GB2038532A - Super-conducting members - Google Patents

Abstract

A superconducting member A is provided with a plurality of regions of normal metal N which act as pinning centres for magnetic fluxoids, the regions being preferably in the form of fine filaments (typically 50 Angstrom units diameter) within super-conducting filaments (typically 5 micrometers in diameter) and improving the current carrying capacity of the member. The regions may be made by placing rods of a normal metal or alloy thereof within rods of superconducting material or building rods of normal metal with rods of superconductor and then drawing down the rods to give the filaments. A compound superconductor may be made in situ by using a normal metal alloy and heating. The superconducting filaments may be stabilized by normal metal and have a barrier layer. <IMAGE>

Description

SPECIFICATION Improvements in or relating to manufacture of superconducting members The invention relates to the manufacture of superconducting members and more particularly to a method of manufacture for increasing the superconducting current carrying capacity of the superconducting member.

British Patent Specifications Nos. 1,333,554 and 1,394,724 describe methods of manufacturing superconducting members comprising filaments of type II superconductive compound in a matrix of a copper alloy. The type II superconductive compound is a compound of the Al 5 crystal structure having the general formula A3B where A comprises niobium (Nb) or vanadium (V) and B comprises one or more of the elements aluminium (Al), gallium (Ga), indium (In), silicon (Si), germanium (Ge), and tin (Sn).

The superconductive compounds Nb3Sn, V3Ga and V3Si have been produced by solid state reaction between niobium or vanadium and a "bronze" matrix consisting of a solid solution of tin, gallium or silicon in copper respectively. This method permits the formation of these compounds in thin filamentary form, twisted if necessary, inside a matrix of a copper alloy. No other process is known by which to manufacture such multifilamentary configurations of these brittle Al 5 compounds on a large scale. The superconducting current densities can be very high in filaments of the Al 5 structure compounds manufactured in this way.However it is desirable to increase the current carrying capacity of these filaments further since the presence of the bronze, and of other features such as regions of pure metal to stabilize the superconductor and of strengthening components of tough alloys will reduce the overall current density in the wire considerably.

It is well known that inhomogeneities must be present inside a type II superconductive material for loss-less superconducting transport currents to flow inside the material. The reason is that above the lower critical field Hc, magnetic flux penetrates into the superconductor in units of the flux quantum (2 x 10-'5Tm2) associated with cylindrical super current vortices.

Within a radius 5 of the vortex axis the superconducting order parameter is depressed falling to zero on the axis. Therefore for many purposes the vortex (or fluxoid) can be envisaged as consisting of a normal metal core of radius t surrounded by a circulating supercurrent. The flux associated with one vortex will interact with the current in an adjacent one causing neighbouring vortices to repel each other. ( is the coherence length, a discussion of which may be found, for example, in the article entitled "The Effect of Metallurgical Variables on Superconducting Properties" by Livingston and Schadler in Progress in Materials Science Vol. 12 No. 3 1964.

Hence the vortices will spread out uniformly and form an ordered array in perfectly homogeneous type II superconductors. If there were any non-uniformity in this arrangement of vortices, flux motion and consequently energy dissipation would result to restore the uniformity. Thus such a material cannot support a loss-less transport current since such a current would imply a vortex gradient. However when inhomogeneities are present within the superconductor it is possible for gradients to build up in the fluxoid distribution and hence for loss-less superconducting transport currents to be maintained. The inhomogeneities provide regions (pinning centres) where the fluxoids find that it is energetically favourable to come to rest, and it is necessary to exert a force on the fluxoid to move it from a pinning centre.Metallurgical features such as precipitate particles, grain boundaries, dislocation networks and atmospheres of impurity atoms are efficient pinning centres and their introduction in or removal from a superconductor can cause very marked changes in the superconducting currents carried by the specimen.

