Technology of Inertial Confinement Experiments

Home , PROTO (fusion reactor), Thermonuclear fusion

IAEA-200

TECHNOLOGY OF INERTIAL CONFINEMENT EXPERIMENTS

PROCEEDINGS OF AN ADVISORY GROUP MEETING ON THE TECHNOLOGY OF INERTIAL CONFINEMENT EXPERIMENTS ORGANIZED BY THE INTERNATIONAL ATOMIC ENERGY AGENCY HELD IN DUBNA, USSR, 19-23 JULY 1976

A TECHNICAL DOCUMENT ISSUED BY THE , INTERNATIONAL ATOMIC ENERGY AGENCY, VIENNA, 1977 TECHNOLOGY OF INERTIAL CONFINEMENT EXPERIMEFNTS Printed by the IAEA in Austria August 1977 PLEASE BE AWARE THAT ALL OF THE MISSING PAGES IN THIS DOCUMENT WERE ORIGINALLY BLANK The IAEA does not maintain stocks of reports in this series. However, microfiche copies of these reports can be obtained from INIS Microfiche Clearinghouse International Atomic Energy Agency Kdmtner Ring 11 P.O. Box 590 A- 1011 Vienna, Austria on prepayment of US S0.65 or against one IAEAmicrofiche service coupon. FOREWORD

The IAEA Advisory Group Meeting on the Technology of Inertial Confinement Experiments was held at the Joint Institute for Nuclear Research, Dubna, USSR, from 19 to 23 July 1976. This was the third Agency Meeting on this subject with the earlier ones being held at the Trieste Center in 1973 and 197V.

The purpose of the meeting was to review the progress in the research of inertial confinement systems, to discuss new ideas in conceptual designs of fusion reactors and particular engineering problems of experiments on inertial confinement.

Considerable progress was reported at the meeting in both laser fusion and relativistic electron-beam fusion experiments. The more important results, conclusions and recommendations of the meeting were summarized by Dr. P. Pashimin and Dr. G. Yonas and, although published in the Journal Nuclear Fusion, are reproduced here. A selected number of original papers prepared by the participants are included in this report.

The meeting demonstrated the importance to fusion research of the exchange of information between scientists from major laboratories thoughout the world.

Facilities for holding the meeting were generously provided by the Joint Institute for Nuclear Research. The Agency also wishes to thank the local committee composed of members of the USSR State Committee for Atomic Energy who contributed to the success of the meeting. TABLE OF CONTENTS

SUMMARY PAPERS

1. Summary of Electron-Beam Fusion Experiments ...... 3 G. Yonas 2. Sum m ary of Laser-Fusion Experim ents ...... 5 P. Pashinin

RELATIVISTIC ELECTRON-BEAM INTERACTIONS

1. REB Interaction Experiments with Plasmas ...... 9 S. Nakai, et al., Osaka University, Japan 2. Transport and Focusing of High-Current REB onto a Target ...... 25 Yu. L. Bakshaev, et al., Kurchatov Institute, USSR 3. New High-Current REB Accelerators at the Kurchatov Institute ...... 41 M.V. Babykin, et al., Kurchatov Institute, USSR

LASER INTERACTIONS

1. Commercial Applications of Laser Fusion ...... 59 L.A. Booth, Los Alamos Scientific Laboratory, USA 2. An Overview of Planning Consideration in the US Inertial Confinement Fusion Program ...... 81 Division of Laser Fusion, ERDA, USA 3. Recent Laser-Drven-Implosion Measurements at KMS Fusion ...... 107 F.J. Mayer, KMS Fusion, USA 4. High Power Glass Laser System "Gekko" for Fusion Research ...... 141 C. Yamanaka, et al., Osaka University, Japan 5. High Power CO2 Laser System for Plasma Research (Lekko I) ...... 157 S. Nakal, et al., Osaka University, Japan 6. Super-Compression and its Stability of Multi-Structured Pellet ...... 173 K. Niu, Tokyo Institute of Technology, Japan 7. Non-Linear Processes in a Dense Plasma ...... 185 A. Samarskij, et al., Institute of Applied Mathematics, USSR 8. Laser Fusion Research in Osaka ...... 203 C. Yamanaka, et al., Nagoya, Japan 9. A High-Power Laser System for Thermonuclear Fusion Experiments ...... 233 Eh. A. Azizov, Kurchatov Institute, USSR 10. Study of the Acceleration of Thin Metal Foils Acted on by High-Power Laser Emission ...... 241 V.N. Belousov, et al., Lebedev Institute, USSR SUMMARY PAPERS SUMMARY OF ELECTRON-BEAM FUSION EXPERIMENTS (G. Yonas)

