© 1985 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE. IEEE Transactions on Nuclear Scmce, Vol. NS-32, No. 5. October 1985 REPETITIVE PULSE ACCELERATOR TECHNOLOGY FOP LIGHT ION INERTIAL ZOKFINE~ENT FIJsIClN” M. T. EuLtraa Sandia National Laboratories P. 0. Box 5800 Albuquerque, NM 87185 Abstract density required to demonstrace nuclear fusion in the -- _--- laboratory on a single pulse basis (i-10 pulses per Successful ignition of an inertial confinement day). The PBFA-II light ion oeam acelerator3 under fusion (ICF) pellet is calculated to require that development ac Sandra National Laboratories will b~ti several megajoules of energy be deposited in the the first ICF facility to come on line (1986) with an oellet’s centimeter-sized shell within 10 ns. T:!llS output energy in the megajoul* range. implies a driver power of several hundreds of 1000 1% terawaLts and po’wer density around lci0 TSj/cm2. Tr.e 800 Sandia ICF approach is to deposit the energy with ii 600 beams of 30 MV lithium ions. The first accelerator g capable of producing these beams (PBFA II, 100 TW) will be used to study beam formation and target 3% fj physics on a single pulse basis. To utilize this technology for power production, repet:tlve pulsing at B rates that may be as high as 13 Hz will be required. 200 iii This paper will overview the technologies being 3 sLudied for a repetitively pulsed ICF accelerator. AS presently conceived, power Ls supplied by rotaLing 4 machinery providing ‘6 MJ in 1 ms. The generator 100 12% I! output is transformed Lo 3 MV, then switched into a 80 pulse compression system using laser triggered spark R 60 gaps. Tnese must be synchronized to about 1 ns. x Pulse compression is performed with saturable inductor 30% < the output being 40 ns, 1 .5 MV pulses. suitches, 012 34 5 6 7 8910 These are transformed to 30 MV in a self-magnetically insulated cavity adder structure. Space charge ENERGY ON TARGET (MJ) limit& ion beams are drawn from anode plasmas with elecLron counter streaming being magnetically Figure 1. Pellet (target) gain as a function of inhibitec. The ions are ballistically focused into energy on target for various assumptions about the the entrances of guiding discharge channels for coupling of the energy into Lhe pellet. Lines of transport to the pellet. The status of com?onenL constant fusion output (energy on target times gdin) development from the prime power to the ion source are also plotted. The right verticle scale gives the will be reviewec. driver efficiency corresponding Co a gl\ren pelleL gain assuming 25% recirculating power. Introduct ion --_- For reactor operation, additional requirements Successful demonstration of ICF in tne laboratory must be met. The driver must operate repetitively at 1 to 1G Hz. At 10 Hz, the driver must supply nearly coulc lead to economic generation of electric pouer, but extensive engineering developments in all aspect.3 one million pulses per day without failure and with of the reactor will be required to realize this only minor maintenance. It must be made of very reliable components which do not degrade appreciably potential. An ICF reactor would include an accelerator (the driver), a system for injecting with accumulated shots or with time on line. The deuterium-tritium fusion pellets, driver must also be relatively efficient, For 3 M.1 targeted at 1C Hz, the Largeted power is 3!: a reaction chamber to absorb the energy from the MW. Because of losses in the driver, some components are exploded pellet, and a thernal-to-electrical working at substantially higher average power. conversion system. The driver includes an eleccrlcal Tt?.e losses are estimated to be 75%, Implying that 90 MW of po’wer source, a means for converting that power to a heat are being absorbed within tne driver. phcLon or an ion beam, and a system to Lransport the This must be done ncndestructively, which typically means that beam to the pellet. The driver must deposit several the heat load must be spread out. megajoules in Lhe centimeter-sizr pellet within 10 ns. Efficiency is also coupled to reactor economics through lirrltations on If this can be done, it snould be possibl? to ignite a the recirculated power. Recirculated power is that fusion reaction with a net energy yield. Figure <1,2 part of tne reactor’s output used to operate tne illustrates the type of yield Lhat mighL be expested reactor. The recirculatec power must be limited to as a function of targeted