lanl-adiabatic-compression-frc-supplemental

y~i !" • — -1^ *& MARCH mi PWl-1755 t£;, PROCEEDINGS or THE^S^PM JOINT SYMPOSIUM ON COfl PACT TORU3ES MB ENERGETIC PARTICLE INJECTION,-^" PRINCETON, NEW JERSEY, 12^12-1^1/79 v •••• w ' urt i^. PLASMA JWSICS LABORATORY K^ail PRINCETON,-Mil* JERSl Hf ==='' ffl STRilu. Hvi-i Or .t.-a!.-Ji^'.>.'i - 1 ••" i L ; V -' ipi Jt aart «r» •«f*art«d ay tlw U.J. ;«t*-t3wnt tf tn tm-.i-SHt ••.-.«-*C8J-74-'.Ke 3073. I'tj.-fudustiori;. t,r»-. =«• «£••, M**"«»t lot., u»« »«o d Upa»; , Jr. who.* ? hjp«r 'or ttt* UMtari ttttu jomrirm H i» ptf-r.-'i', •-. if -•PROCEEDINGS OF THE US-JAPAN JOINT SYMPOSIUM ONCOMPACT TORUSES AND ENERGETIC PARTICLE INJECTION HOSTED BV PLASMA PHYSICS LABORATORY PRINCETON UNIVERSITY PRINCETON, NEW JERSEY 035¥l ON Ql£IRl Bu'ii.js ui ; ma UOLOI-.ICHT IS 'JJJLIMITEQ 12-14 DECEMBER 1979 « Sum OOIWT'I » *n, mm^ ir PREFACE The six;y papers contained in these Proceedings serve to convey some idea of the highly original and diversified discussions that cook place at the US-Japan Joint Symposium on Compact Toruses and Energetic Particle Injection. Perhaps the most striking feature of the Symposium, however, was not its diversity, but its essential unity. Researchers from such seemingly disparate fields as tokamaks, mirrors, theta pinches, zed-pinches, and relativistic-beam injection found themselves confronting identical problems of physics and converging towards a similar reactor goal. The participants were also pleased to experience a second form of con­ vergence: the joining of the Japanese and United States fusion programs in a -ollaborative effort. This collaboration promises to be particularly fruitful in the area of compact torusas and energetic particle injection — where so much depends on the emergence of new ideas and new experimental techniques. Harold P. Furth 21 January- 1980 ii TABLE OF CONTENTS Page Title Page i Preface ii Table of Contents iii Introductory Remarks by J. F. Clarke 1 The Compact Torus Concept and the Spheromak by H. P. Furth 3 Injection of Relativistic Electron Beam into Toroidal Systems by A. Mohri, K. Narihara, Y. Tomita 8 The LASL Compact Torus Program by R. EC. Linford and CT Staff 12' Initial Results of Field Reversed Plasma Gun Experiment by W. C. Turner. C. W. Hartman, J. Taska 16 Experiment on Plasma Confinement by Intense Relativistic Electron 3eam Ring by Y. Tomita, K. Narihara, T. Tsuzuki, M. Hasegawa, K. Ikuta, A. Mohri 20 Intense Relativistic Electron Beams in Toroidal Magnetic Geometries by V. Bailey, J. Benford, R. Cooper, 3. Ecker, H. Helava 26 Compact Torus Research at U.C.I, by A. Fisher, S. Robertson, M. Rostoker 29 The Longshot Injector: A 3/4-fc J, 120-ke V Pu]sed Source of 3 * 10 16 Ions for Ion Ring Formation by J. B. Greenly, D. A. Hammer, R. S. Sudan 33 Reversed-Field Configuration with Rotating Relativistic Electron Beams by J. D. Sethian, K. A. Gerber, D. S. Spector, A. E. Robson 37 Results and Present Status of the Relativistic Electron Ring Experiments and Their Application to Spheromak Problems by H. H. Fleischmann 41 Reversed Field Configurations Generated by Proton Pulses by J. A. Pasour, J. Golden, J. Harsh, C. A. Kapetanakos 45 Thermal Background Effects on the Kink Instability of a Field- Reversing Ion Layer by S. J. Yakura, T. Kammash 49 Magnetized Gun Experiments by T. R. Jarboe, I. Renins, H. '-;. Hoida, J. Marshall, A. R. Sherwood 53 Formation of a Compact Torus using a Toroidal Plasma Gun by M. A. Levine, P. A. Pincosy 57 Reconnection Conditions for Flowing Field-Reversed Plasma from a Plasma Gun by J. w. Shearer, J. L, Eddleman, J. R. Ferguson ... 61 Physics of the OHTE by T. Ohkawa and the OHTE Group 65 Formation of Toroidal Plasma Confinement Configurations by using Hot Electrons by C. W. Hartman, M A. Levine 63 Particle-Fluid Hybrid Simulation of Field Reversal in a tfirror Plasma by B. I. Cohen, T. A. 3rengle 72 Two-Uiaiensional Time-Dependent Transport in Field Reversed Equilibria by S. P. Auerback, H. L. 3erk, J. K. 3oyd, 3. Mc Mamara, D. Shumaker 76 A Steady-State Beam Driven Field-Reversed Mirror by J. H. Hammer, H.'L. 3erk 30 iii Calculation of Ideal MHD Growth Rates and Eigenfunctions in Field Reversed >tirrors in the Large Toroidal Mode Number Limit by D. V. Anderson, W. A. Newcorab , • 82 Toroidal Reversed Field-Pinch Experiments by D. A. Baker ... S6 Some Properties of the Heating and Confinement in the RF? Configuration by S. Ortolan! 39 Relaxation of Toroidal Discharges by L. Turner 94 Effects of Impurity Radiation on Reversed-Field Pinch Evolution by E. J. Caramana, 7. W. Perkins 98 Field Reversal Experiments, FRX-A and FS5-B Results by W. T. Armstrong, R. K. Linford, J. Lipson, D. A. Platts, E. G. Sherwood 102 FRX-C and Multiple-Cell Eicperiments by E. E. Siemon and LASL Compact Torus Staff 106 Compact Torus Theory — XKD Equilibrium and Stability by D. C. Barnes, C, E. Seyler 110 Two-Dimensional Simulation of Compact Torus Formation by D. W, Hewett 115 Two-Dimensional Compression in General Compact Tori by E. Hameiri, W. Grossmann IIS Tearing-Mode Stability Analysis to a Cylindrical Plasma by H. L, 3erk, J. Saver, D. D. Schnack 122 The Tilting Mode in the Reversed-Field Theta Pinch by A. I. Shestakov, D. D. Schnack, J. Killeen .126 Periodic Field-Reversed Equilibria for a Multiple-Cell Linear Theta Pinch by H. Meuth, F. L. Ribe 130 Zero-Dimensional Modeling of Field-Ra Versed Theta-Pinch Machines by E. H. Klevans 135 Spheromak Formation by Theca Pinch by Y. Sogi, H. Ogura, Y. Osanal, K. Saico, S. Shiina, fl. Yoshimura 139 Field-Reversed Plasma' Gun Based on the Inverse-Pinch Discharge by U. D. Getty 143 A Trizgered-Reconnection Compacr Tiroid Experiment by A. L. Hoffman, G. C. Vlases 147 Plasma Rotation in Field-Reversed Theta Pir.ches by I. C. Steinhauer . . 151 Stellaraiak a Hybrid Stellarator — Spheromak by C. W, Hartman ..... 155 utilization of Electron Coils for an Advanced Tokamak and Conjecture About the Cause for Current Step (Down) by S. Yoshikawa 159 Dynamically Formed Spheromak Plasma (PS-1) by G. C. Goldenbaum, Y. P. Chong, G. Hart, J. H. Irby . . . *. 16: Spheromak Equilibrium and Stability and Numerical Studies of a Spheromak Formation Scheme by M. Okabayashi, S. Jardin, H. Okuda, T. Sato, G. Sheffield, A. Todd 166 Design and Fabrication of the S-l Spheromak Device by H. Yamada, J. Sinnis, H. P. Furtb, M. Okabayashi, G. Sheffield, T. H. Stix, A. M, M. Todd 171 Two-Dimensional Simulation of the Formation of the»?PPL Spheromak by A. Aydemir, C. K. Chu, H. C. Lui 176 Bifurcation of Tornidal Plasma in a Poloidal Ouadrupole Field by H- Ikezi, K. F. Schwarzenegger 180 iv Startup Scenario of Compact Tori Based on R£3-Injection Developed in SPAC Group by £. Ikuta 184 The S?S Compact Torus Experiment by A. De Silva 186 Minimum Energy Equilibria by A. Reiman, R. S. Sudan 189 Compact Toroidal Plasma Equilibrium and Implications on Stability by G. K. Morikawa 193 Radio-Frequency Flux Control of Toroidal Plasmas by S. Inoue, K. Itoh . . 197 Field-Reversed Configurations: Theoretical Considerations and Reactor Applications by G. H. Miley 200 The Holomak — A Toroidal Spheronak by T. H. Stix, A. M. M. Todd 204 The LISOS Reactor: Compression of a Compact Torus by a Liquid Metal Liner by A. E. Robson 208 Preliminary Studies of Spheronak Reactors by M. Katsurai, M. Yamada . . . 