Research employing relativistic electron beams will involve a 70-cm-diameter cusptransport and at the I.V. Kurchatov Institute of Atomic Energy, beam compression system. Other experiments to Sandia Laboratories, and the Institute of Laser extend the neutron yield through use of higher- Engineering at Osaka University was described. power beams are also planned. A new 2.09 water Before summarizing the work presented here, one Blumlein device called "Kal'mar" was described should point out that there are extensive electron by S. Fanchenko, and this accelerator is to be used beam research programmes in over two dozen in deposition experiments. laboratories throughout the world, and only some The Osaka programme, which was described by of that work was touched on in this meeting. The S. Nakai, has emphasized studies of the deposition point that should be made is that in the last few of beams in CD2 foils. Nakai also concluded that years there has been a great expansion of effort there is evidence for enhanced stopping, but he and interest in electron-beam research. The attributed this effect to beam-plasma instabilities. reason for this is quite obvious, namely that the He reported the observation of neutron yields of physics uncovered thus far by this research is 109 to 1010 with an electron beam current of 70kA extremely exciting and has suggested several incident on the target. Neutron spectrum measure- fusion applications, such as laser excitation, micro- ments were not carried out, and the question of ion wave production, collective ion acceleration, and acceleration was not conclusively addressed. plasma heating. The subjects discussed at this Streak photographs showing radial blow-off veloci- meeting were more limited, however, and only ties of approximately 107 cm ·s- 1 were the major dealt with direct use of electron beams for inertial evidence of enhanced deposition. confinement. In the Sandia work, presented by G. Yonas, In the work of the Kurchatov Institute, which was experimental studies of beam focusing in diodes presented by L.I. Rudakov, the most significant using pre-formed plasmas and results from diode result was the measurement of the velocity of a design codes using particle simulation were shown. pusher material accelerated by electron deposition It was pointed out that the ion current emitted by in a thin double foil structure. The electron a thin layer of plasma at the anode was a critical accelerator Triton was used with parameters of feature of the electron beam focusing phenomena. 500 kV and 100 kA. The pusher (a 10-pm-thickCH 2 foil) Experimental observations involving X-ray and velocity was found to be 5-7 cm-ps - 1 , with the optical techniques indicated that symmetric irradi- essential physics claimed to be the enhanced ation of spherical targets could be achieved with a coupling of the high-voltage electrons into an outer single beam. By comparison of ablation velocity material (5-pm-thick gold). This enhanced coupling data with results of hydrodynamic calculations, it was attributed to magnetic stopping within the was possible to determine energy deposition gold after its initial heating and expansion. This characteristics of beams in the MA. cm 2 - range. technique was then used to compress and heat The measurements showed that power densities deuterium gas, at 150 torr contained within a conical of approximately 1012 W -g1 could be achieved for cavity in a lead block, thus simulating a portion of thick targets but that for targets much less than a sphere. A peak neutron yield of 1-3X106 was an electron range in thickness, the absorbed power reported and was found to depend strongly on density was found to be roughly one order of magni- deuterium pressure and percentage concentration tude higher. The mechanism proposed to explain of argon added to the deuterium to suppress ther- this enhanced deposition was that of a stagnant mal conduction. A strong dependence of neutron beam with electrons able to couple their energy yield on gas parameters tended to indicate that the effectively into the thin material. neutrons were of compressional nature and not It was also pointed out that the exact mechanism produced by beam-target interactions. The signi- of energy deposition is still not well understood ficance of this result relates in some measure to and that self-consistent time-dependent calculations the rather low power of the electron beam used, treating the coupled hydrodynamic and electro- 5X 1010 W, and a programme is underway to extend magnetic effects are being developed. Both the these results to higher-power sources. A 5-MJ Sandia and Kurchatov groups emphasized the accelerator, Angara V, has been proposed and a importance of self-field effects and suggested that single module of this accelerator, which is to be the role of beam-plasma instabilities should be constructed at the Efremov Institute, was described less important. The work at Sandia on hydrodyna- by M. Svinzin. The purpose of the experiment is to mic stability of electron beam targets involving scale these results, and pellet gains of the order of both theoretical and experimental studies was 100 have been predicted with 2 MJ deposited in the reported. The relatively small gradients asso- pellet. ciated with electron deposition in the ablation The Kurchatov approach to dealing with these region appear to result in favourable stability high yields is to employ transport (in a magnetic characteristics; however, additional questions of cusp geometry) of beams from multiple modules, stability at the fuel-pusher interface were pointed and experiments involving beam transport and out and are under study using a 2-D Eulerian compression were presented by V.P. Smirnov. hydrocode. It was also mentioned that implosion Critical near-term experiments will be carried experiments, which would allow one to compare out later this year on the Angara I machine which the rate of neutron production and compression,

3 were needed. It was postulated that such an The subject of light ions was not discussed exten- experiment could be carried out at absorbed power sively, but it was pointed out in the concluding levels of approximately 1013 W and an 8X 10 12 -W session that the potential advantages of light ions accelerator, Proto II, which i.s presently being could also apply to relativistic electrons if deposi- constructed at Sandia, was described. tion enhancement methods were employed. Sandia The Sandia programme has emphasized focusing is investigating the possibility of producing intense of electron beams within the diode in order to light ion beams in high-current diodes and is devise a relatively simple and economically practi- emphasizing the question of ion-beam focusing. cal fusion ignition system. It was pointed out that There is apparently no similar programme on damage to the electrode system would be a major ion-beam generation and focusing outside of the problem and would necessitate the use of relatively United States. low yields and sacrificial layers or other material In conclusion, the results from all three labora- protection concepts. On the other hand, the tories indicate that electron beams can be coupled Kurchatov approach which involves high-yield efficiently to materials with energy densities explosions requires beam transport in a magnetic greater than thought possible using single-particle system which places the burden on beam concen- calculations. Target compression studies using tration at a distance, but also lessens the diffi- these beams have begun, and the Kurchatov Institute culties of achieving an economically practical pellet. has apparently demonstrated the first compres- The long-range plans at Sandia involve the con- sional neutrons produced with electron beams. struction of a 40-TW accelerator, EBFA 1, to be Sizable programmes are underway at the Kurchatov completed by 1979 with a further modification to Institute and Sandia Laboratories to reach break- reach scientific breakeven in the early 1980's. even levels by the early 1980's.

