The oxidation behaviour of metals and alloys at high temperatures in atmospheres containing water vapour: A review

High temperature oxidation of Fe13CrxAl alloys in vapour mixtures

A hot-stage, environmental scanning electron microscope has been used to observe the in situ development of oxide whiskers, pyramids, and pits in the oxidation of copper and nickel at elevated temperatures. The effects of oxidation temperature, metal deformation, and the presence of water vapor on these irregular oxidation features were studied. In each case, the feature results from the presence of a central screw dislocation which provides ledges for the extension of the oxide lattice, but the specific geometries are decided by factors such as surface diffusion along the dislocation core, the rate of the molecular dissociation step, and the balance of surface energy and dislocation line tension forces.

In situ observation of whiskers, pyramids and pits during the high-temperature oxidation of metals

Oxidation studies were performed at 1100 °C in dry air and air containing fixed partial pressures of water vapor on a number of alloys and coatings that form α-Al2O3 scales under oxidizing conditions. The alloys investigated included RENE N5, PWA 1484, diffusion aluminide coatings (with and without Pt modification) on RENE N5, and a Ni-8 wt pct Cr-6 wt pct Al model alloy. The water vapor affected the oxidation of the alloys in three important ways: (1) The scales spalled more profusely during cyclic oxidation in wet air than in dry air, particularly for those alloys with alumina scales, which are only moderately adherent under dry conditions. The results were consistent with the mechanism previously proposed (Reference 1), whereby the water molecules decrease the fracture toughness of the alumina/alloy interface. (2) Thicker oxides are formed during oxidation in wet air than dry air. This effect comes primarily from accelerated transient oxidation during exposure in wet air. (3) Spinel was found to form on top of the alumina scales during long-term exposure. This phenomenon occurred in all atmospheres but was much more pronounced for exposures in wet atmospheres. Mechanisms for the preceding observations are proposed.

Some water vapor effects during the oxidation of alloys that are α-Al2O3 formers

Cyclic oxidation tests have been performed at 1100 °C in wet and dry air on a number of alloys and coatings that form α-Al2O3 scales upon exposure to oxidizing conditions. The alloys that were investigated included PWA 1480, PWA 1484, CMSX 4, diffusion aluminide coatings on PWA 1480 and PWA 1484, and Co-24Cr-10.5Al-0.3Y. In cases where some cracking and spalling of the alumina scales occurred in dry air, the presence of water vapor caused the degradation rate to be increased by a factor of 2. When no cracking or spalling of the alumina occurred in dry air, as was the case for low sulfur alloys, water vapor had no effect on the oxidation behavior. It is proposed that water vapor causes stress corrosion cracking at the Al2O3-alloy interface during cyclic oxidation.

The Effect of Water Vapor on the Oxidation of Alloys that Develop Alumina Scales for Protection

The oxidation of iron at 950°C in oxygen/water vapour mixtures

Pure iron, low carbon steel, medium carbon Mn-steel and MnMo-steel were studied in view of their high temperature corrosion behaviour in various gases. The gases used were: Oxygen, steam, carbon dioxide, air and five synthetic atmospheres, similar to those obtained by the combustion of sulphur-free fuel oil and town gas. In order to obtain possible characteristic differences of behaviour isothermal experiments were made as well as experiments with temperature cyling. In the first case temperatures were 700 and 1200°C while in the second case temperatures were cycled between 500 and 1200°C. There are different degrees of agreement between the two experimental techniques, but in most cases it is possible to predict scaling behaviour for either type of experiment. It is possible to establish an “equivalent time” as a reference parameter for the comparison of temperature cycles from different furnaces.



Wachstum von Zunderschichten wahrend Warmebehandlungszyklen



Reines Eisen, niedrig gekohlter Stahl, mittelgekohlter Mn-Stahl und ein MnMo-Stahl wurden unter Verwendung verschiedener Gase auf ihr Verhalten bei Hochtemperaturkorrosion untersucht. Als Gase wurden verwendet: Sauerstoff, Wasserdampf, Kohlendioxid, Lug sowie funf synthetische Atmospharen ahnlich den Gasgemischen, die bei der Verbrennung von schwefelfreiem Heizol und Stadtgas entstehen. Um charakteristische Unterschiede herauszuarbeiten, wurden isotherme Versuche ebenso wie Versuche mit zyklischer Temperaturfuhrung gefahren. Im ersten Falle lagen die Temperaturen zwischen 700 und 2200°C, im zweiten Falle schwankten die Temperaturen zwischen 500 und 1200°C. Die Ubereinstimmung zwischen den beiden Versuchsfuhrungen ist unterschiedlich, doch kann man das Verzunderungsverhalten in den meisten Fallen fur beide Modifikationen voraussagen. Dabei kann auch als Bezugsparameter eine „Aquivalenzzeit” bestimmt werden, welche den Vergleich der Temperaturzyklen verschiedener Ofen er-moglicht.

