JPS6161275B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a thermal persistent current switch for large currents.

Superconducting coils used in magnetic levitation trains, energy storage devices, and the like need to be continuously excited at a constant current value over a long period of time, so persistent current excitation is often performed. A persistent current switch is usually used for persistent current excitation.

Although there are mechanical contact types of persistent current switches, superconducting type switches are generally used because of their advantages such as small size, light weight, and ease of operation. This utilizes the transition between the superconducting state (zero electrical resistance) and the normal conducting state (finite electrical resistance) of the superconducting wire to perform a switching operation.

However, when using such a superconducting persistent current switch, the biggest problem is the instability of the superconducting wire. For example, when a current is passed through a switch in a superconducting state, superconducting breakdown (quenching) often occurs at a current value far lower than the critical current of the superconducting wire. In order to eliminate this instability, superconducting wires used in the windings of normal superconducting magnets are stabilized by integrating the superconductor with a normal conducting metal such as copper or aluminum, which has low electrical resistance at low temperatures. We are trying to make this happen. However, in persistent current switches, it is necessary to make the electrical resistance in the normal conducting state (resistance when the switch is cut off) as high as possible, so such low-resistance stabilizing wires cannot be used, and superconducting wires are considerably unstable. It must be used in stable conditions.

Furthermore, the instability described above tends to become more pronounced when the cross-sectional area of the superconducting wire is increased in order to obtain a large current capacity. For example, Figure 1 shows approximately 60μ
A stranded wire consisting of 305 Nb-Ti wires with a diameter of m is used as a unit wire, and the number of unit wires of a persistent current switch using multiple unit wires and the current capacity per unit wire are calculated. It is a diagram showing the relationship, and it can be seen that as the number of strands increases, the current capacity per strand decreases. This means that as the required current value increases, a dramatically larger amount of superconducting wire must be used, making it difficult to realize a small and lightweight persistent current switch for large currents.

As a means to solve such difficulties, the second
As shown in the figure, a unit switch 1 made of a superconducting wire with a small cross-sectional area (that is, a small number of unit wires in the example shown in Figure 1) is fabricated, and the necessary number of unit switches 1 are connected in parallel. This is extremely easy to imagine. Taking a superconducting wire with the characteristics shown in Figure 1 as an example, if we wanted to realize a persistent current switch with a current capacity of 1300A with a single unit switch, we would have to use six unit wires, but only one If the unit switch 1 is made using unit wires, it has an allowable current of about 470A, so it is sufficient to use three of them in parallel, and the number of wires is reduced to 1/1 compared to a single switch. Can be made 2. By using this method of using unit switches in parallel, a small and lightweight switch can be realized by reducing the amount of superconducting wire, and in addition to the advantages of high normal conduction resistance, one unit switch can be used in parallel. Once designed, all current values can be covered by simply increasing the number of devices, which has the advantage of significantly simplifying the design and manufacturing process.

However, the system using such a unit switch has major drawbacks as described below. That is, when connecting a plurality of unit switches 1 in parallel, the individual superconducting wires are connected with uniform connection resistance at the connection part 4 between the superconducting wire 2 and the external superconductor 3 to be connected to it, such as the lead wire of a superconducting coil. This is because it is extremely difficult to connect them properly. This will be described in detail below.

Normally, superconductors are not directly connected to each other by welding or brazing at the joints of superconducting wires. This is because the current capacity of such connections is often lower than that of other parts, and in the case of ultrafine multifilamentary superconducting wires (FM wires) that are commonly used these days, the normal conducting metal base material This is because it is technically difficult to remove the superconducting core wires to expose the superconducting core wires and connect them to each other, and there is a risk that the characteristics at the connection portion will be significantly degraded. For this reason, the superconducting wires are usually connected by brazing with solder or the like while the superconducting wires remain covered with a normally conducting metal base material (often copper). The superconducting wire used in persistent current switches has a low-resistance base material removed in order to ensure high resistance in the normal conduction state, but both ends of the wire, which should be connected to external superconducting wires such as superconducting coil lead wires, are used. For the above-mentioned reasons, it is common practice to leave the normal conducting metal coating and connect the normal conducting metal coating to the external superconducting wire by brazing with solder.

FIG. 3 is a longitudinal cross-sectional view of such a connection part, and the current flowing through the superconducting wire 2, which is the core wire, flows through the normal conductive metal 5, the solder 6, and the external superconducting wire 3 at the connection part 4.
Since it flows to the external superconducting wire 3 via the normal conducting metal 7, an electrical resistance exists at this connection portion 4. This is the connection resistance. The connection resistance is determined by the resistance of the normally conducting metals 5 and 7 and the solder 6, but the dominant one is the resistance of the solder 6.
is the resistance value of Oxygen-free copper, which is most commonly used as a normal conducting metal, and Pb, which is commonly used as a solder.
Comparing the resistivity of 60% Sn solder at liquid helium temperature, the former has a resistivity of about 1 × 10 ^-8 Ωcm, and the latter has a value of about 5 × 10 ^-7 Ωcm, which is about 50 times that of the former. Even if the thickness of the solder layer is considerably thinner than the thickness of the normally conducting metal layer, the resistance of the solder layer becomes dominant. Such a relative relationship does not change much even if we consider aluminum as the normal conducting metal and other low melting point alloys as the solder material.