4 SUMMARY OF LASER-FUSION EXPERIMENTS (P. Pashinin)

Substantial progress has recently been achieved energy deposition in targets with spherical in the experimental study of the fundamental physi- geometry. One should emphasize the importance cal processes responsible for the prospects of the of developing experimental methods able to deter- laser-produced thermonuclear fusion. Included mine the exact energy balance taking into account here are the absorption, the reflection and the all the losses. refraction mechanisms for laser radiation in the b) The rapidly expanding efforts in constructing plasma corona, and also the hydrodynamics, energy the next generation of facilities carrying 103 - 10 4 J transfer from the corona to the ablation target in the beam create the necessary basis for the layer, the stability of compression and the mixing decisive step in laser fusion, i.e. to obtain a sub- of target material. stantial thermonuclear yield. This would be a In the domain of the technology of physical very important step forward because today we are experiments substantial progress was also noted. 10 orders of magnitude away from the region where Experimental facilities have been developed laser fusion might be of practical importance. The operating over a wide range of wavelengths and extrapolation into such a remote region is very providing the possibility of doing experiments with hazardous since the great change of energy scale variable geometry and with programmed temporal involved may completely alter the physics of the and spatial structur e of the laser. pulse. The process. technology of various laser systems involving many Speaking about more remote prospects the situa- different principles of pumping has been greatly tion seems to be somewhat different from that of improved and a range of new possibilities of laser the electron beams. We know today how to build techniques has emerged. laser systems for the next step, by the next step The feasibility of producing substantial compres- is meant target ignition or reaching energy gains sion of nuclear fuel in spherical geometry has been Ef /Ei > 1. According to present calculations this demonstrated experimentally. It was shown that will involve a laser carrying an energy of the compression may be stable. The first results 104 to 105 J, closer to 105 J, but from the existing have been obtained on the production of neutrons laser systems one is not able today to indicate the in the compressed volume. It should be noted that candidate which meets all the requirements of a the experimental results do rather well compare laser fusion reactor and is sufficiently economical. with theoretical predictions. That is why the search for a new type of lasers Notable progress has been achieved in laser continues and the development of laser systems plasma diagnostics. Methods for studying fast with all the accessory technologies still remains a processes with high-speed and temporal resolution major problem. The development of laser target have been developed, particularly those involving technology meeting the requirements of economics optical and X-ray measurements for picosecond also remains a major problem. pulse durations. A major step forward is noticeable The IAEA should be encouraged to organize such in the investigation of fuel compression and the advisory group meetings at regular intervals evaluation of the compression ratio as well as during the next few years. The topics for the next temperature measurement using the energy spec- meeting might include, for example, the following trum of particles by the time-of-flight method. subjects: A very important success was achieved in the a) The study of physical processes in laser and technology of target fabrication for laser fusion and particle beam targets including the interaction with in the improvement of methods of target selection radiation, the heat conductivity and hydrodynamic and control. instability. Speaking about plans for the near future and the b) Prospective laser and accelerator systems. problems involved one may note two most important aspects. Moreover, taking into account the fast expansion of the comparatively young approach of inertial a) The necessity of a more detailed study of the confinement, it would be useful to hold a conference physical processes mentioned above and, in parti- on inertial confinement systems under the spon- cular, the processes responsible for an efficient sorship of the IAEA.

5 RELATIVISTIC ELECTRON-BEAM INTERACTIONS REB Interaction Experiments with Plasmas

S. Nakai, K. Imasaki, S. Miyamoto and C. Yamanaka

Institute of Laser Engineering, Osaka University Osaka, Japan

Abstract The interaction of focused REB with dense plasma on solid target has been investigated. The experimental results show the existence of anomalously strong interaction which is favorable for the pellet fusion by the implosion.

1. Introduction The spherical ablation to compress and to ignite thermonuclear burn of a fuel pellet depends upon an efficient dissipation mechanism to deposit most of the electron beam energy symmetrically in the outer thin portion of a target. In laser fusion, the critical density surface separates the absorption region of laser light and the compressed region, which prevents the direct heating of pellet core to ensure the superdense compression. In REB fusion, the stopping length of an electron beam must be shorter than the pellet radius or the beam energy should be absorbed in corona region which may be created from the target material by a fractional energy deposition of a beam. 13 18 -3 In the density regime 10 1310 cm it is possible to heat plasma only by collective effect which may arise from the beam- plasma instability and from the turbulence generated by the return current.( There are also a few experiments which give evidences of both processes. For pellet fusion, we have to investigate REB interaction with higher dense plasma which is affected by the plasma dynamics from target and also by the properties of a focused beam. We have investigated the energy deposition mechanisms when a tight- ly focused beam has been irradiated onto a plane solid low Z target.(2)(3)(4) The results shows the existence of an anomalously stronger interaction than classical one.

9 2. Experimental Apparatus

We have constructed REB generators REIDEN-I, II,and III. They are consisted of a Marx-generator, a Blumlein type pulse forming line and a diode. In REIDEN-I, a paralell plate Blumlein with mylar insulator are used. Coaxial Blumlein of water insulator are used in REIDEN-II and III. Most experiments on REB-plasma interaction have been done by using REIDEN-II and III. The characteristics of them are shown in Table 1. The diagnostic methods for beam and plasma are as follows. The voltage wave form across the diode and the beam current are monitored by a resistive divider and a current shunt of toroidal resistor on the return path from the target respectively. The response times of them has been checked to be less than 10 nsec. The focal spot size of the beam on the target is estimated from the X-ray image of 3 channel pinhole camera with Be-foil of different thickness as an X-ray absorber, with the assistance of the observation of the plasma by a high speed streak camera. An electric image converter framing camera (5 nsec exposure time) is also used to observe the dynamical behavior of plasmas. The electron temperature of plasmas is estimated by 2 channel X-ray absorbing foil detectors of 200 pm and 400 pm Be-foils. The ion velocity is measured by the time of flight method. Two charge collectors are set at the position of 68 cm and 33 cm distant from the focal spot of the beam on the target. The total neutron yield is counted by paraffin moderated activation of Dysprosium which has been calibrated by AmBe neutron source. Neutron energy is measured by time of flight method using two plastic scintillation neutron counter located at the position of 4 m and 9 m distant from focal point of the beam.

3. Experimental results

3.1 Diode impedance and beam focusing The cathode is a plane brass disk with aquadac coating. At the center on it, a guide pin of tungsten of 1 mm diameter is set with the length of 7 mm. Anode-cathode spacing is 15 mm. Typical wave forms of voltage and current are shown in Fig. 1, and diode impedances in Fig. 2. The diode impedance depends strongly on the length and voltage of pre-pulse. With small and short pre-pulse, the diode impedance does not decrease to give the critical current

10 for pinching as shown by curve (1) in Fig. 2. With adequate amount of pre-pulse, the diode impedance goes down initially and then increases as shown by curve (4). When smaller diameter cathode is used, the diode impedance in initial phase is larger and the onset of impedance increase appears earlier as shown by curve (5). The initial decreasing phase of the impedance can be explained by the Child-Langmuir model with plasma cathode expansion( 5)

17V (d-ut) Z = Kvl-- 2

where d, u, and R are anode-cathode gap, expansion velocity of surface plasma and cathode radius respectively. Assuming that the increase of the impedance is due to the beam collapse at the diode center, the pinching velocity can be derived to be 0.1 cm/nsec. A fairly constant impedance can be obtained as shown curve (2)(3). These experiments indicate tight pinching of the beam of about 1.5 mm diameter and deep craters on the target which show large amount of energy deposition in small spot. With longer pre-pulse, the diode shortening occures and no voltage can be sustained across diode gap. The diode current of these examples are compared with the parapotential current.( 6 ) The current of the focused beam (2),(3) and (4) corresponds quite well with this model.