212 The All Plasma Spheromak: The Plasmak by P. Koloc, J. Ogden 216 Neutral Beam Sustained, Field-Rever3ed Mirror Reactors by G. A. Carlson, K. R. Schultz, A. C. Smith, Jr 220 The Moving-Ring Field-Reversed Mirror Reactor Concept by A. C. Smith, Jr,, G. A. Carlson, H. Et. Fleischmann, T. Kammash, EC. R. Schultz, D. M. Woodall 224 Preliminary Reactor Implications of Compact Tori: How Small is Compact? by R. A. Krakowskl, R. L. Hagenson 223 TRACT: A Small Fusion Reactor Based on a Compact Torus Plasma by H. J. tfillenberg, A. L. Hoffman, L. C. Steinhauer, P. H. Rose. . . • 233 List of Attendees 237 v -1- I7JTR0DUCT0RY REMARKS John F. Clarke, Deputy Director, Office of Fusion Energy, Department of Energy, Washington, D.C. 20545 I wish to add my welcome and that of the (Department of Energy to Mel's. Exchange visits of the past have proven to be valuable to the participants from both countries, and it is gratifying that this is the first to be held urder the newly inaugurated series with the Government of Japan. It may not be coincidental that the topic for this conference is in the area of alternative concepts. The diversity of our national program is probably matched only by the Japanese program among the several national fusion pro'jrams worldwide. Further, the focus of the meeting, what we call compact toroids, is indicative of the readiness of the worldwide fusion community to deal with worthy, innovative ideas. In the U.S. program we continuously evaluate alternate confinement approaches for development as fusion power systems. How do we arrive at a decision to launch development of an approach such as compact toroids and neglect others? This is a difficult technical management issue that we are frequently asfced to addresc. The answer is not simple and ultimately must rely on professional judgments. The process for selecting concepts has evolved as technical successes were obtained in the development of the mainline confine­ ment concepts. That introduces a key element in the selection process; we examine and select alternate confinement approaches by taking f'ull account of ;he status of the principle confinement approaches, and we do not undertake selection of alternate fusion confinement as an abstract exercise. Historically, the mainline confinement approaches, the tokamak and irirror, have reached a prominent role in the development program because of demonstrated experimental success in confining hot, fusion-quality plasma earlier than other approaches. However, the flexibility implicit in the physical principles that unite all confinement concepts has oermitted a great variety of feasible fusion approaches to be conceived and pr ODosed for develop­ ment. These are collectively identified as alternate concepts. Any number of them might be chosen for development and ultimately lead to successful fusion power systems. It is taken as an imperative by the Office of Fusion Energy that any attempt to develop all of them with equal emphasis would be detrimental to the eventual technical success of the most promising and an irresponsible utilization of valuable resources. Several courses of action are conceivable in order to optimize the develop­ ment of the highest potential of fusion for commercial aaplication. The one approach that is clearly unsound is to ignore potential advantages that can come to the development program from ideas outside the mainline effort. It 's precisely this factor that is employed in the review and selection process for alternate concepts. The selection of alternate confinement approaches for development involves three factors; potential reactor advantages with respect to the mainline -2- approaches, technical feasibility, and the readiness (or timeliness) of undertaking the development with respect to the status of worldwide fusion development. The analysis of these factors is a continuing process; the level of performance by which the concepts are evaluated becomes increasingly -ore demanding as the concepts mature (and as the successes of the mainline affort elevate the general standard for all fusion development). Novel alternate concepts such as compact torolds must be perceived by the fusion community as having the potential of significant reactor advantages with •-espect to the mainline concepts. As the conceot matures and becomes ready -or proof-of-orinciple level tests, the perception must be replaced by quantitative studies supporting the reactor advantages. The compact toroid concept is clearly in an embryonic state. As the research we will near about during the next several days evolves, this concept will be submitted to this scrutiny also. We wish the research success for it is eoually clear that this approach has significant potential benefits for users. Despite the enthusiasm for such a fusion system it still takes the dedicated work of the fusion community to make the promise of it a reality. -3- THE COMPACT T03US CONCEPT AND THE SPHEROMAK H. P. Furch, Plasma Physics Laboratory, Princeton University, Princeton, New Jersey 0854<i Photographs of solar activity, using polarized filters [1J, have pro­ vided indirect evidence of the emission of long-lived toroidal plasma con­ figurations confined by linked poloidal and toroidal magnetic fluxes. Plasma configurations of this sort were first produced in the laboratory by H. Alfven [2) and his coworkers (cf. Fig. 1) and reported at the Second International Conference on Peaceful Uses of Atomic Energy in 1958. The injection of a current layer of highly energetic electrons (an "E-iayer") to produce a ste?dy-state toroidal plasma confined by pololdal field was proposed at the same conference by S. Christofiios [3]. A moderr. variant of this idea, using a neutral-beam-injected, medium-energy ion cur- rem [\], is shovn in Fig. 2. Reversed-poloidal-field configurations of che same form have long been produced successfully by means of ordinary plasma currents in theta pinched — both with and without toroidal magnetic field component [5,6] (Tig. 3) • The common faacvre of all these plasma confinement schemes is chat che magnetic field lines are closed, yet the field generating coil system is not required to iiak the plasma toiroid. The principal variants of this ''compact torus" concept are tabulated in Fig. 4. The present paper is con­ cerned mainly with a brief review of the case 3pol < ^plasma- where che current carriers have poloidal-field gyroradii that are smaller than che scale height of che plasma. Experimental tbeta-pinch plasmas with essentially null toroidal-field component (3 po i » Bt 0T ) have exhibited remarkable longevity in earns oi che characteristic time scale for MHD instability [7,8]. VJhile the ideal .'