Magnetization and critical current density of ultra-fine multifilamentary superconducting wires

Abstract

Multifilamentary superconducting wires with a greatly reduced level of losses have been produced with lengths of several tens of kilometers. In spite of the reduction of the filament diameter, proximity effects are avoided, and we make the best possible use of the reversible motion of the flux lines, so that the hysteretic losses are reduced. The concepts lead us to realize conductors comprising filaments of Nb-Ti, with a 0.1 to 0.2 pm diameter embedded in a highly resistive CuNi matrix. In order to characterise the possible applications to indusmal power systems, we should investigate in a quite precise way, the losses in submicronic filaments. We can determine the correct value of the critical current density with critical current measurements and magnetization curves on such wires.In order to promote an industrially competitive application in power systems, losses in multifilamentary superconductors must be lower than those in conventional conductors (1 watt at 4.2 K is equivalent to about 500 to loo0 watts at room temperature).An approximate value of the losses is given by the following expressionl. a I 3 wire dependent constants dF filament diameter P twist pitch Pt matrix resistivity Jov overall current density JC critical current density B magnetic fieldIt follows that a lowering of the losses should be obtained by reducing the filament diameter, decreasing the twist pitch and increasing the matrix resistivity. These considerations have led to the development by GEC ALSTHOM and the Laboratoires de Marcoussis, since 1983, of long lengths of submicronic multifilamentary wires. In an attempt to further reduce the losses by reducing the filament diameter, it was observed that proximity coupling of closely spaced filaments led to a loss increase2. In order to calculate losses more accurately we must have a better estimate of Jc values with the largest possible precision.As a first approximation Jc can be obtained by a deconvolution of the critical current IC (B,T). Ideally, one should be able to model the magnetization curves with such a degree of perfection that a selfconsistent determination of Jc becomes possible.We studied the proximity effects on the AC 87/3B wire, and the study of the reversible motion of the flux line was performed on an industrial wire : the AC 8 8 L 2 wire (The main characteristics of these wires manufactured by ALSTHOM Intermagnetics S.A. (AISA) are given in table 1). The first objective is a critical current density determination with a good modelisation of proximity effects and reversible motion of the flux lines. Proximity effects estimate3We propose to approach these effects through a determination of an effective filament diameter dF* . Filaments arrangement ----NCPI L A Y E R S dF : filament diameter with the Bean model f : wiredependentlaw Superconducting zone d~* : @x.f (wire.B 0 f ( wire, E) = 1 + 2 (NCP1-1)(1 2D : distance between filaments NCP1: number of layers in a stack E, , , : the coherence length of the normal metal Bo : field wire parameter T : temperature We can thu...