3.2 Plasma production and heating When focused beam bomobards the target, plasma is produced at focal point. The size of bright spot corresponds to that of image of hard X-ray which is identified as the size of beam focusing. Then the plasma expands along the target surface and does not expand toward cathode.( 2) This behavior is due to the magnetic pressure of the beam current. Fig. 3 shows the time variation of measured electron temperature of the polyethylene target plasma assuming Maxwellian velocity distribution of electrons for 200 KeV, 40 KA beam bombardment. The electron temperature at 50 nsec from pulse rise is about 8 KeV. At the later time than 150 nsec it decays exponentially. The voltage pulse ended at about 80 nsec but the current continued to flow for about 150 nsec. During the current flowing, the X-ray due to the beam current overlaps to the X-ray from the plasma, so the electron temperature is overestimated. The real electron temperature of plasma just after the pulse is estimated as 1-2 KeV by extrapolation of decay curve in Fig. 3.

11 The ion velocity is measured by the time of flight method. Using a polyethylene tharget the ion kinetic energy of the carbon is about 1~2 KeV for the case of Fig. 3. For Al and Pb targets, the dependence of the expanding velocity of the ions on the diode voltage is shown in Fig. 4 where the solid line is the calculated value assuming the beam plasma interaction to be classical one.

3.3 Neutron generation When the plasma temperature is so high as a few KeV with the density comparable to solid, we can expect neutron generation by thermonuclear reaction. Only in that case when CD 2 target is used, neutron generation is observed. The neutron signals on the scintillation detectors at the location of 4 m and 9 m distant from the focus point are shown in Fig. 5. The energy of neutron are identified to be 2.45 MeV, which is consistent with that of D-D reaction. The dependense of neutron yield on the electron beam power are plotted in Fig. 6 and Fig. 7. The data in Fig. 6 were obtained with solid CD 2 plane target and that in Fig. 7 were with

CD 2 thin film target of the thickness of 100, 160, 200 and 300 pm. It should be noticed that there are no difference in neutron yield depending on the different thickness of target.

4. Discussion

When an electron beam impinges on solid target, it heats, evaporates and ionizes the target material. The main deposition mechanism of beam energy is via beam-plasma interaction. Assuming that the electron beam of radius r is stopped in range X over which the electrons lose their energy, the plasma temperature T may be given approximatly by the energy balance equation 2 Tr 2.nlkT __ IVT (1) 1 where n1 is the plasma density which interacts with the beam, k is the Boltzmann constanat, I is the beam current, V is the beam

accerelation voltage and T is the pulse length of the beam when it is short enough to neglect the plasma motion or the expansion time of plasma from the electron beam radius. Here the energy loss by radiation and thermal conduction is neglected. There are several possible mechanisms for the energy deposition of the beam to the plasma to estimate the range A. They are the classicalprocessof binally collision and Cerenkov radiation, classical process of binally collision and Cerenkov radiation,

12 two stream instabilities in hydrodynamic and kinetic instability depending on the beam properties, and the return current instability. The energy loss of the beam by Coulomb scattering gives the range of about 1.5 mm for 200 KeV electron in polyethylene target. With this value of range eq. (1) gives about 20 eV as the plasma temperature for the same condition of Fig. 3, which is two order of magnitude lower than the measured value. Such a long range can not explain the observed neutron production with thin CD 2 target of 100 to 300 pm. 19 22 -3 In the high density plasma of 10 10 cm , the electron drift velocity due to return current hardly exceeds the ion acoustic velocity. Therefore this mechanism can not contribute to the energy deposition of the beam electrons in solid target experiments. As for the two stream instability in cold beam-cold plasma, the maximum growth rate is given by( 7)

/ n 2 1 / 3 a = 0.7 n)1 l (2)

where wpl is the electron plasma frequency corresponding to the plasma density n1 and n2 is the beam electron density. Assuming that the stopping length of the beam electron is given by

v X e (3)

where v e is the beam electron velocity, eq. (1) predicts quite reasonable value of plasma temperature comparing to the experimental results. In this calculation, the plasma density n, is taken to 22 -3 be solid density -102 cm 3 . In the solid density plasma the damping rate of beam driven plasma instability is marginal comparing to the growth rate even with the low Z hydrogen plasma at temperature of lrv2 KeV. In general the thermal expansion of plasma can not be neglected when longer pulses than 10 nsec are used. Therefore interaction in the expanding corona region must be considered. If we apply the hydrodynamic instability to the corana region, the plasma temperature becomes much higher than the observed value.

13 In the strongly focused beam the hot beam condition will be fulfilled for the applicability of the kinetic instability,(8) that is

n2 n 21/41/6 Ae max \nlY3 (4)

The growth rate of instability with the beam of the velocity spread Ae is given by 2 02 1 jpi (5)

nl¥np A2+kj c Using equation (1) (3) and (5), we can estimate the plasma temperature, which agrees quite well with the observed value when the 19 -3 plasma density nl is taken to be 5 x 101 cm 3 where the electron beam mainly interacts. The neutron yield can be estimated by

N = !nD

Putting the parameters in experiments for r = 1.5 mm, Tr = 80 nsec, the yield of neutron is calculated as shown in Fig. 8 using the integrated fusion cross section < ov> over Maxwell distribution corresponding to the derived temperature by eq. (1). It is obviously noticed that the neutron yield and the dependence on input beam power are consistent with the experimental results of Fig. 7. In the REB-target interaction, the boundary and interaction region may be very complicated. The self magnetic field of REB may be shielded by the magnetic skin depth 6 = 1/2 (4) -1/2 which is 3 x 10 - 3 cm for the plasma of 1 KeV temperature and the injection time of T = 10 nsec. The density gradient of the plasma in corona region must be estimated including the kinetic and magnetic pressures of the beam, and its balance with the pressure of expanding plasma.

5. Conclusion The experimental results presented here, such as neutron yield, plasma temperature and kinetic energy, show the existence of anomalously strong interaction of REB with low Z dense target

14 plasma. The REB dissipates most of it energy in the thin layer or in corona region and does not penetrate deeply into the target material. This fact is favorable for the pellet fusion by the implosion using the intense focused REB.