!HD theory for this configuration does not actually allow all modes to be stabilized, the elimination of th«S strongest instabilities by axial elonga­ tion, combined with the finiteness of the plasma ion gyroradius (cpgi. - a-piasma) na ? be providing effective stability in the experiments. Far Che reaccor applicaeion, the null-B t variant has the obvious advantage of very high beta-value Ui) £ 15, but the possible drawback, of insufficiently stable confinement, since there may be a contradiction between the require­ ments for stability against gross modes and microinstabilities. The fast time scales and high voltages of the fie Id-reversed cheta-pinch formation method (Fig. 2) could be avoided in the beam-injected fie Id-reversed mirror approach (Fig, 3), provided that the problem of ion-current cancelation by the electron-drag current can be resolved satisfactorily [4]. The compact torus with comparable toroidal and poloidal field components (Fig. 1). which has lately coot :o be called che "jpheronak," has long re­ ceived theoretical attention 19,10,11]. the toroidal field is supported by poloidal currents flowing inside che plasma and oust, of course, vanish out­ side the plasma boundary. Ideal MHB stability is achievable when chp. plasma -4- encity is somewhat oblate, as in Fig. 5, and is surrounded by a moderately close-fitting conducting shell (12]. Under these conditions, the stability limit for 3g 3 8TT ({D 2 ^) 1 ' 2 /B| (where 3 0 is the field strength OQ the mag­ netic axis) is typically of order 2-5%, but it can be several times greater in highly optimized configurations [13]. The basic advantage of the finita- 3 tor regime is that stability against all ideal MHD modes can be ensured =ven for the case Bpol <e ap]_ asma , which is congenial co good microstability properties. The main drawbacks are that special provision must be made for generating the toroidal flux and that beta must remain well below unity. It should be noted, however, that a limiting value of aj in the 5-102 range for the spheromak gives as much ((p^)) 1 '* a s a tokamak 8£-value of 50-1002, if the maximum field strength at Che magnet coils is the same in the two cases. This is because the spheromak field is maximal at the plasma center (Fig. 5) while the tokamak field is maximal at the coils. The spheromak is closely related to the conventional reversed-field Z- pinch (RFP) [14): it corresponds to the particular case of null field- reversal and fairly low aspect ratio. The RFP has more shear, and thus is abla to tolerate somewhat higher limiting beta values, but has the drawback of requiring plasma linkage by external toroidal-field coils. Experimentally [15]. Che null-field case of the KFP is found to lie precisely at the transi­ tion point between the "quiescent" state that is achieved with external B tor t 0 and the turbulent state that prevails for external 3 tor > 0. The finite-resistivity MHD kink modes of the spheromak have been studied for large aspect ratio : stability is found to be realizable with optimal current profiles and very close-fitting shells [16] — quite comparable Co the case of the conventional RFP. The low-aspect-ratio limit appears to hav; similar resistive instability characteristics [12], but has not yet been treated for optimized profiles. The stabilization of che resistive inter­ change mode in the spheromak depends mainly on collisionlessness and modera­ tion of the beta value [17]. (For the B tor • 0 version of the compact torus, the resistive MHD analysis has been carried out thus far only for ;he special case of axisymraetric modes [18].) The experimental study of spheromak plasmas began with Aliven [2]- It has been resumed recently, by means of the same coaxial-gun technique [19, 20] (51g. 1), as well as by means of a theta-pinch formation process £?ig- 3) that includes toroidal-field generation [21]. References 22-24 describe a new type of "quasi-static" spharomak-fomation technique chat is ainied at avoiding the high pulsed powers associated with a dynamic forming process in plasmas of reactor size- The degree of gross stability observed during the limited pulse time of the theta-pinch spheromak experiment [21] has been c :cellent — or even too good, since the ideal MHD theory clearly predicts a tilting mode [12] for prolate plasmas of the type '..f Fig. 3, whether a toroidal field component is present or absent. The antitheoretical stability against tilting has also been noted experimentally in the B_ ol >> 3 cor case [7,3]. Very recently, however, the injection of a gun-produced spheromak plasma into a prolata conducting shell has exhibited the predicted tilting [20]. The suppression of the tilting node, as well as of higher surface modes, may be a consequence of the presence of hot plasma at the separatrix (cl. Fig. 3) and jus:, outside it. An effect might be expected vhen the ion gyro- radii are large or when the field outside che sec^ratrix recaius finite shear, due to the presence of external axial plasma currents or helical windings [12]. External-plasma stabilization could turn out to be very Important for the compact-torus reactor application, since stabilisation by close-fitting con­ ducting shells would be inconveniently restrictive, AS is shown in several reactor studies [23-27], one of the attractions of the compact toi'us is its potential ability to undergo compression, expansion, or displacement, frae of mechanical constraints. ACKNOWLEDGMENT This work supported by US Department of Energy Contract So. SY-76-O 02-3073. REFERENCES [I] RIDDLE, A. C, Solar Physics 1^ (1970) 448-457. [21 ALFVEN, H., Proc, 2nd Int. Conf. on Peaceful Us.es of Atomic Energy _3_i (1958) 3. [3] CHRISTOFTLOS, N.. ?roc. 2nd Int. Conf. on Peaceful Uses of Atomic Energy .32 (1958) 279. [4] COHEN, 3. I., 3RENGL;, T. A., Particle-Fluid Hybrid Simulation of Field Reversal in A Mirror Plasma, this conference. [5] KOLB, A. C, DOBBIE, C. B., GSIEM, H. R., Phys. Rev. Lett. 3. (1959; 5. [6] KOLB, A., et al., in Plasma Physics and Controlled Nuclear Fusion Re­ search (Proc. 3rd Int. Conf., Novosibirsk., 1968) II_ (IAEA, Vienna, 1963) 567. [7] ES'KOV, A,. 0., et al., in Controlled Fusim and Plasma Physics (?rcc. 7th European Conf., Lausanne, 1975) I_, 55. [3j LINFORD, R. K., and CT Staff, Tho LASL Compact Torus Program, this conference. [9] LUST, R., SCHLUTER, A., Z. astrophys. 3* (1954) 263. [10] CHAHDRASEKHAS., S., in Proc. of the National Academy of Science? A2 (.1956'. 1. [II] MORIKAWA, G. K. , et al., Phys. Fluids _l£ (1969) 1643. [12] 3USSAC, M. N., e_t al. , in plasma Physics and Controlled Nuclear Fusion Research (Proc. 7th Int. Conf., Innsbruck, 1973) III. (IAEA, Vienna, 1979) 249. [13] GAUTIER, P., et al., in Controlled Fusion and Plasma Physics (Proc. 9th European Conf., Oxford, 1979) paper EP 30, to be published. [14] BAKER, D. A., Toroidal Reversed Field-Pinch Experiments, this conference. [15] 0RTOLAK1, S., Some Properties o* the Heating and Confinement in the RF? Configuration, thi.3 conference. [161 GLASSER, A., SELBE?>G, H., to be published. {171 3USSAC, M. S., Bull. ;jn. Phys. Soc. .23 (1979) 872. [IS] BERK, H. L., &- <l_. , Tearing-Mode Stability Analysis ar a Cylindrical Plasma, this cont j rence. -6- [19] TURNER, W. C, et. al., Initial Results of Field Reversed Plasma Gun experiment, this conference. [20] JAR30E, T. R., et al., Magnetized Gun Experiments, this conference. [21] GOL0ENBAUM, G. C, et al., Dynamically Formed Spheromak Plasma (?S-1), this conference. [22] OKABAYASHI, M., &t_ al., Spheromak Equilibrium and Stability and Numeri­ cal Studies of i Spheromak Formation Scheme, this conference. [23] YAMADA, Ji., et. al., Design and Fabrication of the S-l Spheromak Device, this conference. [Zi] AYDEMIR, A., !t al., Two-Dimensional Simulation of the Formation of the PPPL Spheromak, this conference. [25] MILEY, G. a., Field-Reversed Configurations: Theoretical Considerations and Reactor Applications, this conference. [26] KATSURAI, M., YAMADA, M., Preliminary Studies of Spheromak Reactors, this conference. [27] SMITH, A. C, Jr., ejt^ al., The Moving-Ring Field-Reversed Ilirror Reactor Concept, this conference. FIGURES £NS sm MB IBIPir li'.'j'^ r\ ~\ *—-JV^sr c m§?9* Neutral beams Neutral beams Fig. 1. Spheromak genera­ tion, using a coaxial plasma gun with poloidal field at the muzzle. rig. 2. Neutral-beam-driven field- reversed mirror machine. f«El0NI2ATI0N 1IIE7A Plll Ca coil •^—d UADVZ TUBE duo PLASMA SIAS FIELD Htl D HUE ccr.rj ECTioa AXIAL CONTRACTION Fig. 3- Field-reversed tiieta pinch. Fig. 5. Oblace