S Paul,* C M Lee, M D F Harvey
TWI, Cambridge, United Kingdom
Paper presented at ITSC 2012 International Thermal Spray Conference, Houston USA, 21-24 May 2012, and published in Journal of Thermal Spray Technology.
In this paper evaluation of sealed and unsealed thermally sprayed aluminum (TSA) for the protection of 22%Cr duplex stainless steel (DSS) from corrosion in aerated, elevated temperature synthetic seawater is presented. The assessments involved general and pitting corrosion tests, external chloride stress corrosion cracking (SCC), and Hydrogen induced stress cracking (HISC). These tests indicate that DSS samples which would otherwise fail on its own in a few days do not show pitting or fail under chloride SCC and HISC conditions when coated with TSA (with or without a sealant). TSA-coated specimens failed only at very high stresses (>120% proof stress). In general, TSA offered protection to the underlying or exposed steel by cathodically polarizing it and forming a calcareous deposit in synthetic seawater. The morphology of the calcareous deposit was found to be temperature dependent and in general is of duplex nature. The free corrosion rate of TSA in synthetic seawater was measured to be ~5-8 µm/year at ~18°C and ~6-7 µm/year at 80°C.
Offshore operators are currently looking to extend the design life of offshore facilities, structures and components to improve the affordability, and to increase their availability in later years of operation. Duplex stainless steels, often used in offshore structures, are susceptible to localized corrosion and environmentally assisted cracking when subjected to loads and temperatures approaching their material design limits in marine environment. Conventional organic coatings (paints) provide limited mitigation due to their rapid degradation at elevated temperature, particularly in hot risers. Whilst the use of thermally sprayed aluminum (TSA) to mitigate the corrosion of C-steels and austenitic stainless steels is well-documented, there are little or no published data relating to its behavior on duplex stainless steels (DSS) or other high strength alloys.
To address these issues TWI undertook a Joint Industry Programme (Ref ) which comprised a detailed evaluation of the performance of TSA-coated, 22Cr DSS welds subjected to pitting corrosion, external chloride stress corrosion cracking (SCC) and hydrogen induced stress cracking (HISC). TSA corrosion rate data were produced at ambient and elevated temperature using two electrochemical techniques. The effects of sealant selection on TSA coating behavior were characterized and conclusions were drawn from the above data.
Twin wire arc spraying was used to spray commercially pure Al (99.5%) onto 22%Cr duplex stainless steel (DSS, UNS S31803). A TSA-coated glass sample was also prepared to measure the free corrosion potential of TSA. The TSA coating was applied to a nominal thickness of 250-300 µm in all cases. The sealant/topcoats used were standard commercial Al-silicone and epoxy phenolic formulations. The coated or uncoated DSS specimens were tested with and without welds.
Corrosion Potential (ECorr), Linear Polarization Resistance (LPR) and Zero Resistance Ammetry (ZRA)
The free corrosion potential 'Ecorr' of the TSA-coated glass, uncoated DSS weld, the TSA-coated and the TSA+Al-silicone sealed specimens was monitored in aerated artificial seawater (pH 7.6-8.0, 18±2°C and 80±2°C). LPR was also used where appropriate. One set of TSA coated and the TSA+Al-silicone sealed specimens were electrically connected to an uncoated DSS weld specimen giving a coating (anode) to DSS weld (cathode) area ratio of 95:5. The galvanic current flowing between the coating (anode) and the DSS (cathode) was continuously recorded using a zero resistance ammeter.
Pitting corrosion tests were conducted on the uncoated, TSA-coated and TSA+Al-silicone sealed DSS specimens (50×50×6mm) in aerated and slowly circulated artificial seawater at 80±2°C for 30 days. For the coated specimens a 10mm diameter coating holiday was introduced on the weld cap of the test specimen to expose the underlying DSS surface.
Chloride Stress Corrosion Cracking (SCC)
External chloride SCC tests by drop evaporation method were conducted for the uncoated DSS weld, the TSA coated, the TSA+epoxy painted DSS weld specimens. Tests were conducted by dripping cold synthetic seawater (temperature 18±2°C, pH7.9) at a feed rate of about 4 drops per minute onto the surface of a stressed weld specimen maintained between 130±10°C. The test specimens were strained corresponding to the 0.2% proof stress of the parent steel using a four-point bend jig as shown in Fig.1
Figure 1: Arrangement used for the chloride SCC test
Hydrogen Induced Stress Cracking (HISC)
TSA coating was deposited onto 3.8 mm diameter cylindrical tensile specimens. The substrate material was plain (unwelded) DSS. Coated and uncoated specimens were subjected to 3% plastic strain to induce cracking in the coating prior to the HISC test. HISC tests were conducted on these pre-strained tensile specimens in artificial seawater at an ambient laboratory temperature (18±2°C). A constant load was kept at 100% of the 0.2% proof stress value (575 MPa). After the initial exposure of 165 days, testing on one set of specimens was continued by increasing the stress level by 25 MPa per week. ECorr of the coated specimens was monitored. The uncoated specimens were under -1100 mVSCE cathodic protection potential.