REFERENCES

(1) L. I. Rudakov and R. N. Sudan: ICTP. Trieste, Preprint IC/73/124

(2) K. Imasaki, S. Nakai and C. Yamanaka: Jour. Phys. Soc. Japan 37, 881 (1974)

(3) K. Imasaki, S. Nakai and C. Yamanaka: ibid 38, 1554 (1975)

(4) K. Imasaki, S. Miyamoto, S. Nakai and C. Yamanaka: ibid 40, 531 (1976)

(5) G. Yonas, J. W. Poukey, K. R. Prestwich, J. R. Freeman, A. J. Toepher and M. J. Clauser: Nuclear Fusion 14, 731 (1974)

(6) J. Creedon: Physics International Report No. PIIR-17-72 (1972)

(7) 0. Buneman: Phys. Rev. 115, 503 (1959)

(8) B. N. Brejzman and D. D. Ryutov: Nuclear Fusion 14, 873, (1974)

Table 1 Pulse TypeType Impedance Marx Typical Focus Power length Impedance bank performance spot density

REIDEN II Water coax. Ons400kV 300kV 11 WC 2 100nsec 130 lmm 6x10 blumlein 10kJ 20kA REIDEN III Water coax. 900kV 500kV 12 2 blumlein 80nsec 27kJ 27kJ 880kA 1.5mm 2x10 W/cm

15 (a)

(1)

(b)

(a)

(2)

(b)

(a)

(3)

(b)

*,~F^-(a) ---- ~ ^~

(4)

(b)

50nsec/div. (a) Voltage wave form 100KV/div. (b) Current wave form 40KA/div.

Fig. 1 Typical diode characteristics

16 100 -

(1)

10-

I a,

r0 ro

H 1 m

(5) (3)

(

(4)

- 3 . _ I I I I I I I I I I I () 202 I0 80 10 40 120 Time (nsec)100 Time (nsec)

Fig. 2 Diode impedance variation with time

17 10

(I - .- 1 . ~_- 1

- . . . I . 1 ... I ... 0 100 200 300 400 ns

Fig. 3 Electron temperature of plasma as a function of time

18 O Al

O Pb

.Al

10 7 .- II . Pb

a) -O' 1Z-~~~~~0 >n

4-) rz 0

A Al 0 U J0_ Pb 10 6 _

150 200 250 300 100 150 200 250 300

Diode Voltage (KV)

Fig, 4 Ion velocity measured by time of flight method

19 Neutron wave form

------9m 50mV/div.

4m _..... _4m_200mV/div....._

X-ray Neutron

200nsec/div.

Fig. 5 Neutron signal on plastic scintillation counter

20 (CD 2) n Solid Target 0

00 0 5 r 000 0 0 0 0 0 v 0 o0 0 o 2 0O 0O OPlo 00 o ooo 0O 0 108 - s- O0

z cE 5 +1E-I

2 5 2 5 1010

Beam Power (Watt)

Fig. 6 Dependence of neutron yield on beam power

21 (CD2)n Thin Foil Target

0 300pm A 200pm 5. . 160pm ^ 100pm

2- A

o X 5*

* 0'ox0O

2 0X o A -rl 0 X A 0 04 108

o z -4

4Jrp 5 - 0 00

2 0

1010 109 (Watt)

Beam Power

Fig. 7 Dependence of neutron yield on beam power with thin

film CD 2 target

22 10 10 -

1 Np=5xl0 Np=5x10 1 9

10 9

10 a >d

or 4-) zo c,E-i « i-cd 10 8

10 8 10 9 2 5 0101 (Watt)

Beam Power

Fig. 8 Calculated neutron yield assuming kinetic instability

23 TRANSPORT AND FOCUSING OF HIGH-CURRENT RELATIVISTIC ELECTRON BEAMS ONTO A TARGET

Yu.L. BAKSHAEV, E.I. BARANCHIKOV, A.V. GORDEEV, Yu.V. KOBA, V.D. KOROLEV, V.S. PEN'KINA, V.P. SMIRNOV, A.D. SUKHOV, Eh.Z. TARUMOV I.V. Kurchatov Institute of Atomic Energy, Moscow, USSR

Abstract

TRANSPORT AND FOCUSING OF HIGH-CURRENT RELATIVISTIC ELECTRON BEAMS ONTO A TARGET. Investigations of the focusing of high-current relativistic electron beams for achieving a pulsed thermonuclear reaction have been made on the URAL and MS accelerators. Injection and transport of a disk-type electron beam into a cusp magnetic field allows k 25% of the beam to be deposited onto the surface of a centrally located target. With co-axial lnes electron beams have been focused onto targets with 85% efficiency. The mechanism for sharp and effective focusing by pin-type cathodes is discussed

Investigations in recent years of the focusing of high-current relativistic electron beams in diodes - in connection with a programme for using relativistic electron beams to achieve a pulsed thermonuclear reaction - have shown that beams with energies of up to 10 kJ are effectively focused on a diode anode to a spot measunng a few millimetres [1, 2, 3]. Although there is still no clear interpretation of the phenomena occurring in high-current diodes, the presence of well-focused beams offers the possibility - thanks to high current and power densities - of experimentallly accelerating thin foils to velocities of about 106 cm/s. An alternative solution of the problem is the uniform bombardment of a target by relativistic electrons from a "cloud" accumulated around the target. If the ratio of cloud radius to the target radius is 10-20, most of the electrons will hit the target. A magnetic field of cusp geometry can be used to form the cloud. At the same time, a field of this configuration can be used for transporting many beams injected into the'equatorial slit [5]. In a pulsed thermonuclear reactor based on relativistic electron beams, with energy of 109 J or more released in a pulse, the accelerators will have to be at a sufficiently safe distance from the target (3-5 m). To reduce the transport length, one can try to replace part of it by a vacuum line with magnetic insulation [6]. At the I.V. Kurchatov Institute of Atomic Energy, work is at present being done on the creation of a 5-10 MJ multi-module system with beam injection along the equator to a thermonuclear target located at the centre of a cusp magnetic field. We describe below the experimental work carried out with "Ural", "MS" and other high- current electron accelerators directed towards determining the feasibility of creating a pulsed thermonuclear reactor on the basis of relativistic electron beams.