spheromak. Plasma currenc is localized within the solid-fiald-line region. (PPL 786431) Fig. 4. Tabulation of compact torases . -8- INJ"£CTION OF RELATIVTSTIC ELECTRON BEAM INTO TOROIDAL SYSTEMS Akihiro Mohri, Kazunari Narihara, Yukihiro Tomita Institute of Plasma Physics, Nagoya University Nagoya 4S4, JAPAN In recent years, the injection of high current relativistic electron beams into toroidal systetus has been a subject of absorbing interest in connection with compact torus and OH assist startup for large tokamak. Injection methods tried up to the present are surveyed here and a method using Plasma Anode, which is adopted for REB injection into SPAC.is reviewed. The first question we have is whether an electron ring is suited for confinement of fusion plasma in the practical sense. The electron energy of S-layer of Astron should be so high as the relativist:.c factorc-> l OO in order to produce a strong magnetic field and a sufficient volume for confinement. The* resultant strong synchrotron radiation brings a serious problem concerning the energy loss. However, if we use an appropriate mechanism to keep the ring radius large enough at lower &• in the strong external field for equilibrium, such a problem will not be fatal. When a high current REB ring closely in a force-free state is formed, its major radius can'be controlled with an apf-ied vertical field like Astron-Spherator. We do not need the use of ion-ring in this case. Besides, generator of 1—10 Me V REB is now at the commercial base. When the main toroidal current is generated by REB injec­ tion from outside, we can neglect any other current-driving equip­ ment and,thereby, toroidal configurations cf very small aspect ratio are realizable. The concept mentioned above much depends on the success of SEB injection into toroidal systems. Methods examined up to the present are summarized as follows; * ASTRON ( L1L }, RECE-BERTA and RECS-CHRISTA ( Cornell Univ. ) inject REB obliquely to the mirror axis and stop the axial motion by dissipation. * Cornell uniy.( Gilad, Kusse, Lockner ) divert toroidal field lines to the cathode of a diode by using self-field of the diode current. * Physics International (Benford, Ecker, Bailey ) use a guiding field and the drift motion of REB. * Cornell Univ. and Maryland Univ. use a cusp field• -9- all of above methods are for injection into neutral qas. However, adjustment of the initial density of plasma at the time of the injection becomes necessary so as to control thea value of thus confined plasma for stability. In the case of the injection into neutral <?as, the range of the gas pressure should be chosen from the condition of space charge neutrality. Usually, the oressure is 0.1 to 1 torr. For the reason, REB injection into a plasma of appropriate density is required, where ions of the plasma easily cancel the space charge of the injected beam elect­ rons. In order to catch a ring immediately after its formation in an equilibrium position, we have to suppress the induced return currents which mask the poloidsl field of the ring. Besides, we can not change the vertical field within a time comparable with the pulse duration of injection. These difficult points can be solved if we use Plasma Anode instead of usual foil anode. Figure 1 shows a schematic explanation of the function of plasma anode. csthsde iilasaa ct-jatoer vaii Fig, 1 A cathode is set inside the toroidal chamber so that its face is directed parallel to the toroidal field. The toroidal and the vertical fields are applied before FEB injection and then the chamber is filled with a plasma. The plasma contacts both the cathode and the chamber wall. When a negative-pulse voltage is applied on the cathode, ions of the plasma are accelerated towards the cathode and bombard the cathode surface where multi-layers of gas molecules are oresent. Thus, a high density plasma is producsc on the surface. There appeals a plasma sheath of double layer type and the electrons of the cathode plasma a~e accelerated in the sheath. The olasma-anode is very superior to the conventional -10- foil anode in that there is no need to exchange foils and thus a repetitive operation becomes possible. Owing to the thin sheath, fairly high current density is obtainable for REB without causing a pinch, and the beam electrons are ejected in the same direction. The induced returr. current can not flow towards the cathode, and the current is forced to take its path towards the chamber wall along the magnetic field lines. inside the beam channel there exists a strong self-maguetic field due to the beam and the resultant magnetic field becomes helix because of the presense of the toroidal magnetic field. The channel winds round the toroidal major axis. Another possible path of the return current is on the just outside of the beam channel as shown in Fig.l. This return current wraps the beam channel like a thin sleeve and masks the self-field of the beam. Since the skin time of the sleeve is very short, neighboring paths of the beam channel merge themselves as the self-field appears out of the sleeve. Finally, an axially symmetric REB ring with a single magnetic axis is formed. The rise time of the poloidal field is therefore very short. The time is about 150 ns in the case of SPAC-V, as shown in Fig.2. Strong plasma heating could be expected Fig- Rises of poloidal fields on the upper and the lower sides of a REB ring in SPAC-V. pining •MMBUHH WMMMMMWmm 500G 5OOG u M^MMW. 50 NSEC/D during this merging phase. While the plasma-anode is working, the cathode is subject to ion bombardment. During the time, the material of the cathode is evaporated or sputtered ind may enter the confined plasma region as impurities. Figure 3 pr-i-sants photographs of the cathode surface, taken with an electron-scanning-microscope. Damage of the surface like chunk3 are observed. Materials stronger against the ion bombardment should be used such as Mo. In this injection of REB, the cathode plasma is easily produced by ionizing attached molecules on the cathode surface. The density of the molecules is more than 10 14 cm - 2 which is sufficient for the purpose. It does not need to use the plasma of the cathode material itse.f• -11- 50 jj Fig, 3 Damage of the cathode surface. Irradiation power of protons: x Q 7 W/cm 2 , Number of irradiation : '85 shots, Total irradiated energy : 30 J/cm 2 . Cathode : SUS-304 -12- TH E LASL COMPACT TORUS PROGRAM** ft. X. Linford and CT Staff'"'-, Los Alamos Scientific Laboratory, Los Aiamos. New Mexico 875A5 INTRODUCTION The Comoact Torus (CT) concept includes any axisymmetric toroidal plasma configuration, which does not require the linking of any material through the "lole in th.. torus- Thus, the magnet <-oils, vacuum vessel , etc., have a simc-le cylindrical or spherical geometry instead of the toroidal geometry required for Tokar.iaks and RFP's. This simplified geometry results in substantial engineering advantages in CT reactor embodiments while retaining the good con­ finement properties afforded by an axisynsnetr ic toroidal plasma-Field geometry. The cross section in Fig. 1 of a prolate CT shows the essential features of the 5-fields and plasma dimensions. CT's can be classified into three major types bi rising the ion gyro radius p. and th^ magnitude Of the maximum toroidal field 9 The well-known Ascron configuration is a CT with large -., or as is de­ fined in this paper 0 /a > I (see fig. I). The two other classes of CT's. Shercns Ki and Field Reversed Configurations (FRC's), both have ;./a < I. How­ ever, 3 =0 for an FRC in contrast to B = B (maximum ooloidal field) for t" 1 . tm pm a SDneronak. F> °. 