At 18±2°C: The TSA coated and the TSA+Al-silicone sealed specimens recorded an initial Ecorr of about -700 to -750 mVSCE for the few first hours, after which the potential values began to lower significantly (towards the negative value) during the first 7 days. The TSA coated specimen after 7 days recorded a potential value of about -1300 mVSCE and the TSA+Al-silicone sealed specimen recorded a potential value of about -1100 mVSCE. The potential of the TSA coating remained significantly negative i.e. 'active' and changed with time during the initial 90-100 days. From about 120 days until the end of test (235 days), the TSA-coated specimen and the TSA+Al-silicone sealed specimen displayed steady Ecorr values of about -1000 and -900 mVSCE, respectively.
In addition, an attempt was made to monitor the corrosion rate of a TSA-coated specimen during the long-term HISC test. Measurements of the corrosion rate using the LPR technique for the TSA coated tensile specimen were started after 90 days of exposure. A steady corrosion rate of 5-8 µm/year was obtained for the TSA coated DSS specimen during the period of 120-160 days in static (unaerated) artificial seawater.
At 80±2°C: A stable Ecorr of the TSA coating on glass after 15 days of immersion was about -1050 mVSCE. A low corrosion rate was calculated for the TSA coating and was measured at about 6-7 µm/year using the LPR technique. The Ecorr of the uncoated DSS weld specimen shows a steady corrosion potential around -100 to -150 mVSCE. Steady Ecorr was observed for both the TSA and the TSA+Al-silicone specimens after about 20-25 days and were very similar for both coated specimens at about -900 mVSCE. The measured free corrosion potential values for both the TSA and the TSA+Al-silicone systems were sufficiently negative and indicate that an exposed DSS area of about 5% would be sacrificially protected. A higher initial corrosion rate was recorded from the TSA coating (anode) when coupled to about 5% area of DSS weld (cathode), but this trend lasted only for about 2-3 hours of immersion after which the corrosion rate dropped rapidly as protective calcareous deposits formed.
Detailed examination of the uncoated DSS weld specimen surface displayed evidence of numerous small, shallow corrosion pits. These pits were seen mostly on the surface away from the central weld cap region. The central weld region of the uncoated specimen had a thin film. Pitting or any type of corrosion attack was not seen on the central exposed weld (holiday) region of the TSA-coated samples with or without sealant. The central weld cap region of all the coated specimens had a thick layer of white corrosion product/calcareous deposit (Fig.2).
Figure 2: TSA-coated pitting corrosion specimen showing
(a) exposed weld area, and (b) cross section along the holiday.
The transverse cross sections of the specimens did not display any visible corrosion attack of the weld after 30 days of immersion. The cross section image of the TSA-coated specimen in Fig.2b shows a uniform layer of corrosion product (15-25µm thickness) on the entire TSA surface. The grey contrast phase adjacent to the TSA in Fig.2b was aluminium oxide based corrosion product as indicated by the presence of Al and O peaks from the EDX spectra. The brighter contrast phase is believed to be a CaCO3 scale. No measurable TSA disbondment was obtained at or near the holiday region of the TSA-coated or the TSA+Al-silicone sealed specimens.
A failure by cracking was recorded for uncoated DSS weld specimens after 14 days of exposure. Severe cracking on the middle of the specimen (weld area) was observed and these initiated from the edge of the specimen (Fig.3a). The cracking was mainly in the ferrite phase and either transgranular or around the austenite phase.
After 23 days, no cracking was observed in the TSA coated specimens (Fig.3b). The TSA+epoxy painted specimens, however, showed signs of degradation of the painted layer.
Figure 3: Transverse cross sections of
(a) uncoated DSS and (b) TSA coated DSS after chloride SCC tests.
HISC at 575 MPa tensile stress: None of the specimens failed during 165 days of exposure at 575 MPa (100% of 0.2% proof stress). The loading regime is shown in Fig.4.
Figure 4: Loading sequence used in the HISC test.