1. INJECTION AND TRANSPORT OF A DISK-TYPE RELATIVISTIC BEAM OF ELECTRONS INTO A MAGNETIC TRAP OF CUSP GEOMETRY

Experiments on the radial transport and the accumulation of an electron cloud are being carried out in the "Ural" high-current accelerator, which has been adapted for the generation of a disk-type beam with an energy of up to 500 J (electron energy 200-300 keV, beam current up to 50 kA, pulse duration 60-70 ns) and its injection into a magnetic trap of cusp geometry through its equatorial slit [7].

25 The experimental set-up is shown in Fig. 1. Into the cathode holder of the high-voltage diode of the accelerator we inserted a "cylinder" (1) made of "orgsteklo" (organic glass - plexiglass), which was 80 mm m diameter and had a circular slit (3) in the equatorial plane of the magnetic trap 10 mm wide and sealed by an aluminium foil, 12-15 pm thick, connected to the body of the device by strips of copper foil, which served as the anode of the accelerator mterspace. The cathode (2) was a ring of stainless steel with an aperture diameter varying from 86 mm to 92 mm and a thickness varying from 0.5 mm to several millimetres. A cathode with a sharpened edge was used in a number of experiments. The disk-type beam was injected through the anode foil, in which it burned a hole 1.5-4 mm wide, depending on the experimental conditions - especially the strength of the magnetic field in the beam injection region. Fig. 2 shows the "cylinder" after an experiment. A pulsed magnetic field (period of current 10 ps) of cusp geometry (3) was created by means of two copper windings (4) carrying opposite currents. The distance between the windings was 20 mm or 40 mm. For a distance between the windings of 40 mm, the maximum strength of the field in the slit and in the trap mirrors at the moment of beam injection was 13 kOe and 6 kOe respectively. The dimensions and geometry of the windings and the magnetic field are shown m Fig. 3. Experiments in which energy was transported by means of an electron beam to a target at the centre of the magnetic trap were earned out by us with a residual air pressure in the "cylinder" (1) ranging from 4 X 10- 4 mmHg to several mmHg. The energy of the beam in the diode was determined from current and voltage oscillograms, with a correction for the inductive component of the voltage. Fig. 4 shows typical oscillograms of the current and voltage in the diode and also the X-ray detector signal. For a typical regime, the beam energy determined in this manner was 470-480 J. In addition, the energy of the electron beam beyond the anode foil was determined with the help of thin-wall graphite calorimeters of different shapes. Fig. 5 shows the calonmeter, which consists of two graphite disks 50-60 mm in diameter located in the plane of the two trap mirrors. The calonmeters were protected by radiators of thin aluminium foil (10 pum) in order to prevent heat emission in the event of local overheating. Calorimetric measurements showed that beam injection proceeded symmetrically with a scatter of less than 10% and, when there was no target at the centre of the trap, the total energy released at the calorimeters was 415-420 J, as against 470-480 J obtained from the current and voltage oscillograms. The mean electron energy in these experiments was 250 keV. These results are in good agreement with the estimate of the energy lost by the electron beam on passing through the anode foil. These measurements were performed for a magnetic field with the field windings separated by L = 40 mm and with a residual air pressure of 4 X 10-4 mmHg m the "cylinder" Even when the pressure in the "cylinder" rises to 1 mmHg, the total energy as given by the calorimeter readings does not fall by more than 10-15%. The energy penetrating the central region of the trap was determined from the sublimation and heating of the material of various test objects exposed to the relativistic electron flux. The heating of the objects was determined by means of thermocouples (5). It should be borne in mind, however, that the energy value obtained in this way may be too low, for the material evaporating from the surface continues to be heated. Energy measurements based on the heating and sublimation of graphite and aluminium targets (spheres with a diameter of 6-8 mm, cylinders 8 mm in diameter and 6 mm long) have shown that, with a low residual air pressure (< 10- 3 mmHg), not less than 25% of the energy of the electron beam passing through the anode foil is released at the surface of the targets placed at the centre of the magnetic trap for L = 40 mm. When the pressure rises to several mmHg, the energy fraction released in a target of the dimensions indicated does not change very much, although the nature of the radiation becomes different. In these experiments, the nature of the propagation of the beam in the trap and the degree of uniformity of the irradiation of the targets were determined. To determine the nature of the propagation of a disk-type electron beam in the trap, the track of the beam was photographed

26 in X-radiation by means of a pinhole camera (6). Figure 6 shows pictures from various trap regions taken with the help of a tantalum convertor (cylindrical or conical) at a pressure of about 1 mmHg in the "cylinder". It can be seen from the pictures that the beam keeps its disk shape, passing 20 mm along the radius into the heart of the trap. In the last frame of Fig. 6, one can see clearly the non-uniform luminescence of a lead sphere 8 mm in diameter suspended from a fine thread at the centre of the trap (pressure P = 4 mmHg). At low pressure (P < 10- 3 mmHg), the nature of the propagation of the beam is different. The X-ray luminescence of the surfaces of the test objects placed in the trap is much more uniform than when the pressure is high, indicating "dispersal" of the beam. The symmetry of irradiation and the high efficiency of energy transport to the target in the case of dispersal of the beam entering the trap can be attributed to the fact that the magnetic field used is a good trap for relativistic electrons, the equatorial slit being closed by the electric and magnetic fields in the diode and the area of losses through the mirrors being small. In such a trap, the relativistic electrons entenng it must collect as a "cloud" in the region of zero field. If a target is placed in this cloud, its surface will be heated uniformly by the relativistic electrons penetrating it

2. ENERGY TRANSPORT WITH THE HELP OF CO-AXIAL LINES WITH MAGNETIC INSULATION

Co-axial lines with magnetic self-insulation as a means of directing relativistic electron beams to a target have been investigated in the "MS" device [6]. Whereas in earlier experiments use was made of "short" lines with length £ < rC, where r is the pulse duration characteristic, in this study we investigated a 3.5 m line for which the wave propagation time was more than 30% of the half-width pulse duration. The experimental set-up is shown schematically in Fig. 7 In the experiment, we measured simultaneously the voltage and current in the line, the current to the Faraday cylinder and the X-radiation from the lateral surface of the co-axial and beyond the end-cap. It should be noted that all the measurements were performed for a co-axial line with a diameter ratio of 5.2/2.6 and for a diode gap of 0.3 cm. The line current was 30-40 kA at a voltage of 350 kV. In Fig. 8 we present oscillograms of the voltage, the currents and the X-ray measurements. It can be seen that the current to the Faraday cylinder is delayed relative to the line current, which corresponds to the time necessary for wave propagation along the co-axial. By appropriate treatment of the oscillograms one finds that up to 85% of the energy entering the line reaches the Faraday cylinder. In the simplest case, the propagation of electromagnetic signals in the co-axial can be described by the so-called "telegraph equations" [9].

au 1 aA aJ C at

aZ at .....GU (2) which are derived for a small line section in the quasi-stationary approximation. If one assumes that the wave front width A < £, using equations (1) and (2) and assuming the establishment of magnetic self-insulation behind the wave front, it is possible to obtain the wave velocity V and estimate the maximum leakage at the wave front. The wave velocity obtained, V = C ,- is in agreement with the results obtained from the oscillogram i n, andi te ai Y+1oe in Fig. 8, and the maximum losses do not exceed an estimated 15%.