0 G ft At', OUTLINE The LASL CT Program is focused on the study of the Dhysical properties of Spneromaks and FRC's in order to develop viable reactor embodiments. The "rain 'acility for this study (Fig. 2} is under construction. The two types of CT's •Vork -jer^or-ied under the auspicies of the J. S. Departnent of Energv. '••U. T. Armstrong, R. R. Bart SCh, ?.. J, Cy. Hsso. Z. A. £k Hat-.'. S. -<ei\ns. J . J. Hoi da, T. P.. Jarboe, J. Lipson. J, "arsnall. Jr.. .<. r . :tc Ker.na, 3. A. ?iatts, A. R. Sherwood, £. J. Sher.vood, ?. E. G:«-on -13- will be aroduced bv the Magnetized jun (Soheromak J and c.ie "ft X-C (FRC) . Eventually, both sources will be able to inject CT's into the central CTX tank where they will be trapped in a dc -nirror field of up to 10 k G. Stability, transport, and heating studies will be carried out on these single ceil configurations. Stability and transport studies of multiple cell configurations will also be carried out in a modified version of the 5 "> long Scylla IV-P nheta pinch. 3ss __<—j •m-C GKMTIM Mx-c cowntuctiw UUL* *'-£ -**o» -•7' «.Wl H * —™i»H>? OH IHMMcf' 'MM In areoaration for these experi­ ments, Spneronaks are being produced in a Gun P-otocype faci'ity and FRC's are being studied in two I m long cheta figure 3. CT Program ai-ich svstens, FSX-A and FRX-8. Plan Figure 3 shows che time scales for these experiments. As indicated, the Gun will be the first pi- sma source tested in the CTX tank. The FR.X-C will be operated as a separate facility during FY 37, At the.and of that fiscal year, the transition section will be added to allow the translation of the FRC plasma into tht CTX tank. The significance of the planned experiments on FRX-B systems is discussed later. The remainder of Che aaper is devoted to a brief status report.on the Gun, CTx , and FRX exaeriments. Companion papers in rnese proceedings ' ' ' describe in more detail the experi­ mental and theoretical results and future plans for these systems. SJN PROTOTYPE A 70 cm long Marshal 1-type coaxial plasma gun has been constructed with electrode radii of 15 cm and 10 cm. Slow risetine magnet coils have been placed inside the inner electrode and outside the outer electrode. These coils nraduce a radial field across Che muzzle of the gun, which is stretched by the emerging plasna and eventua Mv forms the poloidal field of the Spheromak. ir. addition, these coils produce a bias field in the gun barrel (between the electrodes) which allows repeatable operation of the gun at nuch lower fill pressures than otherwise possible. Initial fill densities of 3 x 10 to 2 x 10 cm have been used successfully with bank energies of about 30 k J. Magnetic field orooes indicate that Spheromaks nave been formed by the gun and stopped in a 45 cm diameter stainless steel flux conserver with no guide field . CTX The -»ain CTX tank is installed and has been pumped down to I x 10 niarr /»pth one of the three cryo punos . The dc magnets have also been installed but ars not yet connected to the power suoply. The 10 kv ban* for the gun -nagnets and the 60 k V main gun bank are in olace awaiting power SUDDIV and control s 'S tern ins ta I I at ion. -14- FRX SYSTEMS The large body of experimental and theoretical data available from the FRC research cannot be summarized here. However, three important FRC issues are discussed, which include some of these results and also provide some of the •notivation for the experiments outlined in Fig. 3- '!) Gross Stability The oaserved gross stability in the experiments for more than 100 Alfven transit times'appears at odds with simple MHO estimates. Note that as a result of the absence of _oro>dal field in the FRC, the S is high, the safety factor a * 0, and there is no shear. However, detailed ideal MHD calculations, which include the proper geometry effects (high elongation b/a, and low aspect ratio R/a) <ield results that are consistent with the observed gross stability. in contrast tc this ideai MHD stability, the FRC is usually terminated bv a rotationally dri ven^n = 2 (toroidal "node number) mode. The exceotions are the Kurtmultaev results and a few non-repeatab Is shots on FRX-B. Although these observations ars not completely understood, they are largely consistent with the Following theoretical results . The mode is stable if the ion rotational fre­ quency 7:. is kept below the threshold of ,;./?* » 1.4, where ?.* is the ion dia- magnetic frequency. The source of rotation, which drives ?.. beyond this limit, is dominated by particle loss although there are .other effects associated with geometry, flux annihilation, and electron-ion energy equilibration. The most im Dortant result, however, 'is that the mode does not limit the eqer'gy confine- Tent of the FRC. This result, which is supported bv experiments , predicts that a DOut half the e.iergy is lost by transport before the mode goei unstable. f2) Transport Scaling The transport, which limits the plasma lifetime appears to be anomalous. 1H0 turbulance does not appear important because decreasing :. for fixed R and decreases a Finite-Larmor-Radius (FLrl) stabilization but increases observed lifetime. However, the observed scaling is consistent with the results of a !-!/<* 0 transport code w'£h the lower-hy Dri d-dr i f t I'LHD) instapility dominating the transport coefficient . There are several important transport scaling parameters- Variations in R result in the usual surface-to-volume effects. The density scale length l is very important oecause A /P. strongly affects the LHD •''"insnort. In turn. 2 /;. is controlled by the density n, R, and r It where r is the rad'us of the con­ ducting wall. 'n particular, as r /r —1,1/;.— a/:., which is the largest and most favorable value possible. Thus fat plasmas (r'/r - I) are favorable not only because of wall stabilization, but because : - res'ul :' in slower trans­ port. As * consequence, the angular acceleration is decreased and the onset of the rotational mode is aelayed. A major purpose of the planned FRX experiments is to examine techniques for producing fat plasmas (r /r - I) and thus extending the -lasna lifetime. The production of fat plasmas requires the trapping of more bias flux in the cheta pinch during the field reversal. Two techniques for traoping more flux will be -1.5 - tested. ''I'• reversing the field more rapidly ,-arn; (2) trapping the Bias flux .vi c.1 an octacc'e oarrier field during reversal.'' The fast reversal technique nill be tried first on the FRX-C using its 250 '<V loop voltage. The FRX-8 is oei^o modified to rest the barrier field and reconnect ion techniques of <ur:mul'aev. If these prove superior to fast reversal, FRX-C will oe Tiodi- r "iea accordingly. The Multiple Cell experiment is designed to test transport and Stability in the multiple cell geometry- In particular, a study will be made of the effects of increasing i /p. by multiple-mirror confinement of plasms on the open field field lines. (3) Translation and Trapping Translation and trapping are necessary if the FRC plasma is to tie studied in tie dc field of the CTX tank. The formation and translation have already seen demonstrated in FRX-A. rlowever, it is important to use a puffed gas fill Instead of a static fill so that the FRC can be translated into a vacuum. This technique is being studied in conjunction with a quadrapole preionization scnene on FRX-A (see Fig. 3), The results will be incorporated in the transla­ tion and trapping experiment on FRX-3. These experiments should produce all of trie necessary data for the successful rranslacion of FRX-C plasmas inco Che CTX in FY 32. REFERENCES 1. T. R. Jarooe e_t j_L. "Magnetized Gun Experiments," these proceedings. 