Nuclear fusion device of the air-core tokamak type

[57] ABSTRACT

An air-core tokamak having high symmetry for providing an increased ratio of plasma radius over the major radius, stability, high B and minimizing field error, as well as a low aspect ratio, high accessibility to the plasma, and separation of the plasma column from the proximity of the material of the vacuum container wall and limiters.

The toroidal confinement device has an air-core toroidal solenoid disposed along the inside diameter ofthe plasma column for producing a toroidal electric field, and spaced apart, toroidal conductors disposed along the outside diameter of the plasma column for compensating for the return flux of said solenoid and providing a transverse magnetic field for stably confining the plasma.

Claims, 3 Drawing Figures

<img file="US3663361A_D0001.tif" />

'47

Patented May 16, 1972

3,663,361

Sheets-Sheet 1

<img file="US3663361A_D0002.tif" />

Patented May 16, 1972

3,663,361

Sheets-Sheet 2.

<img file="US3663361A_D0003.tif" />

3,663,361

NUCLEAR FUSION DEVICE OF THE AIR-CORE TOKAMAK TYPE

BACKGROUND OF THE INVENTION

In the field of plasma physics, a need exists for magnetically confining a plasma column along an endless equilibrium axis in a vacuum chamber. Various proposals have been made and used to accomplish such confinement, comprising the sellarators described and shown in U.S. Pat. Nos. 3,002,912; 3,012,955; 3,015,618; 3,016,341; and 3,278,384. While these arrangements are useful and have accomplished the desired confinement, the magnetic confining fields have been low or imprecise. It is additionally desirable to provide a high-symmetry, low aspect-ratio tokamak, together with high accessibility to the plasma for injection of laser or other particle beams and/or for diagnostics and separation of the plasma column from the proximity of the material of the vacuum container walls and limiters.

SUMMARY OF THE INVENTION

This invention which is an improvement over the abovementioned apparatus, provides an air-core tokamak for overcoming the heretofore known difficulties, and accomplishing the above-recited desired results. As such, this invention employs a conventional endless vacuum container, axial field coils, and plasma current producing means while providing improved structural configurations and functional means for providing for the desired or improved results. More particularly, in one embodiment, this invention combines the abovementioned features common to both the heretofore known stellarators and tokamaks, while dispensing with the field conducting iron-core required heretofore in the tokamaks.

The above and additional objects and novel features of this invention will become apparent from the following detailed description when read in connection with the accompanying drawings, and the novel features will be particularly pointed out in connection with the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

In the drawings, where like elements are referenced alike:

FIG. 1 is a partial cross-section of the improved device of this invention;

FIG. 2 is a partial top-view of the apparatus of FIG. 1, illustrating the dimensions of the preferred embodiment thereof;

FIG. 3 is a partial cross-section of the apparatus of FIG. 2, through III—III, illustrating the structural and functional features thereof.

DESCRIPTION OF THE PREFERRED EMBODIMENT

This invention is useful in magnetically confining a plasma column having wide variety of plasma particles therein. For example, the plasma particles, comprise electrons and ions of the same or different species. One species of ions, comprises deuterium nuclei alone, but ions of deuterium, tritium, or heavier particles, such as He³ nuclei, may be used alone or in combination. Accordingly, this invention is useful for the wide variety of applications, over the wide range of plasma particle velocities, densities, temperatures, confinement times, and plasma currents to which the previously employed stellarator and tokamak techniques have been applied.

In this regard, it is known that the application of a homogeneous (uniform) axial magnetic field to plasma particles in a vacuum container will cause the paths of the particles to bend in tight helixes encircling the lines of force of the magnetic field in the container. As a result, the particles cannot freely move transversely across the magnetic field lines to the walls of the container, and each particle revolves around a line force at a radius that depends inversely upon the magnetic field strength. While particle collisions cause the particles to diffuse across the magnetic field, the rate of diffusion varies inversely proportionally to the square of the axial magnetic field, (i.e., the rate varies as 1/B/, where B₂ denotes the axial magnetic field strength). As stated in “Project Sherwood — The U.S. Program in Controlled Fusion,” by Amasa Bishop, Addison Wesley, 1958, this relationship has significance, for it shows that if a plasma can be confined in a stable plasma column, the rate of particle diffusion towards the vacuum container wall reduces by increasing the strength of axial magnetic field B_z.