Although practically any defect in the crystal lattice of the superconductor can cause flux pinning the effect is very dependent on defect size and distribution. One very strong pinning centre will produce very little effect in parts remote from it, while many tiny ones will likewise produce very little effect on a fluxoid since the distribution of pinning centres round the fluxoid will be very similar wherever it is situated so that no position is energetically more favourable for it to come to rest. It appears that a distribution of defects is most efficient at pinning when their spacing matches that of the fluxoid lattice. The spacing of the fluxoid lattice will itself vary with the field, the spacing being reduced as the field is increased until the core regions begin to overlap as the upper critical field is approached.Hence different distributions of defects will pin more effectively at different fields. In recrystallized niobium titanium alloys, for example, where the grain boundaries are the only pinning centres the maximum pinning is found towards low fields where the fluxoid spacing is larger, while when small finely spaced precipitates are produced by heat treatment the peak in the pinning curve moves to higher fields.

The critical current density in a ductile superconductor can be changed by several orders of magnitude by cold work and the formation of precipitates of normal metal. It is not possible to build up high concentrations of dislocations in Al 5 compounds by working the material because they are very brittle so that flux pinning from these defects is low. The grain or crystallite size appears to be the important factor controlling the critical current density in Al 5 compounds grown by several different techniques. Impurity elements introduced into the starting material, e.g. carbon in the vapour deposition of Nb3Sn, or zirconium (Zr) and Oxygen (0) introduced into the niobium when forming Nb3Sn by liquid state reaction with tin can produce significant improvements in the current carrying capacity of the Al 5 compound.It is not clear in every case whether the impurity works by producing precipitates of normal metal, or by the effect these precipitates have on the grain size of the Awl 5 phase. Alloying additions have also been tried with the bronze route production of A15 phases. With Nb3Sn for instance, we have tried approx. 1 atomic per cent (a/o) additions of Hafnium (Hf), Zirconium (Zr), Cobalt (Co), Titanium (Ti), Molybdenum (Mo) to the niobium and found slight increases with Zr (if the alloy is anodised), Hf, Co and Mo additions, the zirconium addition having the additional advantage of speeding up the rate of Nb3Sn information.By replacing part of the tin in a 94% copper tin alloy by Al, Ga or In we have produced slight increases in the current carrying capacity of Nb3Sn, with a marked (two to three fold) increase in the case of a 94a/, copper, 1 2a/O tin and 42ago gallium alloy. It has also been reported that dilute germanium additions to tin bronze produce marked increases in the critical current density of Nb3Sn.

According to the present invention potential pinning sites are introduced into a superconductor during the manufacturing process.

The invention provides a method of manufacturing a superconducting member from a plurality of components formed into a unitary structure comprising including in a component (hereinafter referred to as the superconducting component) which is to provide superconducting properties in the superconducting member a plurality, and preferably a multiplicity, of regions of normal metal or alloy of dimensions appropriate to provide pinning centres for fluxoids when the superconducting member is in use. It is to be understood that by "superconducting component" we include a precursor material which is not necessarily a superconducting material initially but becomes so on subsequent treatment, such as by reaction with another element.

By normal metal or alloy we mean a metal or alloy which will be non-superconducting at the temperatures and magnetic field conditions in which the superconducting member is to be used and in which the superconducting component will be superconducting.

Preferably the method of manufacturing comprises providing a plurality of rods of the superconducting component and a plurality of rods of normal metal or alloy either received in bores in the rods of the superconducting component or bundled together therewith, and reducing the rods together by drawing or like mechanical deformation until a final desired configuration of filaments of the superconducting component with finer filaments of normal metal or alloy included therein is obtained.

Since fine filaments of normal metal or alloy of diameter of the order of 50 Angstrom units or less in filaments of the superconducting component of diameter of the order of 5 micrometers are required, it will be appreciated that a manufacturing procedure involving repeated bundling, drawing down, cutting and rebundling will normally be required. The procedure facilitates close control of the number, size and distribution of the fine filaments of normal metal in the superconducting member in its final form.

The normal metal or alloy may contain a constituent which is desired alloying addition for the superconducting component. The method thus provides a technique for introducing such alloying additions to the superconducting component.

The method is particularly appropriate in the manufacture of superconducting members which comprise a carrier matrix of electrically conductive, but non superconductive, material in which is embedded a large number of fine superconductive filaments of a compound of the Al 5 crystal structure having the general formula ASB where A comprises niobium or vanadium and B typically comprises one or more of the elements aluminium, gallium, indium, silicon, germanium and tin.