2. A. T. Armstrong <jt a\_., "Field Reversal Experiments, and FRX-A and FRX-3 Results," these proceedings 3. R. 5. Sietnon and LASL Compact Torus Staff, "FRX-C and Multiple Cell Experiments," these proceedings. '-. 3. C. Barnes ei a I •. 'Compact Torus Theory--,4HD Equilibrium and Stability," these proceedings. 5. A. u. Es'kov si a_i., "Principles of Pla-na Heating and Confinement in a ;omoact Toroidal Configuration," Plasma Phys. and Contr, Mud. Fusion Research, Innsoruck, Vol. II t1373) 187. 5. S. Hamasaki ana R. K. Linford, Bull. Am. Phys. Soc. IU (1575J 1081. L6- INITIAL RSSOLTS OF FTt LD REVISS2D FLASMA 3'JN E.T7E5IMENT* V. C. Turner, C. W. Hartaan, and J, Taska Lawrence Liveraore Laboratory, Livermore, California 94550 A. C. Smith, Jr. Pacific Gas and Slactric Company, San Francisco, California 94106 '*e have begun experiments using a magnetized co-axial plasma gun to produce field reversed piasaa>^'2) jj, e ^j, iajeccs axially into the Beta II facility at Livermore, The experimental program consists of three stages: (1) formation, (2) translation and magnetic mirror trapping and (3) neutral beats heating of field reversed plasma. Experiments to date pertain to the first stage and have demonstrated production of field reversed plasma. The experimental apparatus is shown in Figure 1. The co-axial gun electrodes are i.5 a long. The diameter of Che inner (outer) electrode is .15 ra (.30 m). The electrodes are fitted with solenoid coils that, together with the guide coils of Che 3eta II device, form a magnetic cusp at ths gun muzzle. The inner electrode solenoid provides magnetic flux that is opposite the guide field (field reversed flux). The outer electrode solenoid is used to control the bias field between electrodes. Diagnostics include an array of magnetic loop probes and a Cu calorimeter. The external magnetic loops consist of four single turn diamagnetic flux loops, two Sogowski belts and a loop that measures the azimuthally averaged component of radial magnetic fieic at one axial location. An axial cagnecic probe can be scanned vertically through the interior of the plasma and contains small loop probes to measure local components of magnetic field. The magnetic probes shown in Figure i ara center tapped for.differential readout, stainless steel jacketed, and compensated with passive XC networks. The probes are calibrated with pulsed solenoid and Helmhoitz coils. The data in this paper were obtained by choosing a guide field strength 4.8 k G, a net inner solenoid flux 0.90 k G - cm-, a plenum gas fill of 50 ata - cm 3 Di' 3 "* 1 chen scanning the gun bank charge voltage (V G ) and bias field between electrodes (3i,i as ) to find optiauo vaiuas for production of fieid reversal. These values turned out to be VQ » 35 k V and 3;,i as • 2.25 k G in the same direction as the guiie field, opposite the field inside the inner electrode. Gun discharge voltage and current are shown in Figure 2. The g-jn current peaks at 940 k A, producing a magnetic field of 25 k G at the surf e of the -.nner electrode that is accelerating plasma out of the gun. The .ime integral of voltage and current shown at Che bottom of Figure 2 gives 130 kj energy input to the gun terminals. Time of flight data from the four diamagnetic flux loops show an initial plasma front velocity in the range 82 to 130 x ICcm/sec, filling the region between ths gun and calorimeter with plasma in approximately 2 microseconds. Thereafter signals on "he diamagnetic loops persist for about 25 usee. For the conditions described above the Sogowski belts meas... - isro net axial current dra-.-n from the gun to the calorimeter and returning ir. the vacuum chamber walls. Figure i shows an axial magnetic probe signal giving *Work performed by ILL for USDOE under contract V-7405-tng-43. -17- a peak change in axial magnetic field strength £ji z • 18.9 k G and field reversal f acxor _3,.'3 Q - 4.2 in the 4.3 k C guide field. The hoc lata half of Figure 3 shows separately signals from each half of the center zapped probe. The two components are equal in magnitude, opposite in sign, verifying their magnetic origin. This verification is equivalent to flipping probe orientation or changing sign of all magnetic fields and observing opposite polarity probe signals. Electrostatic signals would have the same sign for each half of the probe and obviously have been reduced to negligible proportions by shielding and isolation. Typically a time integrated energy deposition of 15 k J is recorded on the calorimeter. A radial .«can of the axial component of magnetic field strength is shown in figure 4. Data are plotted on successive shots, moving the probe between shots. Each data point plotted is a five microsecond average over the peak of the recorded waveform. A multi-channel radial probe has besn constructed ;o verify the profile in Figure 4 on a single shoe but no data have been taken yet with chis probe. Several points can be made from Figure 4. First, a field reversed region with B z • -10 le G, corresponding to •J3 z /3() • 3.1, extend* out to a radius 0,10 m. Second the radius of the axial field null is 0.15 m. Third, the magnetic flux inside Che null is 4.9 x lfl3 ^c - cm i t , factor 5.4 times the net flux inside the inner electrode. A flux amplification effect has been noted earlier by Alfven and his eo-uor'*ers. <•- ><3) Fourth, the radius enclosing :ero flax is 0.23 o. Fifth, the radial profile in Figure 4 conserves the vacuum flux out co the wall at 0.75 m. Sixth, the axial magnetic field energy inside the separatrix is ;8 fc J/a. Equilibrium of che profile in Figure 4 requires existence or a ceroids 1 field component since the axial field pressure inside exceeds that outside the plasma by a factor of four. Toroidal field components have been observed but a radial profile verifying the eauilibrium has not been obtained yet. A scan, of .13 Z on axis versus net flux in the inner gun electrode is shown in Figure 5. For negative values of net fiux che inner electrode solenoid is not pulsed strongly enough co reverse the guide field flux, and the plasma produced does not reverse che guide field. Oace :he net flux is positive large field reversals are readily produced, but if che net flux is too large, field reversed plasma does noc emerge from che gun. (1) H. Alfven, ?roc. Second Int'l. Conf. on Peaceful Uses of Atomic Energy, 21, 145^(1958). (2) 3. Alfven, L. Lindberg, ?. Mitlid, J. Nuel. Energy, l, 116 C1960). (3) L. Lindberg, C. Jacobsen, Astro. J., 133, 1043 <196l7. MOTICI This reporl *34 prepared is in iccounl ol' work sponsored by the United Sulci Government. Sinner :he United Sulci not ihe Untied Suies Department ot Energy. not iny ol :neir employees, nor iny •-' :.leu :onr:ac:ors. uioiortractorv 31 :!ieii employees, milees w> *arrjn Tv express jt implied, jr assumes inv -essi !upi Jir. or responsibility '^t :r.t icr-jiacy. .omplerertcs. Jr ,*»ertiinest v jny in!jrmation. apparatus prouuc: n froucn jiwiuwa. 