A visual examination of the surface of the uncoated DSS specimen (with applied cathodic protection (CP) potential at -1100 mVSCE) show a uniform layer (about 20-30 µm thickness) of a calcareous deposit on the entire surface of the test specimen and this film has been detached at several locations (Fig.5a). EDX spectra collected from the calcareous layer showed a two-layer structure- an inner layer rich in Mg and O and an outer layer comprised primarily of Ca and O. Detailed examination of the cross-section revealed very fine cracks in the uncoated DSS specimen. These fine cracks are the initial stages of hydrogen induced stress cracking.
The TSA-coated and the TSA+Al-silicone sealed specimens show white corrosion products on the coating surface, but no visible cracks. The TSA layer has retained a good bond with the DSS substrate (Fig.5b). No visible crack was found in this specimen after 165 days of exposure at a tensile stress of 575 MPa (proof stress) as any crack initially formed was filled by corrosion product when exposed to seawater (Fig.5b). Similar observations were also found in TSA+Al-silicone sealed specimens.
Figure 5: SEM images of cross sections of
(a) uncoated and (b) TSA coated DSS after HISC testing at 575 MPa.
HISC at 669-695 MPa tensile stress: The uncoated DSS specimen failed within 2 hours of increasing load to 669 MPa and showed a primary and several secondary hydrogen stress cracks (Fig.6a). This image also shows some evidence of calcareous deposit on the DSS surface. This, however, was very thin when compared to the TSA-coated specimens.
The TSA-coated specimen and the TSA+Al-silicone sealed specimens were subjected to a higher tensile stress of 695 MPa for additional 48 days. No failure of the specimens was observed and these specimens were unloaded and photographed (Fig.6b and 6c). These images show that at high tensile stress levels (typically 695 MPa), the coatings suffer from severe cracking. Both specimens show some white corrosion product on the surface. Some black globules were also seen on the surface of the TSA+Al-silicone sealed specimen. EDX analyses on these particles indicated the presence of Ca, O and C.
Figure 6: Photographs of HISC specimens after testing showing
(a) uncoated, (b) TSA coated; and (c) TSA + Al-silicone sealed DSS. The uncoated DSS specimen failed after 24 hours at 669 MPa while the coated specimens survived 235 days at 695 MPa.
Cross-sectioning of the TSA-coated specimens subjected to 695 MPa load showed extensive cracking and opening of the TSA coating which would allow the seawater to ingress onto the underlying DSS surface. This may be the cause of fine cracks observed on the DSS surface immediately under the TSA coating where the coating had cracked and opened widely (Fig.7b). TSA+Al-silicone sealed specimen showed similar features.
Figure 7: SEM image of (a) cross section of a TSA coated specimen after 235 days of exposure. The stress level was increased to 695 MPa (120% of proof stress) after 187 days at 575 MPa (proof stress). This is the same specimen shown in Fig.6b.
General corrosion mechanism
Exposed DSS surfaces of TSA-coated specimens did not exhibit rusting when exposed to artificial seawater. This is not unexpected as the TSA coating specimen was galvanically coupled to the DSS weld specimen. Once the corrosion product begins to form on the coating surface and a calcareous deposit begins to precipitate on the DSS cathode, the corrosion rate decreases and stabilises after about 15 days to a very low value. The protective mechanism offered by TSA to exposed steel in seawater is one of cathodic protection followed by calcareous deposit formation.
The morphology of the calcareous deposit is temperature dependent. At low temperatures (<5°C) Mg-rich deposit is expected but at higher temperatures Ca-rich deposits are formed due to the difference in the temperature dependence of the solubility limits of Mg(OH)2 and CaCO3 (Ref ). Solubility of CaCO3 in water decreases with increasing temperatures while that of Mg(OH)2 increases. The observation that the initial deposits formed at ~18°C were Mg-rich followed by a Ca-rich deposit has been reported previously and is related to the degree of polarization and the kinetics of precipitation reaction (Ref [2-5]).
Figure 8: Mechanism of calcareous deposit formation on a cathodically polarized steel surface in seawater.
