Got it cracked? – stress corrosion cracking

Materials World magazine
,
1 Aug 2009

Stress corrosion cracking is difficult to study and forecast. James Marrow, Reader in Physical Metallurgy at The University of Manchester, explores solutions.

Stress corrosion cracking may be one of the most insidious forms of metals degradation, being difficult to study and forecast. In the absence of a pre-existing defect, the incubation period from surface oxidation to corrosion pit then crack nucleus may occupy a significant fraction of the lifetime of a component.

Although sophisticated models have been developed based on current understanding, reliable prediction of crack nucleation, and hence the likelihood of achieving a certain lifetime, remains a challenge.

The processes involved are sensitive to the microstructure and environment. Quite subtle influences of thermo-mechanical processing and surface finish can have profound effects on the duration of the incubation period and the characteristics, such as size and number density, of the crack population. A consequence of this uncertainty may be higher degrees of conservatism in the design and operation of components whose degradation could be critical to safety. The economic consequences can be high, particularly in power generation.

Mixed monitoring

Researchers at the Materials Performance Centre in The University of Manchester School of Materials, UK, are concerned with the mechanisms of stress corrosion cracking in nuclear materials, particularly stainless steels as well as nickel-base and zirconium alloys. A multiscale approach is being taken, ranging from studies of corrosion/oxidation to interactions between factors such as processing, residual stress, deformation, damage and crack development.

The new synchrotron technique of diffraction contrast tomography has been combined with computed tomography to study intergranular stress corrosion cracking in sensitised austenitic stainless steel. Diffraction contrast tomography provides a 3D map of the shape and crystal orientation of each grain in a bulk sample. Computed tomography then gives three-dimensional images of the crack as it grows, using the attenuation of the X-rays.

Combining these observations reveals the grain boundaries that have high or low resistance to stress corrosion cracking, and also provides data to validate crystal plasticity finite element models for crack development. Such models can be constructed directly from the diffraction contrast tomography data. The insight into resistant grain boundaries, which tend to be low energy structures, supports the development of more resistant structures.

To study the crack initiation process and to measure short crack growth rates, it is necessary to make period observations. This may be done by suspending the test, removing the sample and examining it using microscopy tools.

However, for some forms of stress corrosion cracking, the sample must remain in the active environment throughout, as the kinetics of cracking are sensitive to the development of the crack tip chemistry, which may differ from the bulk environment. Interruption of testing for examination may therefore provide inaccurate data. In situ observation using optical tools is possible in ambient environments, but for high temperature and pressure autoclave testing, which is necessary to understand stress corrosion cracking in the energy industry, the tests are generally conducted in a sealed chamber.

Peer within

An ‘imaging autoclave’ has been developed at Manchester to enable in situ optical observation of high temperature aqueous corrosion. This facility applies digital image correlation strain mapping and Raman spectroscopy under realistic environments (refreshed loop, with slow strain rate, static and cyclic loading), at sufficient temperature and pressure (currently 250°C at 40 bar). This enables characterisation of material degradation in energy relevant materials such as stainless steels and zirconium alloys. The surface of the sample can be observed at high magnification through a transparent sapphire window.

Digital image correlation maps surface displacements by tracking features in successive images. It is well suited to the observation and characterisation of surface breaking defects, which may not be detected by conventional imaging. Measurement of the crack opening displacement, which is sensitive to crack shape and stress, also provides data on the rate at which cracks develop in depth (see image above, left). Raman spectroscopy, at temperature, may be used to study the oxidation processes leading to crack initiation.

Manchester aims to achieve full 3D observation of corrosion, crack initiation and growth under realistic conditions relevant to energy generation, combined with full 3D, high-resolution characterisation of the materials microstructure. Such insights will significantly strengthen our understanding of stress corrosion cracking, and our ability to predict and prevent its occurrence.

Further information: James Marrow