31 .eprdiirirt :!ul :s me *ouid not inirincc pnvaie!y«su.nea rieim. Refcience to 1 ;on-,p.un sr produc rat Joes not '.mpis jpntovjl it recommend-ation si :he ^rjdu L' ?i IK '.'m.er«:> 01 Cmtornu ot :he i. J Department 01 Enercy 10 :<ie J\wiu«ull M ."inert ir.ji may ac suiuoie. TYPICAL MAGNETIZED PLASMA GUM VOLTAGE AND CURRENT Plitwia IMMJIIH lo luavu 1— 20 30 ^0 Tunc, miciosecnm Js .U i -M crm o • .j,—^-jrfe-^ II LC=L:ZI| fitp i—te^i 07 m 2 to 3 > c z > a z o o OT m H C •o -i9- Figure 3 TYPICAL FK'LD REVERSAL St GNAL FROM BALANCED MAGNETIC PROBE 10/9/79 Shot 25 A-B Figure 4 RADIAL SCAN OF AXIAL MAGNETIC FIELD 11/1/79, 10 ^^i!2.^79 5 I ., ,. ,. l: ii ^ Sensitivity « 6.5 KG/voit Vacuum B 0 * 4.5 k G a* -5 -10 -15 Of- iai 0<r<C25em Vacuum field / _ / / / — / t J m / i • p—e—sr 10 15 20 25 30 R, cm A only ^-V"^ B only i——. 5.5 it». •» - „-»»MMm V 1 * - - i i ••!•----• -••-j 5 - X (bl 20 < r < 50 cm j Flux conserving wall — at R * 75 cm Radius of diamagnelic loops AS L 15202530 35 40 4550 R, CD! Figure 5 AXIAL MAGNETIC FIELD PROSE SIGNAL VERSUS PLASMA GUN INNEFI ELECTRODE FLUX 11/6/79 data , Bias field • 1.5 — 2.3 k G between gun electrodes Guide field = 4.8 k<3 130 kj input to gun terminals 20 m « 10 I __ 1 • 1 • • • -• -• - • s- S, K - 4.8 k G " -/ - • ^^ — 1 • • « -i x 10 3 o ; x IO 3 Inner electrode flux, k C — cm 2 2X 10 3 -20- EXPERIMENT ON PLASMA CONFINEMENT BY INTENSE RELATIVISTIC ELECTION BEAM RING Y. Tomita, K. Narihara, T. Tsuzuki, M. Hasegawa, K. Ikuta, and A. Mohri, Institute of Plasma Physics, Nagoya University, Nagcya 464, Japan 1. INTRODUCTION Toroidal magnetic configurations formed by injection of intense relativistic electron beans (REB) have many benefits for plasma confinement. The formation time is so short that the state can pass over to a final stable one, quickly going through dangerous, unstable regions. When a beam is injected into a preformed plasma, the plasma is effectively heated in a short time. Since no equipment to generate toroidal currents is necessary around the major axis, compact toroidal configurations are realizable. An experimental program named SPAC has been conducted to inves­ tigate the possibilities mentioned above [1-3]. In SPAC-V, REB rings with a current of 30 k A and milliseconds life were success­ fully formed using a new injection method in which plasma acts as anode and no anode foil is needed. Encouraged by the results from the SPAC-V experiments, we have constructed a scale-up toroidal device SPAC-VI which was designed aiming at longer life and higher current of REB rings. In the new toroidal device SPAC-VI, the strong adiabatic compression is easily applicable, when the beam electrons and the plasma are both energized. In the next section (Section 2) a summary of the results of the SPAC-V experiment is given. In Section 3 the design and experi­ mental results from SPAC-VI are presented. 2. SPAC-V EXPERIMENT Experimental setup is schematic*_ly shown in Fig. 1. The end of the magnetically insulated transmission line (MITL), which is connected to pulsed high voltage generators PHOEBUS II or III, is inserted inside the vacuum vessel where the cathode is set with the surface towards the toroidal direction. After applying the magnetic fields, the vacuum vessel is filled with a prepared plasma. The plasma contacts with the surfaces of the cathode and the vessel. On the application of the voltage, a plasma sheath forms in front of the cathode. Across this sheath, elec­ trons are accelerated to the applied voltage and enter into the plasma. This method of beam injection, which we call "Plasna- Anode Method," is superior to the conventional foil anode in the following points: (I) There is no need for exchanging foils and highly repetitive operation is possible. (2) Electrons are -21- scarcely momentum scattered by the anode material, (3) By changing the plasma density and the area of cathode, the diode impedance is easily changed- (4) The small thickness of sheath enables fairly high current density beam emission without causing a pinch. In the experiment, the cathode was 4-6 cm in diameter and the plasp'. density around the cathode was l O^-l O 11 cm" ! . Features of the REB rings thus formed depend largely on the ex­ ternal magnetic field strengths and on the plasma density before REB injection. When REB are injected into the denser plasma, a ring current as high as 150 k A was obtained. The lifetime, however, was as short as 20 microseconds. The lifetime of the REB ring has been prolonged up to 1.6 msec. Figure 2(a) presents time dependences of the ring current I R , poloidal magnetic fields at the center Bp^ and the major radius of the ring Rp. The ring current reached 42 k A 300 nsec after the injection, and it decayed fairly quickly for 200 microseconds. Then, the current decayed slowly until a final large disruption occurred The major radius decreased in the rising phase of B v and unen the radius was nearly constant for about 1 msec. A train of stepwise damps of current were observed during the fast decay of the ring current in the early stage. At the final large disruption, intense x-ray flash emitted from the central region and a positive spike appeared in the poloidal field at the center. These facts suggest that 'the collapse of the ring vas accompanied with the fast shrinkage of the major radius. Figure 2(b) presents the case where the toroidal magnetic field was adjusted to decay faster. In this case, disruption occurred at a faster time than in the previous case. This fact suggests that the life of the ring is determined by the d-scay of the toroidal magnetic fields. The vertical magnetic field necessary for the equilibrium is 3 paff = 3 p + f Y 2 -D 1/2 i* ' I A = Alfven current. where a is the minor radius of the ring, 1^ the internal in­ ductance, and 3 0e ?f is the effective poloidal beta which in­ cludes the centrifugal force effect of the circulating electrons. From the experimentally measured values of B v , i R , R p , and the roughly estimated value of a, we can estimate 3 pe ff". In the -22- case of Fig. 2(a), 3 oe ££ increased from 0.9 at t = 0.4 msec to l. S at t = 1.2 m Sec T This increment could be caused by the heating of plasma due to beam plasma interaction and/or the ac­ celeration of the beam electrons embedded in the plasma. From a micro-wave interfero.-r.etry , the averaged density of the confined plasma was estimated to range around several times , n 13 - 3 . 10 cm From the Doppler broadening of impurity lines ion temperature was estimated to be about 100 ev at the early phase of the ring current evolution. 3 a rig. i. sefit«*eic dijr:* 3 a£ th« tsr«id*l davlca SPAC-V. *t • » <» vi Fig— . Tina variations of the external eiaenetic fiel4-(D v ,3 c ) «id th« net current fl J. tts paloidal r-jq-ietic riei J . '3 1. and the T.ajor radius (fl J in eh« uae of <a> slower decay and (b» faster decay of B fc . -23- 3. Design and Experimental Results of SPAC-VI A new ioroiial davice S?AC-VI was ccns Cruc^ad contiguously from previous experiments • SPAC-V. This new toroidal device was designed with aiming ac plasma confinement by REB rings of longer life and higher current and stronger major radial adia- batic compression. For this purpose,the vacuum chamber was made larger and vertical magnetic field coils were distributed radially. Figure 3 shows the schematic structure of SPAC-VI. The toroidal field B t is generated by the current(1.2MA maximum) in a single center conductor sat along the major axis of the torus and its flat top time within ±7% change is 20msec. The verti­ cal magnatic field coils are installed to keep the decay index less than 1.1 for stabilizing the positional instability of RGB rings extending the whole space of plasma confiment and a maxi­ mum vertical magnetic field" is generated 12k