Scale Growth during Re-heating Cycles

Die Versuche wurden durchgefuhrt in Sauerstoff, Luft, CO2 und Wasserdampf sowie in Luft mit bis 15% Wasserdampf; die Versuchsdauer betrug bis 200 min. Fur die Oxydation von Reineisen und Stahl in Sauerstoff und Luft ergab sich ein parabolisches Zeitgesetz, das jedoch bei langerer Versuchsdauer unter dem der idealen Parabel entsprechenden Wert sank. In Wasserdampf wurde Reineisen bei 850° C linear oxydiert, bei hoheren Temperaturen lag das Zeitgesetz zwischen einer Geraden und einer Parabel. Der Weichstahl verhielt sich ahnlich, doch war hier die Abweichung bei hoheren Temperaturen weniger ausgepragt. In CO2 ergab sich fur beide Probenarten ein lineares Zeitgesetz. Ganz allgemein wurde Reineisen doppelt so schnell oxydiert wie Stahl; die Oxydationsgeschwindigkeit nahm in der Reihenfolge Sauerstoff — Wasserdampf — Kohlendioxid ab.



Bei parabolischer Oxydation bestand der Zunder immer aus drei Schichten, bei linearer Oxydation nur aus einer (Wustit).



In Luft mit Wasserdampf ergab sich immer ein parabolischer Verlauf, ohne das ein deutlicher Einflus des Wasserdampfzusatzes (2,5 bis 15 % zu erkennen gewesen ware.



Scaling rate of pure iron and mild steel in oxygen, water vapour and carbon dioxide at temperatures ranging from 850 to 1000° C



The tests were carried out in oxygen, air, CO2 and water vapour as well as in air with up to 15 per cent water vapour. The duration of the tests was up to 200 minutes. The oxidation of pure iron and steel in oxygen and air, as a function of time, was found to follow a parabolic law; with tests of long duration, however, the results tended to be lower then those corresponding to the ideal parabola. In water vapour, pure iron was found to oxidize at 850° C at a linear rate; at higher temperatures, the curve followed an alignment between straight line and parabola. Mild steel showed similar characteristics except that the deviation at higher temperatures was less marked. In CO2 the curve was linear with both samples. Generally speaking, the oxidation rate of pure iron was found to be twice as high as that of steel. The oxidation rate decreased in the sequence: oxygen, water vapour, carbon dioxide.



With parabolic oxidation, the scale always consisted of three layers; with linear oxidation, it consisted of one layer only (Wustit).



In air with water vapour content, the curve was parabolic in all cases although no distinct influence of the water vapour addition 2.5 to 15 per cent.) was recognizable.

Die Verzunderungsgeschwindigkeit von reinem Eisen und Weichstahl in Sauerstoff, Wasserdampf und Kohlendioxid bei Temperaturen zwischen 850 und 1000° C

The work described in this paper is part of a current programme that has two objects: (1) to investigate further the reasons for the different scaling behaviour of steel in steam and carbon dioxide, although these gases have similar oxygen potentials; (2) to provide background information for an investigation into the effect of variations in re‐heating furnace atmospheres upon scaling and scale adhesion.

SCALING RATES of pure iron and mild steel in oxygen, steam, carbon dioxide in the range 850°–1000°C

Oberflachenfehler infolge eingewalzten Zunders konnen durch die Verzahnung zwischen Metall und Zunder und moglicherweise auch durch die Ansammlung von Fayalit an der Grenzflache bedingt sein. Beide Faktoren werden durch hohe Temperaturen, d. h. Temperaturen von 1200° C und daruber, begunstigt.

Im Falle niedriglegierter Stahle sollte die Ofenatmosphare stark oxydierend, fur Stahle mit nur „Rest-Nickelgehalten” hingegen schwach oxydierend eingestellt werden.

In den fruhen Stadien der Oxydation bleibt das Oxid in Kontakt mit dem Metall selbst, indem es „nachkriecht”.

Durch das nach innen gerichtete Kriechen der Zunderschicht wird in einer dunnen Schicht Nickel konzentriert. So bald das Kriechen durch Eindiffundieren von Sauerstoff ersetzt ist, bleiben die Metallteilchen raumlich fixiert wahrend die Grenzflache noch weiter in das Metall hineinwandert. Die Dicke der erweiterten Verzahnungszone ist ein Mas fur die durch Eindiffundieren von Sauerstoff entstandene Zundermenge.

Die Metallfilamente verlieren Eisen an das umgebende Oxid. Die Nickelanreicherung ist andererseits mit einer Volumenverringerung verbunden.

Geschmolzenes Fayalit-Eutektikum erleichtert das Kriechen der Oxidschicht an den vorstehenden Filamenten vorbei, so das der Kontakt zwischen Oxid und Metall dann erhalten bleibt.

K. Sachs

Papers

The oxidation of iron at 950°C in oxygen/water vapour mixtures

Scale Growth during Re-heating Cycles

Die Verzunderungsgeschwindigkeit von reinem Eisen und Weichstahl in Sauerstoff, Wasserdampf und Kohlendioxid bei Temperaturen zwischen 850 und 1000° C

SCALING RATES of pure iron and mild steel in oxygen, steam, carbon dioxide in the range 850°–1000°C

Die Verzahnung von Zunder und Grundmetall in niedriglegierten Stählen