Methods for the manufacture of such superconducting members are described in British Patent Specifications Nos. 1,333,554 and 1,394,724, which methods are readily modified to incorporate the method of the present invention.

Conveniently the normal metal comprises copper but other ductile refractory metals such as niobium, vanadium, tantalum, titanium and zirconium may be employed.

Where a superconducting member is manufactured by a method as described in Patent Specification No. 1,333,554 or 1,394,724 modified in accordance with the present invention, it is convenient to employ as the normal metal a Cu-B alloy. In this case it is noteworthy that the inclusion of fine filaments of Cu-B alloy within A metal filaments provides an additional source of B metal for reaction with A to form A3B compound when the heat treatment is carried out.

Specific methods of manufacture embodying the invention will now be described by way of example and with reference to the drawings filed herewith, in which: Figure 1 is a diagrammatic cross-sectional view of part of a block or ingot of a metal A with inserted rods of a normal metal N; Figure 2 is a diagrammatic cross-sectional view of a bundle of rods of metal A having interspersed therewith a plurality of rods of a normal metal N; Figure 3 is a diagrammatic cross-sectional view of a nesting array of tubes of metal A with intervening tubes of a normal metal N; and Figure 4 is a diagrammatic cross-sectional view of an array similar to that of Fig. 3, but in which in place of the tubes of normal metal N there are circumferentially spaced arcuate strips of alternately normal metal N and metal A.

In the examples to be described the basic technique of manufacture of the superconducting member is that of British Patent No. 1,333,554. This method was developed to manufacture hard brittle compounds in the multifilamentary configuration. The Al 5 structure compound A3B is produced in the following manner. Rods of metal A are embedded in holes drilled in an ingot of an alloy comprising a solid solution of the element B in a carrier metal such as copper which does not react with the other components in the subsequent heat treatment. For example when producing Nb3Sn niobium rods are inserted in holes in a copper-tin bronze, while for V3Ga vanadium rods are inserted into a copper-gallium alloy ingot. The rods of metal A are converted to filaments by reducing the diameter of the ingot containing these rods by some mechanical process such as wire drawing.If necessary the rod produced from the initial bronze ingot with inserts of metal A may be cut up and the shorter lengths stacked together and this bundle (preferably contained inside a tube) reduced further by mechanical means to produce finer filaments of the metal A. These filaments are converted to the compound A3B by heating the composite of A filaments and bronze to a temperature at which the bronze matrix is still solid but where reaction to form A3B can occur at the interfaces of the A filaments with the bronze matrix. Variations in this method of manufacture are possible, for example the initial assembly of A rods in bronze can be made by stacking together rods of A with bronze rods.

In the example of method embodying the present invention, pinning sites are introduced into the compound in the following manner. The process for manufacturing the A3B compound follows the method described in British Patent Specification No. 1,333,554 with the important difference that instead of introducing rods of metal A into the holes drilled in the bronze ingot, composite rods comprising a normal metal or alloy N in a matrix of element A are used. After the composite wire containing filaments of A with regions of normal metal or alloy N inside it has been drawn to size the wire is heat treated to produce the A3B compound. The normal metal or alloy N is chosen so that it is not incorporated in the lattice of the A3B compound but remains as discrete particles or filaments embedded inside the A3B compound.The normal metal or alloy need not be compoletely insoluble in the A3B compound provided sufficient undissolved metal remains and the superconductivity properties of the A3B compound are not deleteriously affected. Examples are described below of the use of an alloy to provide the normal metal filaments and to supply B and other elements to the A3B compound.

It will be appreciated that this adaptation of the method of British Patent Specification No.

1,333,554 is readily extended to the method of British Patent Specification No. 1,394,724 where it is desired to include a pure metal for stabilisation in the final unitary superconducting member.

Particles of normal metal such as fine precipitates are known to act as strong pinning centres.

This is because it is energetically more favourable for the normal 'core' of a fluxoid to be situated in a normal metal than in a superconductor. An advantage of the present method is that the amount of the normal material can be varied as desired by altering the relative amounts of the A and N components of the composite rod. The large volume expansion which occurs when metal A is converted to compound A3B (about 37 volume per cent when Nb is converted to Nb3Sn) must be taken into consideration when designing the composite. Too large a proportion of the normal metal or alloy N is clearly undesirable because it will reduce the amount of superconductor in a filament of given diameter.

