Demagnetisation by crossed fields in superconductors

Abstract

A study has been made of the decay of the trapped magnetisation in superconductors when exposed to a crossed field. Numerical results have been compared with the theory of Brandt and Mikitik (4) which solves the problem for a thin strip superconductor. FlexPDE with the A formulation and COMSOL with the H formulation were both used. Simulations of a strip with a cross section aspect ratio of 20 showed good agreement with theory both for the case of a transverse field larger than the transverse penetration field and for one smaller. In the latter case the magnetisation saturates as predicted, however the simulations show a slow decay after many cycles. In the case of stacked YBCO tapes the movement of flux lines is very small and the effects of the reversible motion were investigated. This can decrease the decay initially for very thin decoupled tapes, but cause a steady decay after very large numbers of cycles. Simulations on stacked strips showed that the decay constant increased approximately linearly with the number of strips. When combined with the theory for one tape this can explain the very slow decay observed in previous experiments. Experimental results were qualitatively in agreement with theory and simulations but showed some discrepancies. However there are a number of differences between the experimental situation and theory so good agreement is not expected. IntroductionIn most applications of bulk superconductors using trapped fields the bulk will be exposed to external fields which will change in both angle and magnitude. While changes in in magnitude are easily incorporated in the Bean model, changes in angle lead to much more complex situations, in many cases involving flux cutting and force free configurations. However many experimental measurements have shown that in general trapped fields are always reduced by the application of crossed fields and this effect may limit the application of bulk superconductors in electrical machines. Most of the experiments were at relatively high fields compared with the crossed fields and directed at elucidating the physics, rather than practical applications of bulk superconductors in motors and generators. In this paper we consider geometries in which the induced currents are mostly perpendicular to the fields so that the standard Bean model can be applied. One of the first observations was that of Sychev et. al. (1), who applied an oscillating field to a wire carrying a current and observed a DC voltage. Sakamoto et. al. (2) used a thin film carrying a current in an applied field and measured the decay of the trapped magnetisation when a transverse field was applied, as in this paper. They showed that the flux lines pivoted about two points alternately and 'walked' out of the film, a bit like penguins heading for their breeding grounds in the Antarctic. They proposed an approximate theory which assumed a uniform current density in the film. The DC voltage generated by an AC field was termed the 'dynamic resistance' and the clearest explanation of t...

Cross-field demagnetization of stacks of tapes: 3D modelling and measurements

“…with t i and t f being the time at the beginning and end of the demagnetization process, respectively. For our configuration, which is similar to that in other works [5,24,25,28], the error regarding this quantity is more strict than for the trapped field (table II) in errors below 20 % for ripple fields of 100 mT or below.…”

“…The ). This process can be explained by the Bean model of the infinite thin strip [27] or other numerical 2D modelling [5,25]. In our stack, the penetration front contains both the penetrates from the top and bottom of each individual tape.…”

“…Recent measurements of stacks showed the main behaviour under cross and rotating magnetic fields [5,[24][25][26], the latter reporting measurements for up to 10 000 cycles. However, theoretical study by cross-sectional approximation (infinitely long 2D approach) showed qualitative agreement only [5,25]. In addition, [25] compares an A-formulation with Brandt and Mikitik theory [27], showing good agreement.…”

Vortex shaking study of REBCO tape with consideration of anisotropic characteristics

“…This phenomenon is mainly due to the property of the power supply, which cannot produce very stable triangular waveform under high frequency of 1/10 Hz, leading to the unstable demagnetization of trapped magnetic field. These results agree well with the Brandt and Mikitik theory, which proves that the relaxation cycles have no relevance with the frequency of shaking magnetic field, as shown in Equation 9 [15,28]. Therefore, it is reasonable to simulate the vortex shaking process of crossed magnetic field with a higher frequency of 100 Hz, instead of the actual frequency of 1/30 Hz.…”

“…Besides these theoretical works, finite element methods (FEMs) based on Maxwell equation and E-J power law have also been widely used for simulating vortex shaking processes of superconductors [25][26][27][28]. The cross-magnetic-field effect on YBCO bulk was investigated by Vanderbemden et al [25] with both experiment and finite-element model based on H formulation and E-J power law.…”