27 3. FOCUSING OF AN ELECTRON BEAM IN LINES WITH MAGNETIC INSULATION

As the voltage of the electric field in the gap increased to 2-3 MV/cm, we noted the formation of a focused beam propagating in the end gap of the co-axial over a distance of 0.3-2.0 cm. These measurements were performed in the "MS" device in a line 10 cm long operating under quasi-stationary conditions. Co-axial conductors with an external radius of 0 2-2.5 cm and an internal radius of 0.035-0.2 cm were installed at the exit of the accelator tube. The length of the co-axial conductors varied from 1 cm to 10 cm, measurements showing that the impedance of a co-axial is virtually independent of its length. From an examination of the osclllograms it follows that, for b/a = 0.4/0.2, as the gap increased from 0.3 cm to 2 cm there resulted a 34 ns delay of the current to the Faraday cylinder relative to the start of the shunt current. With an accuracy of 10% one can say that the time delay Ar increases linearly with the increase of the gap d. The beam transmission efficiency determined by the ratio of the charge reaching the Faraday cylinder and passing through the shunt is shown in Fig 10 as a function of the gap spacing. As the Faraday cylinder is moved further away, the charge reaching it decreases. At the same time, the current to the sides of the co-axial conductor near its end increases - this is also confirmed by X-ray measurements from the side surface. The decrease of the impedance of the co-ax and its time shift may be due to the formation of plasma in the diode gap. In fact, a dense plasma forms at the inner electrode of the co-ax as a result of a "micropoint" explosion. When one has large diode gaps, the current initially impinges on the side surface of the outer co-ax, as a result of which plasma is accelerated in the longitudinal direction ana "shot" into the diode interspace [10], the main part of the beam impinging on the Faraday cylinder. The time delay of the current at the Faraday cylinder gives a plasma velocity value of up to 5 X 107 cm/s. The presence of such a plasma is also confirmed in a number of cases by shorting of the diode interspace in a time less than the pulse duration, which is recorded on the basis of the characteristic "pulling" of the current to the Faraday cylinder. Figure 11 shows the degree of focusing of the beam, determined on the basis of X-radiation with a pinhole camera. Here the relative darkening in the proportionality region of the photo- graphic material is converted to current density. The size of the focusing region at half-width is 1 mm and the maximum current density 0.5 MA/cm 2. Measurement of the X-radiation bounded by a lead cone 1 cm thick showed that up to 30% of the current passes in a region 1 mm in diameter. It should be noted that focusing in a co-axial geometry makes the connection between generator and load more flexible.

4. FOCUSING OF A HIGH-CURRENT ELECTRON BEAM IN A DIODE WITH THE HELP OF A PIN-TYPE CATHODE

We have previously reported the results of sharp and effective focusing of a high-current beam in a diode with the help of a pin-type cathode in the "Ural" accelerator [2]. As we pointed out, the attempt to describe the phenomenon of focusing with the help of a pin-type cathode on the basis of pair-potential theory proved to be a failure for two reasons: (a) the results depended only slightly on the length of the pin-type cathode - in other words, its length was virtually independent of the ratio of the diode radius to the cathode-anode distance; and (b) focusing depended strongly on the distance from the pin-type cathode to the anode.

28 In order to throw light on the mechanism of focusing with a pin-type cathode, additional experiments were carried out with the "Ural" accelerator and in another high-current accelerator with parameters differing considerably from those of "Ural": its diode voltage could be 40-60 kV and the wave resistance of the co-axial shaping water line was about 0.3 ohm. The electrical length of the line corresponded to 30 ns, and the nominal energy stored in it was 100 J. The accelerator is shown schematically in Fig. 12, while Fig. 13 shows oscillograms of current, voltage and X-radiation for two distances between the pin-type cathode and the anode (d = 1.5 mm and d = 0.5 mm). In Fig. 14 we show pictures of the focal spot taken in X-ray light with a pinhole camera for d = 1.5 mm (optimum distance between the pin-type cathode and the anode). The mean current density at a focal spot 0.5 mm in diameter was about 25 MA/cm 2 In some of the experiments we observed beam focusing with a spot having a complex structure (Fig. 14 b). X-ray hardness measurements performed in "Ural" behind a lead collimator with an aperture diameter of 1 mm and an axis passing through the focal spot showed that the hardness corresponded to the voltage at the diode beyond the region of sharp current rise (the diode current rises to 100 kA for a diode voltage of 90-100 kV). It should be noted that, despite the considerable difference between the parameters of the accelerators in which beam focusing with the help of a pin-type cathode was investigated, the main characteristics of the current, voltage and X-ray oscillograms are similar. In our opinion, the current, voltage and X-ray oscillograms and the (time and space) focusing characteristics can be explained on the basis of the formation of a double layer between the cathode and the anode plasma in the interspace between the pin-type cathode and the anode. The existence of a layer for a long time (focusing duration 30-40 ns) is possible due to the fact that, for a certain gap between the pin-type cathode and the anode, the maximum current in the /Te layer Ilayer ~ e n , / S-(here n is the plasma density along the layer boundary, Te and m are Vm the temperature and mass of an electron and S is the area of the emission surface) is less than the short-circuit current of the accelerator Is-c -le (where JjUme is the voltage in the shaping hne p and p the wave resistance of the line), which ensures normal functioning of the diode in this regime. We would point out that the functioning of the diode can be broken down into two stages: regime of vacuum take-off of current (region on the oscillograms up to the sharp current rise) when there is a vacuum interspace between the cathode and the anode plasma, and the "plasma diode" regime upon the formation of a layer between them. Estimates show that for "Ural", with an optimum gap of 3.5 mm between the pin-type cathode and the anode, the thickness of the layer can be 0.5-1 mm. The maximum current in the layer in this case - given reasonable assumptions regarding plasma density (n - 1016 particles/cm 3 on the basis of data presented in 2 Ref. [11]), electron temperature (Te ~ 4-5 eV) and emission area (S ~ 0.5 cm : size of molten spot on copper anode) - is about 100 kA, while the short-circuit current for Uline ~ 500 kV is about 200 kA. In fact, with decreasing gap between the pin-type cathode and the anode there is an early shorting of the interspace, apparently due to the considerably greater density of the plasma in the layer in this case. The beam focusing mechanism in the proposed model is determined by the action of the proper magnetic field of the current both under "vacuum diode" conditions and with the formation of a double layer, when the current reaches a value exceeding the critical value