In seawater, Ca and Mg salts are present along with salts of Na, K and many other cations. For precipitation of Mg(OH)2 or CaCO3 or both, supersaturation is essential. Supersaturation of Mg(OH)2 can arise near the steel surface as a consequence of the production of OH- at the cathodic sites. However, precipitation kinetics for CaCO3 are slower than for Mg(OH)2 due to inhibition influence of Mg2+ (and other anions present in sea water) on aragonite (CaCO3) nucleation and on nucleation and growth of calcite (CaCO3). At 25°C, the acidity/alkalinity in the vicinity of cathodically polarised steel surfaces in sea water has been reported to be ~pH9.5 (Ref ). Another factor to note is the isoelectric point (IEP) of the two precipitating compounds, namely CaCO3 (IEP~9) and Mg(OH)2 (IEP~11) (Ref ). The isoelectric point represents the pH below which the particles (of a compound) are positively charged and above which they are negatively charged. Thus, in the vicinity of a cathodically polarised surface (interface between steel surface in the scribed region and seawater) where the acidity/alkalinity is likely to be pH9.5, supersaturation of Mg(OH)2 will result in its precipitation. At this pH, which is >IEP of CaCO3 and <IEP of Mg(OH)2, the CaCO3 particles are negatively charged and thus would not precipitate on the negatively charged (polarized) steel surface. The pH should decrease with distance, and supersaturation of Mg(OH)2 should occur only within a certain distance from the metal surface. Beyond this distance, the sea water pH is lower than the IEP of CaCO3 and is supersaturated with it. Observations in this work at ~18°C that the calcareous deposits formed on cathodically polarized DSS were Mg-rich close to the steel surface and Ca-rich away from it are consistent with this theory.
The uncoated DSS weld specimens displayed a number of small corrosion pits after 30 days of immersion in aerated artificial seawater (~80°C). These pits may grow further during a longer-term exposure. The pits were located on the surface away from the weld. Chemical analysis confirmed a higher amount of Ni, and slightly higher amounts of Cr and Mo in the weld than the parent metal. This may explain why corrosion was not observed on the weld area and had occurred preferentially on the parent metal.
Exposed areas of the DSS specimens coated with TSA and the TSA+Al-silicone sealant did not exhibit any localised or general surface attack. Ecorr values of the TSA-coated DSS and the TSA+ Al-silicone sealed DSS (even when 5% substrate is exposed) were around -900 mVSCE. This potential value is sufficiently negative to cathodically protect an exposed DSS and is in line with a protection potential of -850 to -1050 mV Ag/AgCl that has been specified for carbon steels by standards such as NACE, NORSOK and DNV (Ref [6-8]).
Uncoated weld specimens failed in less than 14 days in an intermittent hot chloride environment. The results from the TSA-coated and TSA+epoxy painted specimens were very encouraging, without failure after a minimum of 23 days when tested in the same environment. It is noteworthy that these tests were conducted with a large scribe mark in the coating system to expose the weld metal. Although, after exposure, most of the painted specimens failed by rapid degradation, the TSA layer was in very good condition and had retained its full thickness. It is clear that this epoxy phenolic paint will not survive for a long time under these severe conditions.
The reaction mechanism of TSA coated DSS in chloride SCC environment is believed to be in three stages:
- During initial wetting of the hot DSS surface with cold seawater, the exposed DSS surface is quickly polarized by the adjacent TSA coating and provides effective cathodic protection.
- This results in the formation of a calcareous deposit on the exposed DSS (cathode) surface by mechanisms discussed earlier. On the TSA (anode) surface, a corrosion product of aluminum oxide is formed.
- The calcareous film at this elevated temperature (~130°C) will quickly dry off and will minimize the exposure of the weld (cathode) from further wetting and rapid thermal shocks, whereas the aluminum oxide based corrosion product on the TSA surface will retard the rapid consumption of TSA.
This study was carried out on a worst case basis i.e. by scribing the TSA coating layer down to the substrate. In reality (properly coated components), even if the coating is scratched or damaged there will be some coating left on the grit blasted surface which would help to build up this protective corrosion product layer of aluminium oxide and will extend the life of a component.
Under external CP of -1100 mVSCE, a calcareous film of about 15-25µm thickness was formed on the uncoated DSS specimen. This film had a poor adhesion to the DSS substrate, even on a grit blasted surface. Under a load of 575 MPa (100% of the proof stress), this calcareous film was found to have ruptured and detached from various locations of the exposed gauge length. This would allow the seawater to ingress onto the DSS cathode surface where hydrogen would generate. Although, at 575 MPa the uncoated DSS with CP did not fail after 165 days of exposure, several fine cracks were found to have initiated (Fig.5a). These cracks could further propagate over longer-term exposure.
The role of TSA or TSA+Al-silicone was to act as an effective barrier to minimize ingress of seawater into the underlying DSS surface. Although hydrogen may have generated during the first few weeks of exposure due to ingress of seawater onto the DSS surface via through thickness cracks in the coating (formed during plastic strain), these were repaired and completely filled with aluminum oxide based corrosion product. The formation of a calcareous deposit in these large fresh cracks occurred quickly and this would have minimized the continued exposure of the DSS surface to seawater. However, such high global stress levels are unlikely in reality and the coating might be expected to work as an effective barrier even when the local stress in the DSS material (e.g. at defects, weld toe stress concentrators and with residual stress areas) is high.
