In recently discovered self healing creep steels, open-volume creep cavities are filled by the precipitation of supersaturated solute. These creep cavities form on the grain boundaries oriented perpendicular to the applied stress. The presence of a free surface triggers a flux of solute from the matrix, over the grain boundaries towards the creep cavities. We studied the creep cavity filling by finite element modelling and found that the filling time critically depends on (i) the ratio of diffusivities in the grain boundary and the bulk, and (ii) on the ratio of the intercavity distance and the cavity size. For a relatively large intercavity spacing 3D transport is observed when the grain boundary and volume diffusivities are of a similar order of magnitude, while a 2D behaviour is observed when the grain boundary diffusivity is dominant. Instead when the intercavity distance is small, the transport behaviour tends to a 1D behaviour in all cases, as the amount of solute available in the grain boundary is insufficient. A phase diagram with the transition lines is constructed.@en

Advances in self-healing creep steels are driven by the understanding of its mechanism. Previous work on self-healing creep metals has shown selective precipitation inside the creep cavities, which has a beneficial effect on the creep lifetime. This effect occurswhen a solute supersaturation exists at the creep temperature. In the case of self-healing creep steels, the precipitation of gold was found to have an exceptional effect. Gold precipitation was found to be efficient because of the atomic size mismatch of gold and iron, which induces a high selectivity for precipitates to form inside creep cavities and not in the bulk. However, gold is not the most widely used alloying element in steel, to put it mildly. Ideally the self-healing effect could be achieved by another solute element dissolved in steel. In order to identify some other possible solute element for self healing creep steels, the mechanism must be understood. In this thesis the mechanism of self healing creep steels, andmetals in general, is investigated.@en

When metals are mechanically loaded at elevated temperatures for extended periods of time, creep damage will occur in the form of cavities at grain boundaries. In the present experiments it is demonstrated that in binary iron-tungsten alloys creep damage can be self healed by selective precipitation of a W-rich phase inside these cavities. Using synchrotron X-ray nano-tomography the simultaneous evolution of creep cavitation and precipitation is visualized in 3D images with a resolution down to 30 nm. The degree of filling by precipitation is analysed for a large collection of individual creep cavities. Two clearly different types of behaviour are observed for isolated and linked cavities, where the isolated cavities can be filled completely, while the linked cavities continue to grow. The demonstrated self-healing potential of tungsten in iron-based metal alloys provides a new perspective on the role of W in high-temperature creep-resistant steels.@en

The autonomous filling of creep-loading induced grain-boundary cavities by gold-rich precipitates at a temperature of 550 °C has been studied as a function of the applied load for Fe-Au alloys using synchrotron X-ray nano-tomography. The alloy serves as a model alloy for future self-healing creep resistant steels. The size, shape and spatial distribution of cavities and precipitates are analyzed quantitatively in 3D at a nanometer resolution scale. The filling ratios for individual cavities are determined and thus a map of the filling ratio evolution is obtained. It is found that the gold-rich precipitates only form at cavity surfaces and thereby repair the creep cavity. The shape of the cavities changes from equiaxed to planar crack like morphologies as the cavities grow. The time evolution of the filling ratio is explained by a simple model considering isolated cavities as well as linked cavities. The model predictions are in good agreement with the measurements.@en

A self-consistent model is applied to predict the creep cavity growth and strain rates in metals from the perspective of self-healing. In this model, the creep cavity growth rate is intricately linked to the strain rate. The self-healing process causes precipitates to grow inside creep cavities. Due to the Kirkendall effect, a diffusional flux of vacancies is induced in the direction away from the creep cavity during this selective self-healing precipitation. This process impedes the creep cavity growth. The critical stress for self-healing can be derived, and an analysis is made of the efficiency of self-healing elements in binary Fe–Cu, Fe–Au, Fe–Mo, and Fe–W alloys. Fe–Au is found to be the most efficient self-healing alloy. Fe–Mo and Fe–W alloys provide good alternatives that have the potential to be employed at high temperatures.@en

The diffusivities of substitutional impurity elements in iron have been computed with ab inito electronic density functional techniques, using exchange-correlation functional PW91. Excess entropies and the attempt frequency for a jump were determined by calculating phonon frequencies in the harmonic approximation. The influence of the degree of spontaneous magnetization on diffusivity is taken into account by means of the Girifalco model. The activation energy for diffusion has been determined by computing the vacancy formation energy, impurity-vacancy binding energies, migration barrier energies, and the effective energy associated with correlation of vacancy-mediated jump. For each type of impurity atom these contributions have been evaluated and analyzed up to and including the fifth nearest-neighbor shell of the impurity atom. It is found that impurities that have a low migration energy tend to have high effective energy associated with vacancy migration correlation, and vice versa, so that the total diffusion activation energies for all impurities are surprisingly close to each other. The strong effect of vacancy migration correlation is found to be associated with the high migration energy for iron self-diffusion, so that movement of vacancies through the iron bulk is in all cases, except cobalt, the limiting factor for impurity diffusion. The diffusivities calculated with the PW91 functional show good agreement with most of the experimental data for a wide range of elements.@en