If the connection resistance is mainly determined by the resistance of the solder 6 as described above, it is inevitable that the connection resistance values will vary considerably. This is because the thickness and width of the solder layer 6 may vary greatly depending on the heating temperature during brazing, the surface condition of the superconducting wires, the pressure bonding force between the superconducting wires, and the like. Figure 4 shows a 60 μm diameter Nb-
Two FM wires embedded with 365 Ti superconducting core wires are connected to each other with Pb-60%Sn solder, length 5 cm.
This is the result of measuring the connection resistance of the connection part. Measurements were carried out on five samples prepared under exactly the same conditions using the usual four-terminal method, applying an external magnetic field of 5 KG and passing a current of 500 A. As is clear from FIG. 4, although the samples were manufactured in exactly the same way, the connection resistance varied considerably, and the maximum and minimum values differed by a factor of about 4.

Now, let us consider the case where the unit switches mentioned above are connected in parallel. FIG. 5 is an equivalent circuit diagram when n unit switches are connected in parallel. In Fig. 5, L ₁ , L ₂ ..., _Lo are self-induction coefficients of each unit switch, r ₁ , r ₂ ..., r _o and r' ₁ ,
r′ ₂ ..., r′ _o are the connection resistances between each unit switch and the lead wire connected to it. When current I ₀ flows through this, currents I ₁ and I ₂ flow through each unit switch.
..., I ₀ is given by the following equation in steady state.

I ₁ = R _t /R ₁ I ₀ , I ₂ = R _t /R ₂ I ₀ , ... I _o = R
_t /R _o I ₀ (1) Here, 1/R _t = 1/R ₁ +1/R ₂ +...+1/R _o (2) R ₁ = r ₁ + r' ₁ , R ₂ = r ₂ + r ' ₂ ,..., R _o = r _o + r' _o (3) If the allowable current of each unit switch is I _q ,
The largest value among the shunt currents given by equation (1),
That is, the total current value when the current value of the unit switch with the smallest connection resistance reaches _Iq becomes the total allowable current. Let us now consider as an example the case where four unit switches with I _q =500A are used in parallel. Connection resistance of four unit switches R ₁ , R ₂ ,
If R ₃ and R ₄ are all equal, I ₁ , I ₂ , I ₃ , I ₄
When these values reach _Iq , the total allowable current _Iqt is determined. Therefore, I _qt =4I _q =
It becomes 2000A. However, if there are variations in the connection resistance, for example, R ₁ : R ₂ : R ₃ : R ₄ = 1:2:3:4, I ₁ : I ₂ : I ₃ : I ₄ = 1:1/ 2:1
/3: becomes 1/4. Since the total allowable current is determined by I ₁ = I _q , I _qt = 25/12 I _q = 1042 A, which is about 1/2 of the previous case.
I'm getting used to it.

Considering the transient state, L ₁ , L ₂ ..., _Lo are R ₁ , R ₂
...If it is sufficiently large compared to R _o , I ₀ during excitation
When is changing, I ₁ , I ₂ , ... I _o are approximately L ₁ ,
A current distribution determined by L ₂ , ..., _Lo is taken, and when I ₀ becomes constant, the current distribution gradually shifts to a current distribution determined by R ₁ , R ₂ ..., _Ro . The time constant of current change at this time is τ
₁ =L ₁ /R ₁ , τ ₂ =L ₂ /R ₂ ...τ _o =L _o /R _o
becomes. In this process, when the current value of the unit switch with the smallest connection resistance reaches _Iq , the process is quenched. In other words, a quench phenomenon accompanied by a time delay appears.

As described above, with the conventional method of connecting multiple unit switches in parallel to create one persistent current switch, it is difficult to obtain a highly reliable switch that has the desired current capacity and continues to flow current for a long time. The drawback was that it was not possible.

The purpose of the present invention is to eliminate such drawbacks of the conventional ones, to equalize the current distribution of each unit switch and to allow the current to flow up to its maximum allowable current, thereby increasing the allowable current of persistent current switches formed by connecting these units in parallel, and To provide a thermal persistent current switch which is small in size and has a large current capacity by eliminating delay quench and allowing current to flow stably for a long time.

That is, in the present invention, a superconducting wire for a unit switch having a device for transitioning between a superconducting state and a normal conducting state is made of, for example, Cypronickel (registered trademark), which has a high electrical resistance of 10 ^-5 Ωcm to 10 ^-4 at low temperatures.
A superconducting wire is embedded as a core wire in a normal conducting metal base material such as, for example, and the wire is arranged in a non-inductive manner, for example, by arranging two spiral wires with opposite winding directions in parallel. In this way, the connection resistance value at the connection part is made uniform, and at the same time L ₁ ,
L ₂ , L _{3 .} . . _Lo becomes a value close to zero, and τ _i =L _i /R _i becomes small. Therefore, the current distribution of each unit switch is made uniform immediately after the switching operation.

