Two aluminized heater tubes were received for analysis to investigate the cause of failure. Unused tube material also was supplied for comparison. Results indicate that the tubes failed via hot oxidation corrosion at extreme temperatures. Both the tube material and processes appear normal. Tube failure appears to be related to burner geometry and flame impingement
Metallographic examination of the failed and comparison tubes revealed that the aluminized coating was an aluminum-silicon alloy (Type 1) and has the specified coating thickness (T1-40). The tube is manufactured from ASTM A463, CS, Type A carbon steel (~1008 plain carbon steel), as specified.
Near the holes, both failed tubes exhibited high temperature oxidation degradation extending through the tube wall. Examination of a cross-section through one of the failed tubes, taken from the opposite side of the tube from the hole, revealed a normal fine grain microstructure with the ID and OD coatings intact. The second failed tube revealed a much coarser grain size than either the first failed tube or the unused comparison tube. The coarse grain size indicates excessive heating.
Two failed heater tubes ("A" & "B") and the two unused tube samples for comparison were received for analysis to determine the cause of failure. Reportedly the tubes had been in service for five years at a temperature of 1200oF. One of the unused comparison samples was in the as-coated condition, and the other had been baked at 1200o F (which is the usual condition after manufacture). All the tubes were 4-in. diameter with a wall thickness of 0.065-in.
Reportedly the tubes were specified to be made from AISI/SAE Type 1008 tube material with aluminum-silicon alloy (Type 1) coating applied via a hot-dip process per ASTM A463. The specified coating has a 0.4 oz/ft.2 coating weight, which is designated as a T1-40 coating and corresponds to a coating thickness of 20.2 microns on each side.
Figure 1 shows an overall view of the two failed heater tubes ("A" & "B") and the two unused tube samples as received for analysis.
Spectrographic chemical analysis of the metal from failed tube "A" and an unused comparison tube was performed in accordance with ASTM E415-99a. Chemical analysis results provided in Table 1 indicate the tube is manufactured from 1008 low carbon steel (more specifically equivalent to ASTM A463, CS, Type A) as was specified. Both steels were deoxidized with an aluminum addition, which should result in a fine grain size. Both steels also contained an intentional titanium addition, which will stabilize the carbon as titanium carbide particles.
Figure 2 provides an interior view of the hole and adjacent areas on failed heater tube "B". An elliptical oxidation pattern is noted suggesting impingement of an angled flame.
Cross-sections were taken through the baked comparison tube and both of the failed tubes. Samples were prepared for metallographic analysis in accordance with ASTM E3-01. Etching in accordance with ASTM E407-99 revealed the microstructures that were examined using an optical microscope in accordance with ASTM E883-02.
Figure 3 shows an optical microscope view of a cross-section through the baked comparison tube. The microstructure is ferrite grains with a fine grain size, and is normal for an aluminum-deoxidized low carbon steel. A uniform coating of aluminum is present on both the OD and ID surfaces. Figure 4 shows a high magnification optical microscope view of the coating on the ID surface.
The specified T1-40 coating specifies a coating weight of 0.4 oz/sq. ft., which according to ASTM A463 corresponds to a two sided coating thickness of 40.4 microns. Typically about 50% of the coating is on each side, so a thickness of 20.2 microns would be expected. Figure 4 shows that on the ID the coating is 25 microns thick, as specified.
Figure 5 presents an optical microscope view of a cross-section through the failed tube "A" near the flanged end away from the hole. A uniform coating is present on both the OD and ID surfaces. However, a comparison with Figure 3 reveals that the grain size of the failed tube is very much coarser than on the comparison tube. The very coarse grain size indicates excessive heating.
Figures 6 and 7 provide high magnification optical microscope views of the ID and OD coatings on a cross-section through the failed tube "A" near the flanged end away from the hole. The coating thickness is 36 microns on the ID surface, and 44 microns on the OD surface. The coating thickness may have been increased by diffusion of aluminum in service, but was probably originally above the specified thickness. Figure 8 shows a localized area where the coating on the ID surface has been replaced by oxide scale.
Figure 9 provides a low magnification optical microscope view of a cross-section through failed tube "A" near the hole edge. There is severe oxidation extending through the tube wall. The boxed area is shown at higher magnification in Figure 10, which reveals islands of very coarse grains (similar to those in Figure 5) surrounded by oxide scale. The very coarse grain size and severe oxidation indicates exposure to flame temperatures.
Figure 11 provides a low magnification optical microscope view of a cross-section through failed tube "B" near the hole edge. Severe oxidation extending through the tube wall is again observed. An adjacent area is shown at higher magnification in Figure 12, which exhibits a fine grain size (similar to that in Figure 3) surrounded by oxide scale. Figure 13 shows an optical microscope view of a cross-section through failed tube "B" oriented 140o from the hole. Here the steel is intact, with a fine grain size and intact coating on both the ID and OD surfaces. The intact coating and metal on this side of the tube indicates flame impingement at the hole site.
The coating on a baked comparison tube and on failed tube "A", and deposits on the interior of tube "B" (Figure 2), were examined using a scanning electron microscope (SEM). The metal and scale deposits were analyzed by energy dispersive spectroscopy (EDS) per ASTM E1508.
Figure 14 presents an EDS analysis of the coating on the baked comparison heater tube. A large aluminum peak and smaller iron peaks are detected. Figure 15 presents an EDS spectrum of the coating on the failed heater tube "A". Large iron peaks and a smaller aluminum peak are detected, along with a small amount of silicon. This is consistent with the use of an aluminum-silicon coating bath. The higher iron peak is expected due to diffusion of the aluminum into the steel.
Figure 16 presents an EDS spectrum of the deposit on the interior of failed heater tube "B" near the hole. Large iron peaks are detected, along with a smaller oxygen peak. This indicates that the deposit is iron oxide. No unusual elements that would contribute to increased attack via high temperature corrosion are noted.
The service temperature for the tubes reportedly is below the 1250 to 1290o F limit expected for long term service without breakdown of the coating and catastrophic oxidation of the steel substrate. The grain coarsening observed on Tube A indicates that it was exposed to a very high temperature such as direct impingement of a flame. Aluminum deoxidized steels contain submicroscopic aluminum nitride particles, which keep the grain size from growing until the steel is heated to a very high temperature sufficient to dissolve the nitride particles. The exact temperature depends on the nitrogen content of the steel, but is approximately 1800 oF.
The tube failure is more likely than not, related to burner geometry and flame impingement rather than the tube materials and processes (which appear normal).
The failed tube "A" and the comparison tube meet the specification for ASTM A463 Grade CS Type A carbon steel, and also AISI/SAE 1008 tube steel. Both steels were deoxidized with an aluminum addition, resulting in a fine grain size.