It is also known that a plasma current will stably confine the plasma in a column in the described uniform axial magnetic field B₂. To this end, as described for example in the abovementioned patents, conventional means are employed. One well-known means for advantageously producing the plasma current is a conventional transformer coil having an iron core around the plasma column on the outside of the vacuum container. This transformer coil and iron core induce an ohmic (i.e., resistive) plasma current discharge along the endless container axis. One suitable means, which produces such a plasma current, is shown in FIG. 5.5 of “Controlled Thermonuclear Reactions” by Glasstone &amp; Lovberg, Van Nostrand 1960, which as described in this publication, causes a pulse of current to pass through the toroidal plasma column by making it serve as the secondary of the ion-cored transformer. In this regard, the primary of the transformer is energized from a low-frequency generator, although a discharge from a capacitor bank may alternately be used. In the case of the tokamak, the primary circuit is laid along the outside of a torus shaped container wall that functions as a conducting shell.

One explanation for the stabilization of the plasma column in the above-mentioned apparatus of U.S. Pat. No. 3,016,341, is based on the pinch effect along the plasma current. For example, as described in the above-mentioned Bishop publication, and shown therein in FIG. 3-1, the plasma current produces a magnetic pinch field having field lines that encircle the plasma column and exert an inward force on the plasma column thus pinching the plasma particles therein to the center of the vacuum container with a strength corresponding to the plasma current under the influence of voltage induced therein. The described axial magnetic field in the interior of the pinched plasma stabilizes the sharp (short wave-length) kink instabilities in the plasma by inhibiting the development of bulge-type “sausage” instabilities therein. To this end, the described axial field, referred to above as the B₂ field, provides a “back-bone” along the axis of the pinched plasma column in the encircling magnetic field lines. These short wave-length instabilities are referred to also as magnetohydrodynamic (i.e., MHD) instabilities.

On the other hand, the broad or long wave-length (i.e., nonMHD) instabilities are also stabilized to this end, e.g. with conventional tokamaks, the conducting shell is employed. In this regard, the conducting walls in the tokamak have heretofore formed a metal sheath around the plasma column wherein image currents induced in the conducting sheath produce forces tending to push the mentioned long wavelength instabilities in the pinched plasma back toward the central axis of the plasma column. This feature, as well as the other above-described features, are described in detail in the above-mentioned 1958 Bishop reference. Alternately, however, these long wave-length instabilities, comprising the wellknown Universal Instability, can also be controlled and suppressed by feedback stabilization, as described in the co-pending application, assigned to the assignee herein and entitled “Plasma Control By Feedback,” Ser. No. 12,309, filed Feb. 18, 1970, by Hendel et al. In this system, described in the abovementioned co-pending application, the phase and frequency of the Universal Instability are detected by various suitable detecting means, e.g. conventional probes, and after suitable processing by amplification and phase shifting by 180°, the processed signal is fed back into the interior of the plasma column for suppression of the instability and/or for density smoothing of the plasma in the plasma column. However, as understood from the above-mentioned co-pending application and another co-pending application, entitled “Dynamic Sta3,663,361 bilizer For Plasma Instabilities To Improve Plasma Confinement and To Increase Plasma Density,” Ser. No. 12,310, filed Feb. 18, 1970, by Hendel et al., and assigned to the assignee hereof, various other energy or particle injection systems can alternately be employed for this long wave-length instability suppression, comprising micro-wave feedback and neutral particle feedback, or micro-wave injection above the ion cyclotron frequency of the plasma.

Also, referring to the described axial B₂ magnetic field, and the encircling field around the plasma column due to the described plasma current, the combination of these fields produces a helical twist to the encircling field lines. This produces some non-negligible shear.

