Cavitation studies in materials: new insights from modern techniques in 2D/3D/4D characterisation

Cavitation is usually thought to be associated with the final stages of failure in structural materials. Since this view pervades the approach to component design properties and performance in service prior to these final stages have been emphasized with the result that, material developments to counter degradation by cavitation have often been neglected in favour of those concerned with microstructural strengthening. As a consequence, current understanding of the origins, proliferation and mechanisms of cavitation is inadequate to provide a robust theoretical framework to predict materials failure from cavitation related degradation. This lack has been perhaps most acutely felt in the development of ferritic–martensitic heat resistant steels for elevated temperature application. In efforts to extend the service temperature of these steels to 600°C and beyond, better knowledge of the factors responsible for the cavitation led embrittlement (a cumulative process present from incipient strains onwards) that leads to creep failure at low stresses would be indispensible.

The importance of this aspect has long been known to the scientific community; for example, the maiden theme of the inaugural issue of Materials Science and Technology’s predecessor Metal Science in 1967 was concerned with the physical origins of cavitation at grain boundary precipitates.¹ However, until recently tools to investigate damage phenomena in materials, to develop and confirm a robust theoretical framework, were lacking. In the past decade rapid development in the technique of micro-tomography has led to an explosive growth in its application to understand the mechanisms of damage and failure of diverse materials. A highlight of this development is the application of tomography to dense structural materials such as steels at large synchrotron sources (e.g. SPring-8, ESRF and APS). Coupled with new destructive techniques capable of volumetric microstructural characterisation, such as serial sectioning, this has contributed to the rise of the 3D/4D paradigm that provides a new approach to complement structure–property correlation in understanding the mechanical behaviour of materials. This approach is being increasingly applied to understand cavitation behaviour in materials by interrogating the bulk in a multi-dimensional, multi-modal manner. The present special issue of MST highlights the application of these 3D/4D techniques to provide new insights on the cavitation behaviour of structural materials, through a set of 12 invited peer reviewed papers.

The issue of creep naturally finds prominence in the several papers in this special issue. Type IV failure is a pernicious problem during creep of weldments² in 9–12%Cr steels and efforts to design microstructures³ to mitigate this problem have led to the development of an experimental grade MARBN steel, on the basis that reducing the size of the fine grained zone in the heat affected zone would enhance resistance to this form of embrittlement. Although on a macroscale this microstructure was achieved, combined synchrotron studies along with complementary EBSD investigations by Schlacher et al.⁴ provided the insight that individual fine grains at the local scale can act as sites for intense cavitation, promoting failure at stresses below 100 MPa. This work underscores the need for investigating microstructural damage in a multi-scale, multi-modal manner.

However, such characterisation is highly challenging and a set of techniques/protocols needs to be developed to obtain and combine volumetric information. One such example is the newly developed technique of correlative tomography applied by Burnett et al.,⁵ alongside complementary small angle neutron scattering (SANS) studies by Jazaeri et al.⁶ on an austenitic stainless steel steam header removed from service (65 000 h, 16 MPa internal pressure, 525°C) near a region at the mouth of a large reheat crack. While SANS established that the crack mouth region rather than its tip contains extensive cavities, up to about 350 nm in size, correlative tomography was able to associate these cavities with second phases (M₂₃C₆, sigma and G phase) and to establish the chemical composition of stringers. On the basis of multi-scale information on cavitation and second phases, an insightful discussion on the role of second phases, grain boundaries and residual stress in the development of creep voids is provided.

A missing factor in such an analysis is the information on grain misorientation available from EBSD characterisation. Two papers address this aspect. The work of Abbasi et al.⁷ is of significance in providing a unique example of combining the power of serial sectioning over a large volume with information from EBSD to classify grain boundaries in copper in terms of their propensity for creep cavitation. In contrast to the 3D approach, Yardley et al.⁸ provide a detailed exposition and application of a new automated method to map creep damage in 2D onto the crystallographic hierarchy, quantified by EBSD, of the tempered martensitic structure of heat resistant steels. These destructive techniques capable of volumetric characterisation of damage principally have resolution at the micrometre length scale. To capture the nature of damage at sub-micrometre length scales, FIB serial sectioning has emerged as an appropriate tool for 3D characterisation of creep cavities from the incipient stages close to nucleation. It is believed that at low stresses, damage in tempered martensitic steels begins at early stages of creep and hence a tool such as FIB serial sectioning could be applied to determine the point of explosion of cavitation in the tertiary stage. With these aims, Yadav et al.⁹ applied FIB serial sectioning to coupons extracted from interrupted creep specimens of P91 steel (60 MPa, 650°C for 9000 h). Small cavities, about 600 nm in size, could be distinguished in the microstructure of samples possibly undergoing viscous creep. Significantly the sample also contained pre-existing cavities (prior to creep) that rapidly increased in size, number and volume fraction as compared with the newly developed small cavities formed in the microstructure. Of late there has been a renewed interest in the role of entrained defects in structural materials on subsequent cavitation behaviour during service.¹⁰ The influence of the presence of prior damage generated in materials due to poor melting or casting practice on the propensity to cavitation led failures is reviewed by Campbell,¹¹ emphasising the role of oxide bi-films.

Apart from its non-destructive nature, tomography is unique in its capability to provide 4D information that can reveal the kinetics of damage proliferation in materials. Metal matrix composites have been extensively investigated using micro-tomography, especially via in situ experiments. The bulk imaging of these advanced structural materials typically reveals damage arising from factors such as heating and cooling during processing or temperature excursions during service that may not be revealed at the surface. Chapman et al.¹² investigated the damage due to thermal cycling of a SiC reinforced AA2080 aluminium matrix composite using micro-tomography to detect and quantify the growth of cavitaties developed at the particle/matrix interface due to the difference in their expansion coefficients. These weak interface properties have been improved using metallic glass as a reinforcing phase rather than alumina or SiC. The paper by Ferre et al.¹³ explores damage evolution in an Al+AA5083 matrix containing Zr based bulk metallic glass particles, formed by two processing routes. Sket et al.¹⁴ attempt to provide a new insight into the deformation and damage mechanisms of carbon fibre reinforced polymers via in situ studies.

Cavitation in the form of controlled porosity can also be a beneficial rather than a deleterious phenomenon, as perhaps best illustrated by its functional role in bone implants. Porous super-elastic metals such as NiTi for bone implant applications combine functional and structural requirements by their specific architecture of interconnected cavities. The paper by Gupta et al.¹⁵ provides a quantitative assessment of the interconnected porosity using a new connectivity–tortuosity index, evaluated from tomographic scans, in a porous NiTi alloy fabricated by self-propagating high temperature synthesis. The concluding review by Gupta et al.¹⁶ seeks to examine critically the potential of the various 3D techniques, principally based on tomography, to reduce the empiricism associated with life assessment of heat resistant steels exposed to creep phenomena. Directions for future research to realise the potential of tomography for residual life assessment are outlined.

The guest editors would like to express their heartfelt thanks to all the authors and referees of the papers that constitute this special issue. C. Gupta, on behalf of the guest editors, wishes to expresses his sincere gratitude to Professor J. F. Knott, then Editor of MST, for his encouragement and positive support to the proposal mooted for the special issue. The support from Maney Publishing and tremendously affirmative efforts of Mark Hull and Rose Worrell are gratefully acknowledged.

Dr C. Gupta, Professor H. Toda, Professor P. Mayr, Professor C. Sommitsch

Guest Editors