The size of the filaments can also be varied over a wide range by altering the initial size of the rods of metal or alloy N and the amount of deformation the composite undergoes. At a boundary between a superconductor and a normal metal the superconductive wave function extends some way into the normal metal while at the same time the order parameter is depressed for some distance into the superconductor. The distance over which these changes occur is of the order of 5. Hence in a very small superconductive region in a normal metal the superconductivity will be somewhat quenched while a very small normal region in a superconductor will to some extent show superconductive properties. This phenomenon, referred to as the proximity effect, has limited the size to which superconducting filaments can be reduced and retain their full superconducting properties.For example it has been reported that the critical temperature and upper critical field of niobium filaments a few hundre Angstroms in diameter embedded in a copper matrix are lower than for niobium filaments 2 ym or larder in diameter.

However in niobium titanium alloys precipitates of co phase particles about 50 A in diameter can pin very strongly so the lowest effective diameter for filaments of metal or alloy N embedded iii A3B is likely to be lower than this.

The spacing between the filaments of metal or alloy N can also be varied by varying their separation in the initial composite of metal A and metal or alloy N, and by altering the amount of deformation the composite receives. Very strong pinning would occur if the fluxoids could come to rest with the normal metal filaments situated along the 'cores' of the fluxoids. Such a situation is unlikely to be encountered in practice. The deformation of the composite and the volume change that occurs on A3B formation will disturb the perfect regularity of the filament arrangement so that matching of fluxoids and filaments will not be possible everywhere.

Moreover in most practical applications (e.g. in solenoid windings) the fluoxids will be nearly perpendicular to the filaments. Qualitatively however it is clear that a closer spacing of filaments will be required for stronger pinning at high fields than would be necessary at low fields.

A quantitative indication of the desirable spacing of filaments for matching the fluxoid spacing at a selection of field strengths may be derived from the following relationship: n B=- fo=N N fo S where B is the magnetic induction n is the number of fluxoids S is the cross-sectional area of the superconductor N = n/S is the number of fluxoids per unit area is is the flux quantum 2 X 10-'5Tm2 Calculation shows the following approximate results for the spacing "d" of fluxoids on a triangular lattice: when the magnetic field H= 1 Tesla, then d-480 Angstrom units H= 5 ,, ,, ~d-215 H= 10 ,, ,, d-152 H= 15 ,, d#cl 24 ,, d-124 H 20 ,, ,, d-107 This is true for any superconductor.The fluxoid core size 4 does vary with the superconducting properties of the superconductor material. For example this has been calculated to be 50 Angstrom units for V3Ga.

There is some evidence in the case of pinning by discrete particles that flux flow occurs at high fields by shearing of unpinned fluxoids past pinned ones. This cannot occur for continuous filament shaped pinning centres as a fluxoid cannot move very far without encountering a filament. Hence continuous filaments could be particularly effective pinning centres at fields approaching Hc2.

The manufacturing of the initial composite rods of the metal A containing regions of normal metal or alloy N can take several forms. One method is illustrated in Fig. 1 where a block or ingot of metal A is drilled with a multiplicity of holes. A regular triangular pattern of holes is shown in the figure but other configurations could be employed.

Rods of metal N are inserted in these holes and the ingot is then mechanically deformed (e.g.

by swaging, extrusion, rolling or drawing) to produce fine filaments of metal N embedded in metal A. If the size of the filaments is too large the rod may be cut up and pieces (given a hexagonal profile for convenience of stacking) stacked together to form another composite. This composite (preferably inside a tube) is then mechanically reduced to produce a rod suitable for insertion into the drilled bronze ingot.

A second method is illustrated in Fig. 2. Here a stack of rods (given hexagonal profile for ease of stacking) is constructed with rods of metal N distributed among the A rods in a definite pattern. An equilaterial triangular pattern is shown but other patterns may be employed. The stack, preferably inside a tube (profiled if possible to fit the stack) is then reduced to a rod form suitable for incorporation in the drilled bronze ingot. A double (or multiple) bundling of the rods of N and A is also possible if fine filaments (or spacings) are required.