The mentioned shear is analogous to the well-known shear of favorable gradient produced in the well-known C-stellarator at Princeton University. There, however, the shear of favorable gradient is produced by helical conductors having oppositely directed current in the adjacent helical conductors, which are arranged around the end loops of the vacuum container. These helical conductors can thus be eliminated by the described combination of axial B_z magnetic field and encircling field due to the described plasma current along the axis thereof. The described twisting for stabilization, (i.e., twisting the described encircling equipotentials of the encircling magnetic field lines into long spirals) also symmetrizes the wellknown E X B drifts.

While the above described theories of stabilization have been confirmed by actual tests, the effects of the plasma current have additionally been described as providing a selffocusing constriction of the stream of rapidly moving charged particles in the plasma column. In this regard, the reduction of the plasma column diameter due to this self-focussing effect, can be compared to the significant and well-known shrinking by half or more in the diameter of the beams of high velocity charged particles in high energy accelerators where the generation of attractive forces in the beam currents is increased during acceleration. In the latter case, the electrostatic forces of repulsion of the charged particles exceeds the mentioned attractive forces until the velocities ofthe particles in the beams reach the velocity of light, when the two forces exactly cancel. In this beam current, comprising mass particle movement in a stream past stationary (crossed) field lines, short-wave length instabilities, such as sharp kink or sausage instabilities are inhibited. However, while the particles in the accelerator beam, comprise one or more species of positive and/or negative particles moving in one or opposite direction, having a decreasing density gradient radially outwardly, and having a uniform particle velocity distribution, the particles have a Gausian particle velocity distribution in an accelerator beam in contrast to the particles in the described plasma column, where the particles have a Maxwellian particle velocity distribution. Thus, in accordance with this explanation, various means can be used to produce the described plasma currents for achieving self-focusing constriction for causing the described diameter shrinking, which in the abovedescribed plasma column is produced by induction during the time the plasma is resistive.

It has been found in accordance with this invention that the above-described apparatus has certain limitations, which will be described hereinafter in connection with the tokamak design followed at Kurchatov. These limitations, comprise an attainable aspect ratio of the major to minor axis in the endless toroidable vacuum chamber i.e., R:a) limited at about 3-4 by the use of an iron core (the T-10 design being about the best that can be done); the degree of axisymmetry is limited by the toroidal-field coil design to low standards; and the use of a close-fitting conductive copper shell creates a conflict between the need for insulated gaps and access ports therein and the resultant perturbations in symmetry.

The aspect ratio consideration is important, because the possibility of obtaining finite B in tokamaks depends sensitively on this issue. Here B denotes the ratio of the outward pressure Pp of the magnetically confined plasma, and the inward magnetic field pressure P_fi that the confining magnetic field exerts. Likewise, the symmetry problems are significant, since magnetic surface calculations at PPL have shown the existence of “islands” of magnetic field along the axial magnetic surfaces in configurations with bumpy toroidal fields such as the described axial field B„ e.g. as found in the T-3 tokamak. Here, the magnetic surfaces are the well-known concentric magnetic apertures formed by the described axial field lines of the described B₂ field. Also, computations at LRL have demonstrated a serious effect due to access ports in current carrying surfaces surrounding axisymmetric toruses. It is thus possible for imperfectly designed tokamaks to have less favorable magnetic surfaces than well-designed stellarators.

Finally, the question of particle orbits deserves special mention. In this regard, particles trapped in magnetic mirrors between toroidal, i.e., axial field coils (i.e., about 10 percent of the particles in the T-3 tokamak) are on escape orbits, due to the toroidal curvature drift. Here, the magnetic mirrors are the well-known, spaced apart axial magnetic field coils arranged to produce an axial magnetic field that is weak in the central region, but strong at the two ends, such as described in the above-mentioned stellarator patent to Lenard et al., U.S. Pat. No. 3,278,384, whereby the strong fields at the ends constitute “magnetic mirrors” that tend to repel the charged plasma particles back toward the central region. As in stellarators, this feature of the trapped particles is not critically important for one-component plasmas with substantial collision frequencies, but it does prevent the full isotropization of the high-energy particle components, and therefore could have an influence on the “anamolous heating process” of the heretofore known tokamaks.