If the resistivity of the normal conducting metal is less than 10 ^-5 Ωcm, the superconducting wire becomes unstable, and if it is greater than 10 ^-4 Ωcm, the resistance between the unit switch and the external superconducting wire becomes too large, which is harmful. .

An example of the test results conducted by the inventors is: L = 1 x 10 ^-6 (H), R = 2 x 10 ^-8 (Ω) τ = 1 x 10 ^-6 /2 x 10 ^-8 = 50 (sec) By devising the winding, L can be made to a small value, and τ can be made small at the same time.

On the other hand, the resistivity of Cypronickel at liquid helium temperature is about 4 × 10 ^-5 Ω, which is about 80 times that of P _b -60% _So solder, so it is thought that the connection resistance is mostly determined by the Cypronickel base material. However,
This is because the resistance of the Cypronickel base material takes approximately the same value if the geometric shape of the superconducting wire is the same. Figure 6 shows an FM wire in which 114 superconducting wires are embedded as core wires in a Cupronickel base material with a diameter of 0.3 mm, which is a normal conducting metal, and an FM wire in which the normal conducting metal is a copper base material with a diameter of 2.1 mm. The results of measuring the resistance of the connection section when external superconducting wires are connected with P _b -60% Sn solder over a length of 15 cm are shown. Measurements were performed on 18 samples prepared under the same conditions with an external magnetic field of 5KG applied and a current of 100A.
As is clear from the figure, the variation in connection resistance is 12%.
It can be seen that the value is within 100%, which is extremely small compared to the case shown in FIG.

Fig. 7 shows an example of the present invention in which a unit switch is fabricated using six strands of Cypronickel-based FM wires with a diameter of 0.3 mm, which are the same as those used in the measurements shown in Fig. 6, and these are connected in parallel. FIG. 3 is a diagram showing the relationship between the number of parallel switches and the allowable current value of the switch when a thermal persistent current switch is manufactured. One unit switch has an allowable current value of approximately 550A, but when these are connected in parallel, the allowable current increases almost in proportion to the number of parallel switches, and there is no decrease in the allowable current value per unit. . In addition, a current of 1400A was continuously applied to a device using three of these unit switches for more than 6 hours,
During this time, no quench occurs, and there is no quench phenomenon accompanied by a time delay as described above.

As described above, a highly reliable persistent current switch for large currents can be easily obtained by connecting in parallel a plurality of unit switches each made of a non-inductively wound superconducting wire coated with a Cypronickel base material.

In the above embodiment, six stranded FM wires made of Cypronickel base material with a diameter of 0.3 mm were used for the unit switch, but the number of wires, size, etc. are not limited to this. Further, it is not necessary that the superconducting wire exists over the entire length of the superconducting wire constituting the switch on the Cypronickel base material, and it is sufficient that the superconducting wire exists only at the connecting portion.

Furthermore, the base material is not limited to Cypronickel, but other metals that have a similar resistivity at low temperatures, that is, a resistivity of 10 ^-5 Ωcm or more, such as manganin, constanta, monel, inconel, and austenite. It is clear from the above description that similar effects can be expected even if stainless steel or the like is used.

[Brief explanation of the drawing]

Figure 1 is a graph showing the relationship between the number of strands of superconducting wire and the allowable current per strand in a conventional thermal persistent current switch, and Figure 2 is a block diagram showing the case where multiple unit switches are connected in parallel. , Fig. 3 is a cross-sectional view showing a connection section where two superconducting wires are connected with solder, and Fig. 4 is a cross-sectional view showing a connection part where two copper-covered superconducting wires are connected to each other by P _b
- 60% S _o A graph showing the difference in connection resistance depending on the sample when connected with solder. Figure 5 is an equivalent circuit diagram when multiple unit switches are connected in parallel. Figure 6 is an equivalent circuit diagram when multiple unit switches are connected in parallel.
The figure shows P _b
- 60% S _o A graph showing the difference in resistance of the connection part depending on the sample when connected with solder, Figure 7 shows a thermal persistent current switch based on an embodiment of the present invention.
It is a graph showing the relationship between the number of unit switches and allowable current. In the figure, 1 is a unit switch, 2 is a superconducting wire,
3 is an external superconducting wire, 4 is a connecting portion, 5 is a normal conducting metal, 6 is a solder, and 7 is a normal conducting metal.

Claims (1)

[Claims] 1. A plurality of unit switches in parallel each having a plurality of superconducting wires, each of which is at least partially coated with a normal conducting metal, and a device for transitioning the superconducting wire between a superconducting state and a normal conducting state. The superconducting wire is connected to the external superconducting wire by solder through the normal conducting metal at the connection point where the unit switches are connected in parallel. A thermal persistent current switch characterized in that the superconducting wire is made of a metal having an electrical resistivity of 10 ^-5 Ωcm to 10 ^-4 Ωcm at a temperature at which it exhibits a superconducting state. 2. The thermal persistent current switch according to claim 1, wherein the superconducting wire is arranged in a non-inductive state. 3. Thermal persistent current according to claim 1 or 2, wherein the normally conductive metal is any one metal selected from the group consisting of cube nickel, manganin, constantan, inconel, austenitic stainless steel, etc. Switch.