G. For the purpose Fig- 3 Schematic structure of SPAC-VI -26- of major radial compression of plasma, the wave form a£ the vertical magnetic field B v is able to be adjusted. The vacuum '."assal fsr plasma car.fir.er.ent is mada of thin stainless steel with a shell effect. This vessel 1.28m in outer diameter does not have a flat lid aiming at adiabatic compression of confined plasma. In this'experiment, for RES injection "The Plasma Anode Method" was adopted to form KE3 rings. Pulse of negatively high voltage delivered from a Maxx generator PHOEBUS-HI (Pulserad-445Wr 1.8£l V,37Qk A,5Qk J) is sent to a cathode through a magnetically insulated transmission line(MITL). This cathode is inserted inside the vacuura vessel and is set with the sur­ face towards the toroidal direction. The MITL can be operated in ultra high vacuum and the transmission efficiency is more than 3G% at the pulse strength of 1.6MV and 190S«A of SOnsec du­ ration. A partialis ionized cold hydrogen plasma, which is prepared by a coaxial plasma gun, serves for the plasiaa-anode. The experiment of plasma confinement by REB rings using SPAC-VT and PH0E3US-:n (Pulserad-445W) was firstly carried out to have a long lifetime. The external magnetic fields of SPAC- VI can be maintained ten times as long as in the case of SPAC- V where the lifetime of H£B rings had been determined by that of the external magnetic fields. In a high q mode operation (q>2), the ring current I R continued for 10msec which was ten tines the previous record"of SPAC-V{Fig_4) . The peak ring Fig. 4 Wave form of a REB ring current current was 4 3k A and the major radius of REB rings, which was decided from measurement of poloidal field near the wall using magnetic probes, was 23cm, where the toroidal magnetic field B*- was 5.4k G. Irthis case, the slowest decay time [dij/(l R -dtl ]"'" was 50msec. The injected RE3 had the particle energy of -t.l Me V and the current of T-SOJCA. However, highly repetitive short bursts of boch x-rays and impurity lines appeared during the decay of the ring current. Origin of the phenomena is now in full pursuit. -25- The major subject of this experiment using the toroidal device SPAC-VI is to form long-lived REB rings with very low q-value and to apply a strong adiabatic compression on them, because this would lead to a compact torus reactor- Compression ratio of 2.5 with respect to the major radius is possible in SPAC-VI. Experimental results in these operations will be reported. Diagnostic instruments which can work in strong x-ray environ­ ments are under commission, so that detailed parameters of con­ fined plasma will be found out. REFERENCES [1! MOHRI, A., MASUZAKI, M., HARIHARA, K. r HAMANAKA, K., IKCTA, K., in Plasma Physics and Controlled Nuclear Fusion Research (Proceedings 6th International Conference, Berchtesgaden, 1976) Vol. ITI, IAEA, Vienna (1977) 395. [2] MOHRI, A., SARIIIARA, K., TSUZUXI, T-, KKBOTA, Y., TOMITA, Y., IKUTA, K., MASUZAKI, M., in Plasma Physics and Con­ trolled Nuclear Fusion Research (Proceedings 7th Inter­ national Conference, Innsbruck, 1978) Vol. Ill, IAEA, Vienna (19 79) 311. [3] NARIHARA, K., TOMITA, Y., TSUZUKI, T.-, MOHSI, A., in Pro­ ceedings 3rd International Topical Conference on High- Power Elect, and Ion Beam Research and Technology (Novosibirsk, 1979) to be published. -26- INTENSE RELATIVISTIC ELECTRON BEAMS IN TOROIDAL MAGNETIC GEOMETRIES V. 3ai1ey, J. Benford, R. Cooper, B. Ecker, and H. Helava, Physics International Company, San Leandro, CA. 34577 Intense relativistic beams (REBs} can perform important functions in toroidal fusion devices. Among these are: (1) starting up and/or maintaining the current in a steady-state toroidal reactor,'"' (2) produc­ ing the confining magnetic fields for the plasma, 3 " 6 (3) rapidly heating the plasma to ignition conditions,7 (4) providing start-up plasma heating and current maintenance for a steady-state, compact, high $ torus.8,9 The REB is injected parallel to the toroidal magnetic field and trapped in the toroidal plasma column by drift-injection, energy loss trapping J 0 " This technique, which has been demonstrated experi­ mentally,^ allows multi-turn injection, whereby the injection time is many times the single transit time of electrons around the torus, and has been shown experimentally to be efficient. This method uses toroidal- plus-betatron-type field geometry with trapping resulting from partial beam energy loss by self-magnetic field generation and/or by heating of the plasma. The increase in the net toroidal current causad by the REB is an increasing function of the kinetic energy of the injected REB and is inversely proportional to the inductance per unit length of the beam- plasma system. The predicted scaling of the increase in net toroidal current compares well with recent experimental data. The time scale for this increase is significantly shorter than the normal L/R decay time of the plasma. The fraction of the injected REB energy that is initially converted to poloidal magnetic field energy or plasma thermal energy can be adjusted by charging the parameters of the injected beam. This allows a great deal of flexibility in using the RES for plasma heating and current-drive. The average " 3 power level required to maintain the toroidal current in a stecjy-state toroidal reactor is only slightly larger than the ohmica 1 ly dissipated power and is significantly less than that required by rf current drive. For a typical tokamak reactor the ratio of the toroidal current to the average REB power required to maintain that current is 10 A/W. The generally accepted value of this ratio for rf current drive is 0.1 A/W.T4 injection of a 6.6 MV REB having an average power of 1.5 MM could maintain a toroidal current of 15.6 MA in a tokamak reactor, while 150 WW would be reauired if rf current drive were used. In this example approximately SO percent of the injected beam energy is converted to increased poloidal magnetic field energy. The higher efficiency of the RE3 current drive permits the larger reactor -27- Q-value (Qmax = fusion power/current drive power) for a steady-state reactor. Stab Te beta of tofcamaks can be increased by either Towering the aspect ratio A or by intensive heating during the formative stages of the discharge to control the growth of the ballooning modes.1 5 >>6 Jassbyl? used the small-aspect-ratio technique in his design for "SMARTOR" and PITR, and Peng9 achieved sim'lar results by compressing a large-aspect-ratio plasma. The intense heating technique is used at Columbia in TORUS II and at Nagoya in SPAC-V to obtain a high-beta (> 10 percent) stable equilibrium. These high-beta configurations may best be realized by repetitive injection and trapping of a REB in a torus. Starting with a cold plasma and no toroidal current, a pulse sequence can be defined to bring the reactor to ignition at constant poloidal beta, without ohmic heating coils. In principle, this beta can be chosen large enough to be in the stable regine for the balloon­ ing modes, L'sing the high-beta equilibrium of Peng and Oory, me have selected a r-ppact steady-state toroidal reactor driven by fleas.8 The reactor has a low aspect ratio [2.0), high beta (0.18), a 1.6 m major radius, ana 250 MWe net outnut. It is started with 225 oulses of a 1 MV, 0.4 MA, 2 -JS beam at 100 Hz, and current is sustained b\ the same beam operatinq at 0.33 Hz. REFERENCES 1. K. Ikuta, Jap. J. Appl. Phys., M_, 1684 (1972). Z. '1. Bailey, Steady-State Current Drive in Tokamaks, woriunop