Identifying high temperature oxides on steel

TWI Bulletin, July 1987

by David Sparkes

David Sparkes, BSc, is a Research Metallurgist in the Materials Department.

It is often useful to know at what stage in a manufacturing process a particular defect was introduced. The following article explains how the examination of oxide scale present on a metal surface can indicate the thermal historyexperienced by a piece of steel, and this information is vital when investigating a component's structural integrity. For example, it may be essential in a failure analysis to determine whether or not a defect existed prior to thecomponent entering service. In principle, valuable information can be gained, at least for post-weld heat treated materials, from studying the oxide scale on the material surface, together with assessment of any decarburisation.

Oxidation behaviour of steel

A portion of the iron-oxygen equilibrium phase diagram is shown in Fig.1, and it can be seen that, depending on the temperature and oxygen content, three separate oxides can be formed.

Scale always forms with the lowest oxide ( i.e. least oxygen content) next to the metal surface, and thus increases in total oxygen content with increasing thickness. From the equilibrium diagram ( Fig.1), it is clear that at temperatures below 570°C only hematite and magnetite are stable. ^[1] It has been found that at temperatures less than 250°C only hematite will form, and between 300 and 570°C the two oxides are present as a duplex film. Although wustite can form above 570°C,its growth is slow and only a discontinuous layer is present until 700°C is reached; between 570-650°C the wustite is present only inside the magnetite layer. The relative proportions of the three oxides ( Fig.2) will depend on the time at temperature, except for temperatures in excess of 700°C when the ratio of the oxide thicknesses has been found to be constant (wustite: magnetite: hematite = 100: 5: 1). This behaviour issummarised in Fig.3.

The total thickness of oxide scale will depend on both the temperature and time of the thermal cycle experienced. Expressions have been produced for determining growth rates, and these have been used to construct diagrams such as in Fig.4, which assumes a parabolic growth relationship. For oxidation to proceed, the scale needs to remain in contact with the metal surface. Because the scale occupies a greater volume than the original metal, plastic deformation of the oxide scale takes place, which can cause it to fracture and become detached, especially at corners where there are competing deformation directions. This must be taken into account when interpreting the information given by the thickness of the oxide scale.

Alloying additions to steel tend to cause a slight decrease in oxidation rate, but the temperatures at which changes occur ( Fig.3) remain nominally unchanged. However, if high alloying levels are present, the oxides formed may not be simple iron oxides, so careful interpretation will be required. Moreover, during service, corrosion products may be formed which could lead to confusion if metallography alone is used to identify the scale constituents.

Identification of the oxides

It is possible to distinguish between the three oxides by optical metallography. If the specimen is etched in a 25% HCl solution, sufficient contrast can be developed to identify which oxides are present ( Fig.2). The use of such a strong etchant will tend to destroy the steel surface, so that any metallography on the parent material should be carried out prior to examination of the oxide scale. If sufficient scale is present microhardness can be used to back up the metallography procedure, as the individual oxides have different hardnesses, see Table. But to determine which oxide phases are present unambiguously it may be necessary to remove the scale and carry out X-ray diffraction.

Typical microhardness values for iron oxides at 50g load

Decarburisation

The presence of carbon in steel allows further information to be obtained about the thermal cycle, because at temperatures greater than approximately 700°C decarburisation will take place ( Fig.5), with the evolution of carbon monoxide at the metal surface. This will produce a porous oxide scale and can even cause it to break up, but decarburisation does give metallurgists another mark on the thermal scale ( Fig.3). The depth of decarburisation is diffusion controlled, and will therefore increase for higher temperatures and longer times, but it will also depend on the condition of the free surface which makes accurate interpretation of the thermal history via the depth of decarburisation difficult, unless additional experiments are performed.

Concluding remarks

Examination of oxide scale and depth of decarburisation should only be used to give an approximation to the thermal history experienced by a steel component. Unless further experimental work is carried out on a particular steel bearing in mind the conditions of service, the temperature can only be estimated to within about 100°C. This is because the individual oxidation characteristics will depend not only on the temperature and time of the heating cycle but also the level of alloying and the atmosphere in contact with the metal surface.

Acknowledgements

This work is a summary of a survey of published data carried out by R F Lawn during an eight week project at The Welding Institute.

Reference