Abstract
The thesis presents results on two different aspects of flux penetration into superconductors. It is predominantly an experimental study using magneto-optical imaging, a microscopy technique for visualising magnetic fields, but also contains some simulation results. The experimental technique is explained in some detail.
The bulk of the thesis deals with dendritic avalanches in a ring- shaped thin film. These avalanches appear as beautiful, tree-like structures which form extremely fast: the advancing tip velocity has been measured to roughly 100 km/s. While they are fascinating and visually appealing objects, they are a significant problem from a practical point of view since they limit the current capacity of superconductors.
The ring geometry opens up for several new phenomena because of the characteristic way flux enters a superconducting ring. While positive flux enters at the outer edge, negative flux enters at the inner edge, but in such a way that the superconductor actually screens the hole. When avalanches nucleate at the outer edge, they may span the ring width and thereby inject flux into the central hole. The flux injection is studied in some detail in the thesis, and it is shown how the temperature in the core region of the avalanche may be inferred from measurements of flux changes in the hole: it is found that the maximum temperature is 100 K in typical cases.
The avalanches may also halt just before they reach the inner edge, and sometimes trigger anti-flux dendrites.
Single vortices are also observed as they enter a NbSe$_2$ sample. The observations show that there is a strong magnetic field at the sample edge where vortices nucleate, but thereafter they quickly move across a certain region leaving a vortex free band near the edge. The observed features can be explained qualitatively within the geometric barrier model.
Finally a surprising interaction between Bloch walls and vortices is discussed. Observations show that Bloch walls may attract vortices of opposite polarity, which seems a counter-intuitive behaviour if the Bloch wall is pictured as a simple bar magnet. However, by including the effect of the surrounding domains this conundrum is resolved. The current author only contributed interpretations of experimental results and the identification of the problem in this case.