In this dissertation, we investigate the macroscopic magnetic properties of bulk high-temperature superconductors (HTS) containing an array of artificial holes in view of enhancing their performances. The study involves a numerical modelling part and an experimental characterization part. In each part, novel concepts are highlighted and detailed. In particular, we develop a three-dimensional finite-element method (FEM) for calculating the magnetic field penetration in HTS where a single time-step is used in the case of a linearly varying applied magnetic field, and we probe the magnetic field in the volume of drilled samples with the help of microcoils inserted inside the holes.
The thesis starts with an introductory chapter that describes the general concept of high-temperature superconductivity and particularly draws the attention on the interests and on the synthesis of drilled structures. Then, we detail the modelling tools that are used for evaluating the magnetic properties of drilled samples. Three models are taken into account: (1) the numerical Bean model which is a generalization of the Bean model for arbitrary cross sections where the samples are assumed to have an infinite height; (2) a 2D finite element model implemented in the open source solver GetDP for samples with an infinite height and assuming a power law relationship, that is characterized by a critical exponent n, between the electric field, E, and the current density, J; (3) a 3D finite element model with the same equations as those of model (2), but where these are solved in a three-dimensional sample with a finite height. For large values of n, both FEM models use the properties of a slow magnetic diffusion to reduce the number of time steps. In particular, the trapped flux can be calculated with only two time-steps: during the first step, the applied magnetic flux density is increased with a constant sweep rate to a maximum value, it then decreases to zero with the same sweep rate during the second step.
The models are first used in simple geometries where they are compared to other available techniques. These are next applied to drilled samples. A systematic numerical study of the influence of the holes on the magnetic properties of the sample is reported. A single hole perturbs the critical current flow over an extended region that is bounded by a discontinuity line, where the direction of the current density changes abruptly. In samples with several holes and a given critical current density, we demonstrate that the trapped magnetic flux is maximized when the centre of each hole is positioned on one of the discontinuity lines produced by the neighbouring holes. For a cylindrical sample, we construct a polar triangular hole pattern that exploits this principle; in such a lattice, the trapped field is 20% higher than in a squared lattice, for which the holes do not lie on discontinuity lines. These results are experimentally validated. Two parallelepipedic samples are drilled with two different hole lattices. The trapped magnetic flux density of these samples is characterized by a Hall probe mapping before and after drilling holes. The sample in which the holes are aligned on the discontinuity lines exhibits the smallest magnetization drop that results from the hole drilling.
Then, we resort to a novel experimental technique using microcoils inside the holes to characterize the local magnetic properties in the volume of drilled samples. In a given hole, three different penetration regimes can be observed when the sample is subjected to an AC magnetic field: (i) the shielded regime, where no magnetic flux threads the hole; (ii) the gradual penetration regime, where the amplitude of the magnetic field scales with the applied field; and (iii) the flux concentration regime, where the magnetic field exceeds that of the applied field. A comparison of the measurements with simple models assuming an infinite height shows that the holes may serve as a return path for the demagnetizing field lines. In the case of a pulsed field excitation, that measurement technique also allows us to estimate the trapped magnetic flux density in the volume of the sample and compare it with that on the surfaces. Moreover, the penetration of a magnetic pulse from hole to hole is described in the median plane and on the surface and the differences of penetration speeds are explained.
Finally, we investigate the magnetic properties of drilled samples whose holes are filled with a ferromagnetic powder. To this aim, we use experimental techniques (Hall probe mapping techniques, together with measurements of the volume magnetization and of the levitation force between the HTS sample and a permanent magnet) and a numerical model (3D FEM) to characterize the modification of the magnetic properties resulting from the impregnation of the holes with AISI 410 ferromagnetic powder. Numerical results support the experimental observations and give clues to understand the mutual interaction between the HTS sample and the ferromagnetic powder inserted in its holes. In particular, the Hall probe mappings of the distribution of the trapped flux above the non-impregnated and impregnated samples reveal an increase of trapped flux after impregnation that is confirmed by simulations.