In a variation of the method shown in Fig. 2, rods of metal A are drawn through a special die which imparts a hexagonal cross-sectional profile to the rods with a semicircular groove at the centre of and running along the length of each face of the hexagon. Circular section rods of metal or alloy N are received in the grooves, which mate together to form circular section apertures between abutting faces of the hexagonal section rods of metal A when these are stacked together. Thus, in the stack, each hexagonal section rod of metal A will be surrounded by six smaller diameter circular section rods of metal or alloy N. On subsequent mechanical reduction of the composite stack to a rod, the insertions of metal or alloy N will become a matrix of fine filaments within the metal A.

It will be appreciated that where the metal A is niobium, a coating on the niobium is necessary for preventing galling when the niobium is drawn through the die. Any such coating must, of course, be compatible with the subsequent operations for forming the final superconducting member.

This particular problem is avoided in a further alterantive configuration in which the normal metal or alloy N is incorporated within the filaments of the metal A in the form of an interconnected network. This configuration is achieved, for example, by providing hexagonal section rods or bars of metal A with a uniform thin layer surface coating of metal or alloy N. The coated hexagonal section rods are then stacked and mechanically reduced as before to produce a composite rod suitable for insertion into the drilled bronze ingot, or otherwise incorporated into a matrix with the Cu-B bronze for the reaction formation of A3B superconducting compound.

The interconnected network of normal metal or alloy N will not in practice maintain the metal A on conversion to A3B in the form of separated sub-filaments as might appear at first sight. This is because the volume expansion which occurs when metal A is converted to A3B compound will, in places, breach the very thin layers of normal metal or alloy N so that the sub-filaments of A3B become interconnected.

In a specific example made by this technique, a niobium rod was drawn into a thin walled copper tube and given a hexagonal cross section by drawing through a hexagonal shaped die.

The rod was cut into shorter lengths and 163 of these stacked together within a thin walled bronze tube, which was then swaged and drawn down to a composite rod of 3/1 6 inch (-4 mms) diameter. The regularity of the copper network was quite well preserved during the deformation sequences. A further cutting, re-bundling and reduction would give a network of mixed normal metal (copper and bronze) on a finer scale inside the eventual filament of niobium.

Other less regular arrangements are possible. Fig. 3 shows one with concentric annuli of metals A and N, which can be manufactured by assembling co-axial tubes and deforming these to produce annuli of metal N of the requisite spacing. If the metal N is impervious to the component B diffusing from the bronze then the inner annuli of metal A will not be converted to A3B. This problem can be overcome by replacing the annuli of metal N by segments of metals A and N (Fig. 4) so that the segments of metal A provide paths for the element B to diffuse from the outside to the interior. Although more difficult to manufacture such an arrangement is preferable even when B can diffuse through metal N since the metal N is thus reduced to smaller regions and the A3B regions that form within the filmant can be formed together.

It is desirable that the normal metal or alloy N should not be ferromagnetic to avoid any hysteresis behaviour of coils. Metal or alloy N must be ductile enough to be co-processed with metal A, and it is desirable that its mechanical properties are similar because in this way the general arrangement of the N filaments in the matrix of metal A is better preserved at greater deformations. If during manufacture the composite of A and N has to be softened then both co.i,ponents must be able to withstand this heat treatment without embrittlement either from the other component dissolving in it and altering its mechanical properties or from the formation of a brittle compound of A and N at their interfaces.During the solid state reaction to form A3B the metal or alloy N should not inhibit or unduly slow this reaction, and as discussed earlier if any component from N does dissolve in A3B it should not degrade the superconducting properties of this phase.

Pure copper is a suitable material for use as the normal metal N because it is ductile and as the carrier metal generally used for the bronze alloy is known not to produce any very great degradation of the properties of the A3B compounds. However since even after prolonged annealing some B is left in the bronze the copper regions are likely to dissolve some of the B and hence slow the conversion of A to A3B somewhat. A possible solution is to use a Cu-B bronze as the normal alloy N. This has the advantage of supplying some of the B component for A3B formation, and it is harder than pure copper so that its mechanical properties are often closer to those of metal A so that more regular deformation of the composite of A and N is possible.