The advantages of this invention will be understood with reference to the parameters of FIGS. 1-3, by comparing them to the parameters of a conventional tokamak. In this comparison, in one embodiment the endless equilibrium axis of the apparatus of FIGS. 1-3 is R=110+20+20+50=200 cm, according to the formula R=i+8_c+8„+a, where b=l 10 cm, δ_Γ=2β cm, 8,.=20 cm, and a=50 cm. Here the diameter of the plasma column is 2α, the distance across the vacuum container from the plasma column to the inside of the toroidal air-core of this invention is δ_ν, the thickness of this toroidal coil is 8_C, and the radius of the air-core of this coil is b.

Further relevant parameters for comparison with the heretofore known conventional tokamak, comprise B_£ (experimental)=(a²/B²)( l/ρ²) = 1.6 percent

B_£ (theoretical)~(a/R)( 1/q²) = 6.25 percent /«//= -—B₍ - = 1.25 X 10“ amps μ»Κ q τ = 300 msec

Flux of the core = 3.8 volt-sec.

Power of the toroidal coil = 92.6MW ^toroidal -----B, = 4.0 x 10⁷ amps ____ μ»

In the conventional tokamak, we note that because b/a is limited, where b is the radius of the iron-core solenoid in the tokamak, R/a cannot ordinarily be reduced to less than 4. Even with the conductivity increased by a factor of 4, corresponding to 250 eV, then b/a = 1.76. Thus, R/α reduces to 3.6, and the practical limit of R/a for the heretofore known iron-core tokamak is placed near 3.5-4.0. Therefore, the aircore solenoid of the toroidal confinement device of this invention is an improvement over the heretofore known systems, as will be understood in more detail from the following.

3,663,361

In accordance with this invention, the desired low aspect ratio Ria is obtained with suitable and accurate magnetic field strength, and without the conventional tokamak iron-core, known heretofore. To this end, the toroidal electric field of the invention described herein, is induced primarily by the air- 5 core coil 19 [C₃], which as illustrated in FIG. 1, is pulsed to 20 kilogauss (kG) on a 10-msec time scale. The (pair of) coils [C₄] 13 and 15, on the other hand, are used to shape the return flux of coil [C₃] 19, creating a vertical magnetic field with a mirror bulge, suitable for equilibrium of the plasma column. Advantageously, the current in coils [CJ is programmed by suitable means well known in the art of computers, etc. according to the plasma circuit response.

In a practical embodiment of the apparatus of FIG. 1, the plasma aperture 21 is initially defined by a conventional limiter 23, i.e., in the initial stages of the discharge for forming the described plasma current. Such a limiter is well known as being in actual operation in connection with the model C stellarator at Princeton University. Advantageously, this limiter 23 is then retracted by a suitable fast-acting pneumatic biasing means 25 (as has been done on earlier experiments on the described Model C stellarator). The toroidal-field coil [C₂] 27, which advantageously has about 10⁶ ampere turns pulsed from a suitable source 29, can then be used to compress the plasma column 31 away from the cylindrical wall 33, which is formed on the inside of vacuum container 35. Initial rf preionization and/or subsequent cyclic compression can also be used for heating the plasma if desired. The toroidal-field loops [C₂-A] 37 and [C₂-B] 39, which are shown also in FIG. 1, are used to compensate for the flux from coil [C₂] 27, the total flux being conserved within the main axial solenoid coil [C,] 41.

The vacuum system, which utilizes pump 43, is closed in container 47 or immediately surrounding the inner coils 27 at V₂ in FIG. 1. Also, the heretofore known, conventional tokamak copper conducting shell, ordinarily placed at V, in FIG. 1, can be removable, since in the geometry of this invention,,it is not required for equilibrium. Thus, the removal of this conducting shell, eliminates transient magnetic field irregularities associated with eddy currents, and enhances accessibility to the plasma in column 31, e.g. for conventional diagnostics, such as by laser interferometry for determining plasma density by Thompson scattering and/or x-ray emission, and/or for injection of plasma or particles, or heating in situ, such as by laser heating, ion cyclotron resonance heating, adiabatic heating, plasma heating by micro-wave generation of high-energy electrons or by a combination thereof.