beam energy. The type of approximately 25% of Lhe total that is generated. The beam is unimportant, provided only that it deposits product of the fraction of the power recirculatec Its energy in the outer shell of the target pellet. times the driver “wall plug to target” efficiency (rl! From Figure 1, iL is clear thaL conservative times the pellet gain (Q) Limes the tnermal to calculations imply that at least 3 to 4 MJ shoulc be electrical conversion (thermodynamic) efficiency must absorbed by the target and thaL a reasonable gain be unity. Assuming a 35% thermal-electrical expecLation is 50 to 100. Eelow 3 MJ targeted the conversion efficiency, the product -10 must be greater gain becomes a sensiLive function of energy. This than 12. For C = 50, n must exceed 212%. The paper will use 4 ?lJ as its design point. At the efficiencies of a light ion beam driver is estimated present time Lhe major efforts in all ICF research to be ?O-25% which is compatible with this programs are devoted LO achieving the beam power requiremenL. (Other potential drivers, e.g. lasers, may fall at this point.) *This work was supported oy the 51. S. Department of The beam mus’; be capable of being targetec on the Energy under ConLract No. CC-AC04-76DP00789. pellet. Brightness, measured in Tw/cm’/rad2, is a 0018-9499!85/1000-1571%01.00~ 1985 IEEE 1572 rtleasure cf the abill:y of the driver beam to hit its Figure 4 is an artist’s COnCeptlOfl Of a Slngie module target. Brightness is degraded by a number of factors Tke fOUr pulse COmpresSiOn lines are combineo beyond including imperfections in the ion source and saturable inductor L into two 4 feed lines, on one defocusing in flight toward the target. Data4 suggest eitner side of the HV transformer. The pulses from that &am brightness is a strong function of voltage these lines are split into 20 equal (1.5 MV) parts, as stow” in Figure 2. ;i’ has been chosen for PBFA- then added in the induction cavities. Table I gives II and will be assumed for reactors because the estimated energy efficiency for each element in approximately 30 MV is the proper voltage to couple the pulsed power system. The transformer, spark gap the energy from these ions into a pellet and should switch, and saturable inductor efficiencies are give a beam of adequate brightness. extrapolated frcn existing data. The splitter and cavity efficiencies are estimates. The adder Given that 3-4 M; of energy from light ion beams efficiency will be discussed later. The primary must be delivered to the fuel pellet in 10 ns, the points to be made by the table are that the loss per peak power of the beam at the target will be component is small but the overall efficiency is still 50% and that the losses are distributed and therefore 300-400 TN. By controlling the temporal distribution tolerable. of the ion beam voltage, lt should be posstble tc bunch the beam, that is, to compress the pulse E2l r%~s- duration. A factor of four is projected to be FDRMERSSWIys*FE* rCOHHONFzF::;:cow possible while the beam is being transported between the accelerator and target. If we $ssUme 50% efficiency f or generating and transporting the beam to +&%&-q+~ iTUNA tne target, the electrical power pulse must be 15C to 200 TW with a 40 ns pulse duration. It does not appear to be feasible to produce these eXceptiOndllY I I I I I @F,,G $+ $$&+ ,;;; ,~ GTTi$z-i IGNITION c a$ 0 (;A”’ * 0.1 n o PINCH REFLEX 0 c b t, APPLIED FIELD 0.01 0 ‘3 q AMPFION A Figlure 3. Electrica: schematic of the pulse 0.001 t compression section of one 10 TW module. 0.1 0.2 0.5 1.0 2.0 5.0 10.0 VOLTAGE MV Figure 2. Experimental ion diode beam brightness as a function of accelerator voltage. high peak powers in a single pulsed power system. Thus, the pulsed power system will be divided into several mod,Jles. PBFA-II) for example, will have thirty-six, 3-4 TW modules.3 Since the number of beam penetration points into the reactor chamber should be minimized, this paper will consider the &sign of mcdules that generate 10 TW electrical pulses. up to 20 such modules operating in parallel woiild be required for a reactor. -L’ Description of the Driver :I: 4 -_-_ -__-_-___-_-_- .“. :a, ) ,,$I <“, :_:I; ]- ----/-------- .~.. e.Y~-.T” - -- Generically, . the dhiver consists of an energy II accumulator, a control switch which begins the pulse I-- 1--‘- _.- ~._ _ _ ~. J compression cycle, a transformer tc get tc a hign voltage C- 3 MV), some high speed switching to Figure 4.