However, on deformation the area of interface between the filaments of the normal alloy Cu-B and the matrix of metal A will be increased. It is therefore possible that the reaction to form AsB will occur more readily and at lower temperatures as deformation increases, and this may interfere with the fabrication of the composite. The use of Cu-B for the normal metal filaments may consequently be limited to certain systems or to certain ranges of concentration of the element B.

Introducing alloying elements into the A3B is facilitated by incorporating them in the normal metal or alloy, since the alloying element is provided at the place it is needed instead of being spread throughout the bronze. In the case of the production of Nb3Sn, for example, it is known that gallium can enhance the critical current. Also in concentrations of -6"/, gallium or less there is no reaction between a copper gallium bronze and niobium to form a compound at the interface between these materials. Hence the gallium can be incorporated inside the rod of niobium and either dissolve into the niobium or react with the Nb3Sn as Sn penetrates to the centre of the wire.

Ductile refractory metals (e.g. Ta, Ti, Zr) may be used as the normal metal N. These may show some reaction with the B element, but providing the compound that is formed is nonsuperconducting, satisfactory pinning is provided.

Claims (11)

1. A method of manufacturing a superconducting member from a plurality of components formed into a unitary structure, which method comprises including in a component (hereinafter referred to as the superconducting component) which is to provide superconducting properties in the superconducting member a plurality of regions of normal metal or alloy of dimensions appropriate to provide pinning centres for fluxoids when the superconducting member is in use.

2. A method as claimed in Claim 1, wherein there is included in the said superconducting component a multiplicity of said regions of normal metal or alloy.

3. A method as claimed in Claim 2, wherein there is provided a plurality of rods of the superconducting component and a plurality of rods of normal metal or alloy either received in bores in the rods of the superconducting component or bundled together therewith, and reducing the rods together by drawing or like mechanical deformation until a final desired configuration of filaments of the superconducting component with finer filaments of normal metal or alloy included therein is obtained.

4. A method as claimed in Claim 3, wherein steps of bundling, drawing down, cutting, rebundling and redrawing down of the rods are repeated until fine filaments of normal metal alloy of diameter of the order of 50 Angstrom units or less included within filaments of superconducting component of diameter of the order of 5 micrometres are obtained.

5. A method as claimed in any of the preceding claims, wherein the normal metal or alloy contains a constituent which is a desired alloying addition for the superconducting component.

6. A method of manufacturing a superconducting member as claimed in any of the preceding claims, wherein there is formed an alloy consisting of a carrier material and at least one element from the group consisting of aluminium,, gallium, indium, silicon, germanium and tin, and the alloy is contacted with a superconducting component comprising niobium or vanadium in which there is included a plurality of regions of normal metal or alloy and heat treated to cause a solid state reaction between the niobium or vanadium and the element or elements from the said group to form a superconducting compound therewith, the carrier material being such as will not react substantially with the base material under the heat treatment, the heat treatment temperature being controlled for avoiding melting of the alloy of the carrier material in contact with the niobium or vanadium at any stage during the reaction, and the said regions of normal metal or alloy being of dimensions appropriate to provide pinning centres for fluxoids when the superconducting member is in use.

7. A method of manufacturing a superconducting member as claimed in Claim 6, wherein the combination of the alloy of the carrier material and the said superconducting component are formed prior to the heat treatment into a unitary structure together with a metal, which is eventually to provide stabilisation, and a barrier material (as defined in Patent Specification No.

1,394,724) is present and positioned in the composite unitary structure to protect the said metal which is to provide stabilisation from diffusion thereinto of any of the other components.

8. A method as claimed in Claim 6 or Claim 7, wherein the alloy at the heat treatment temperature comprises a solid solution of the element or elements from the said group in the carrier.

9. A method as claimed in Claim 6, Claim 7 or Claim 8, wherein the reaction conditions are such as to form an intermetallic compound between the niobium or vanadium base material and the element or elements, which compound has a crystal structure designated Al 5.

10. A method of manufacturing a superconducting member substantially as herein described in any of the examples.

11. A superconducting member when made by the method of any of the preceding claims.

5 10 15 20 25 30 35 40 45 50