In the example of this invention illustrated in FIG. 3, the apparatus 45 provides an azimuthally symmetric diffuse pinch, which has application to fusion research due to the ease with which B can be increased. The fact that the plasma in column 31 is heated, especially at the center where the magnetic selffocusing well, due to the described plasma current, is the deepest, makes heating by heating means 19, comprising an ohmjc heater, attractive for the plasma production, e.g. from material from a suitable source 49 for injection of suitable gaseous or plasma forming material 51, into vacuum container 35. In contrast, plasma injection from a suitable conventional source involves dissipation of the kinetic energy thereof. Here, by analogy, the plasma may be compared to water in a cup. If the water is poured into the cup from high above the cup or at high relative velocity, care must be taken to keep the water from splashing out of the cup, even though the cup provides a perfect gravitational well. Instead, if snow packed into the cup melts, the water, of course, will not splash out.

The axial magnetic field forming means 41, comprises suitable windings S3 that surround the vacuum container minor cross-sectional diameter 55 along the entire length of the major, toroidal, endless axis 57. As described above, the axial magnetic field twists the pinch field lines encircling the plasma column 31 due to the plasma current. This combination of axial and pinch fields produces plasma stabilization, as described above, wherein weak but non-negligible shear, analogous to the conventional shear known heretofore, is present.

The apparatus 45 of FIG. 3 provides a high B, while keeping the toroidal field reasonably high. Also, the plasma radius 59 is maximized in accordance with this invention. This requires a/R as large as possible, while q=2II/6 is larger than 2, which in turn is not limited in accordance with this invention by the volts-seconds of an iron-core. A mathematical discussion of this latter feature of this invention is given in Princeton University Technical Memorandum TM-253 by the inventor of this invention.

The operation of the apparatus 45 of FIG. 3, based on the above described length a of 50 cm, will be understood from ₁₅ the following. The axial field coil 41 provides -lOkGatthe plasma in plasma column 31. The current required for this is 2 x 10® amps turn. The field is de op quasi de. Coil 27 is a pulsed toroidal field coil, which provides - 10^e amp of ringing pulsed magnetic field, thus producing the compression of 20 the plasma in column 31 to provide plasma heating. Loops 37 and 39 are cancellation loops for compensating for coupling between coils 27 and 41. Coil 19, which is the air-core coil of this invention, produces a toroidal electric field. The field produced by coil 19 may reach 20-kG (pulse in 10 msec). 25 Coils 13 and 15 compensate for the return flux of coil 19, as well as providing a transverse magnetic field with mirror shape. This makes the plasma in plasma column 31 stable against transverse motion. These coils 13 and 15 may be pulsed with the same time content as coil 19, i.e., 10 msec.

The time sequence of the above-described example of the apparatus 45 of FIG. 3 is as follows:

1. Axial (toroidal) field coil 41 is energized.

2. Coils 19, 13 and 15 are pulsed to provide the required ₃₅ electric field and the vertical field ( the coils 13 and 15 being programmed).

3. The limiter 23 is retracted.

4. Coil 27 is pulsed to compress the plasma in plasma column 31.

₄₀ For ready reference the following Table gives a review of terms used in the specification:

TABLE OF TERMS q — safety parameter (equal 2τι, where i is the rotational 45 transform of the magnetic field line);

lo u— total current passing in the plasma in the toroidal direction;

B, — toroidal magnetic field;

μ₀ — unit of magnetic induction in MKS systems (e.g. 4π X 50 10~⁷ henry/m);

limiter — (molybdenum) for restricting the channel of current flow in the vacuum vessel so that the current and plasma do not interact directly with the vacuum vessel, thus to avoid possible damage to the vacuum vessel and contamination of the plasma by the influx of impurities.

This invention has the advantage of providing an improved air-core tokamak for magnetically confining a high temperature plasma. The system of this invention also provides a sta60 Ne plasma column having a plasma current in a uniform axial magnetic field without the need for a conducting wall or sheathe around the magnetically confined plasma column. This invention additionally provides an accurate magnetic plasma confining field with high symmetry, high ratio of 65 plasma radius to major radius, high B, low aspect ratio, and high plasma column accessibility.