The steady free corrosion potentials of the TSA and the TSA+ Al-silicone sealed DSS specimens were shown to be around -1000 mVSCE and -900 mVSCE after about 120 days of immersion. This would suggest that the TSA and the TSA+Al-silicone would reduce the risk to HISC compared to the external CP of -1100 mVSCE over the longer exposures. However, at excessively high stress levels, e.g. above 695 MPa in the early days of immersion, or when a DSS weld (holiday) is directly exposed, the cathode surface will be rapidly polarised to below -1100 mVSCE by the TSA coating. In such conditions, the use of the TSA coating alone can only be marginally beneficial as compared to an external -1100 mVSCE CP. The use of Al-silicone sealant on the TSA coating was shown to attain a more positive Ecorr during the early days of immersion and over the long-term immersion in artificial seawater, with the steady potential being around -900 mVSCE. This could be beneficial when HISC is an issue.
Summary and conclusions
- A free corrosion rate of TSA after 25 days of exposure in circulated and aerated artificial seawater was measured at about 5-8 µm/year at 18°C and 6-7 µm/year at 80°C using both LPR and ZRA techniques. Similar corrosion rates were observed for Al-silicone sealed coatings.
- Pitting corrosion initiated on the uncoated DSS weld specimens within 30 days of exposure in aerated artificial seawater at ~80°C. TSA and TSA+Al-silicone sealed coatings protected the DSS from pitting corrosion. Ecorr of the TSA and the TSA+ Al-silicone sealed coatings in aerated artificial seawater at ~80°C was around -900 mVSCE as compared to -100 to -150 mVSCE of the uncoated DSS. Ecorr of the coated systems was sufficiently negative and would cathodically protect about 5% area of exposed DSS.
- While the uncoated DSS specimens failed within 14 days the TSA-coated specimens survived chloride SCC environment for >23 days.
- Uncoated DSS under external CP of -1100 mVSCE and at a constant load of 575 MPa (100% of the proof stress) in artificial seawater did not fail after 165 days of exposure but had fine hydrogen stress cracks. At a higher stress level (~669 MPa or 116% of the proof stress) the uncoated DSS failed within a further 24 hours of exposure.
- In artificial seawater at a constant load of 575 MPa (100% of the proof stress) TSA coated and TSA+Al-silicone sealed DSS did not fail and had no signs of hydrogen stress cracking after 165 days of exposure.
- TSA coated and TSA+ Al-silicone sealed DSS did not fail after increasing the stress to 695 MPa (~120% of the proof stress) for a further 48 days of exposure. However, at this stress level, the TSA and the TSA+ Al-silicone sealed coatings experienced extensive cracking and exposed the DSS surface under open cracks in the coating. This resulted in fine hydrogen stress cracks in the underlying DSS adjacent to larger cracks in the coating.
This work was funded by the Sponsors of TWI Joint Industry Project 14661: Chevron, Exxon Mobil, Health and Safety Executive, Marathon Oil, Petrobras, Shell UK and Technip Offshore UK.
- S. Shrestha and C-M Lee, 'Thermally Sprayed Aluminium Coatings for Preventation of Corrosion of Duplex Stainless Steel at Elevated Temperature', TWI Report 14661/8/05, 2005, p 674-679.
- M.M. Kunjapur, W.H. Hartt, and S.W. Smith, Influence of Temperature and Exposure Time upon Calcareous Deposits, Corrosion, 1987, 43(11), p 674-679.
- S-H Lin and S.C. Dexter, Effects of Temperature and Magnesium ions on Calcareous Deposition, Corrosion, 1998, 44, p 615-622.
- K. Akamine and I. Kashiki, Corrosion Protection of Steel by Calcareous Electrodeposition in Seawater: Part 1- Mechanism of Electrodeposition, Eng. Rev., 2003, 36, p 636-642
- J.S. Luo, R.U. Lee, T.Y. Chen, W.H. Hartt, and S.W. Smith, Formation of Calcareous Deposits under Different Modes of Cathodic Polarization, Corrosion, 1991, 47, p 189-196.
- R. Cottis and S. Turgoose, Electrochemical Impedance and Noise, NACE International, Houston, U.S.A., 2003
- 'Cathodic Protection Design,' RP B401, DNV Standards, DNV, 1993.
- 'Cathodic Protection,' M-503, NORSOK Standard, Standards Norway, 2007.