths of attack for Alloy A in both the OQ+T and AC+T conditions, exceeded that incurred by the control, Alloy C. All other specimens sustained negligible attack. At 30 fps, the same two Alloy A specimens ( OQ+T and AC+T ), and the Alloy B specimen in the AC+T condition, exhibited maximum depths of attack at least as great or greater than seen in the control, Alloy C. In the case of the Alloy A specimens in the OQ+T and AC+T conditions, increasing the impingement velocity had a greater effect on the diameter of the attacked zone, as listed in Table 5, than it did on the maximum depth of attack. From Table 9, it can be also be seen that the depths of attack in the 30 fps in the other Alloy A and Alloy B specimens in the 30 fps test were less than those measured for the control, Alloy C.

Table 9. Maximum Depth of Impingement Attack (mm) after Thirty-Day Natural Seawater Jet Impingement Testing
Heat Treatment CodeMaximum Depth of Attack
15 fps30 fps
Alloy A (9.4% Al)
OQ nil nil
OQ+T 0.15 0.19
AC nil 0.07
AC+T 0.15 0.16
FC nil nil
Alloy B (8.1% Al)
OQ nil nil
OQ+T nil 0.06
AC nil 0.11
AC+T nil 0.17
FC nil 0.05
Alloy C (8.8% Al)
As-Cast (Control) 0.05 0.16

In some cases, penetrations deeper than those shown in Table 9 were found within the impingement zone. However, because they associated with intergranular or selective phase attack, as opposed to erosion corrosion, they were disregarded in the present analysis. Furthermore, in some specimens, localized attack, as listed in Table 6, which was observed elsewhere on the specimens, exceeded that in the impingement zone. What, if any, effect these other sites had on impingement susceptibility is presently uncertain. While the Alloy A specimen in the OQ condition exhibited negligible impingement attack in both tests, the same specimen showed the highest percent mass loss. The same can be said for the corresponding Alloy B specimen in the 30 fps test. On the other hand, Alloy B specimens in the AC and AC+T conditions exhibited both high percent mass loss and deep penetrations relative to some others.

Depth of attack has been previously found to increase when the cathodic or noble area outside the impingement area is increased ( 16, 17). In this study, Alloy A specimens in the OQ+T and AC+T conditions, as well as Alloy B in the AC+T condition, had the greatest surface areas, because of thickness. While these exhibited the deepest penetrations, their surface areas were almost the same as that of the control, Alloy C.

Nevertheless, beyond the observation that Alloy A, in both the OQ+T and AC+T condition, and Alloy B in the AC+T condition exhibited greater values of maximum attack than the control, Alloy C, it is difficult to find clear correlations with alloy composition, heat treatment and microstructure. However the maximum depth of attack, even in the worst case, which is 0.19 mm in Alloy A in the OQ+T condition after the 30 fps test, is quite small.

© NACE International. All rights reserved.
Paper Number 05223 reproduced with permission from Corrosion/2005 Annual Conference and Exhibition, Houston, Texas