Icrit. = 8.5 -'3y-R(kA), where R is the effective transverse dimension of the emission surface and d d is the distance between the virtual cathode and anode or the layer thickness. R Estimates show that for "Ural" Icrit. 50-60 kA on the assumption that - - 5-7 (d ~ 1 mm d and R ~ 5-7 mm) and hence focusing starts only at the moment of the voltage drop at the diode.

29 An important factor for sharp focusing is apparently the early formation of the anode plasma, so that the meeting place of the cathode and anode plasmas and the formation of the layer are at some distance from the anode surface, which ensures the formation of an electron beam "crossover" at the anode surface in accordance with the picture considered in Ref. [12]. The authors are very grateful to L.I. Rudakov for his valuable advice on the setting-up of the experiments and for his constant interest in the work covered by this paper.

REFERENCES

[1] YONAS, G., POUKEY, J.W.,PRESTWICH, K.R., FREEMAN, J.R., TOEPFER, A.J., CLAUSER, M.., Nucl. Fusion, 14, 5, 731 (1974). [2] KOBA, Yu.V., et al., paper presented at the Fifth Internat. Conf. on Plasma Physics and Controlled Nuclear Fusion Research, Tokyo, Nov. 1974, Vol. II of proceedings, IAEA, Vienna, p. 337. [3] BLAUGRUND, A.E., COOPERSTEIN, G., Phys. Rev. Lett. 34 461 (1975). [4] Number not used. [5] BARANCHIKOV, E.I., et al., paper presented at 1st Conf. on Relativistic Electron Beam Applications, Albuquerque, USA, Nov. 1975, Vol. I of proceedings, p. 284. [6] GORDEEV, A.V., et al., Transactions of the New York Academy of Sciences 251, pp. 668-678. [7] KOBA, Yu.V., et al., Pis'ma Zh. Tekh. Fiz. 2, issue 7, 12 April 1976. [8] Number not used. [9] TAMM,I.E., Osnovyteoni e]ektrichestva (Fundamentals of the theory of electricity), GITTL (State Publishing House for Techmco-theoretical Literature), Moscow (1954). [10] Voprosy teorii plazmy (Questions of plasma theory), issue 8 (Leontovich, M.A., Ed.), 1975. [11] KELLY, J.G., MIX, L.P., Journ. Appl. Phys. 46,1084 (1975). [12] GOLDSTEIN, S.A., DAVIDSON, R.C., SIAMBIS, J.G., LEE, R., Phys. Rev. Lett. 30, 25 (1975).

30 FIG.1. Experimental set-up

FIG 2. "Cylinder" of plexiglass or lucite with circular slit for the radialinjection of a disk-type electron beam The hole burned by the beam is clearly visible on the anode foil Width of hole 1 5-4 mm.

31 FIG.3 Geometry of magnetic field.

50KA Li______I - _ a JL 200QKEB

Log 6

L... . ._ . .. ri ------^ . -. -.- - . - ^ . -.

50HC FIG.4. Typical oscillograms of (a) current, (b) voltage and (c) X-ray detector signal.

32 FIG.5. Disk-type graphite calorimeter I - graphite disks, 2 - radiators(aluminium foil - 10 mm); 3 - thermocouples.

33 FIG.6. Photographof beam track taken in X-ray light.

6 7

2 3 5

\ -»vwr» ^ j 1N \\1 x . . _ - ,, i i AT I WrP 1 JAM

/V LA L

FIG.7. Experimental set-up. 1 - cathode foot, 2 - inner tube of vacuum line, 3 - outer tube of vacuum line, 4 - cathode; 5 - anode; 6 - capacitive divider; 7 - shunt.

34 370KB

T~Hs~I

20KA!I X

FIG.8. Oscillograms of voltage (U), current at shunt (Ish), currentat Faraday cylinder (IF and X-radiation (X~s).

35 Hee,HCe-

30-

0.-

0 0.5 <0o 5 to 2.5 .o CM

FIG.9. Delay of current to Faraday cylinder (IFc) as a function of gap

quQw

1 L=0,3c L= 3CM L= 4CM L= 5ce d,CM 0' 5 <5 20

FIG.10. Ratio of charge reaching Faradaycylinder (IFc) to the total charge as a function of the cathode-anode gap for co-axials of different length (L).

36 (r)OTH.ea

0,5

FIG 11 Currentdensity in plane of anode; obtained with the help of a pin-hole camera.

FIG.12. Schematic view of the accelerator:1 - accumulating capacitor;2 - controllable gas switch; 3 - shaping line, 4 - non-controllablegas suited, 5 - cathode; 6 - high-voltage diode; 7 - pin-type cathode, 8 - anode foil; 9 - anode insert; 10 - anode, 11 - resistive voltage divider, 12 - current shunt, 13, 14 - X-radiation detectors; 15 - Polyethylene diode insulator.

37 IOOKA a

50KB

50Hc cd=t,5mm. dc=O,,5mm.

FIG.13. Typical oscillograms of (a) current, (b) voltage and (c) X-radiation detector signal. I - for gap d = 1 5 mm (X-ray signal weaker by a factor of 34); II - for gap d = 0.5 mm.

38 K TOO cp 0, 5

- -,/,. ,/,/

ATblPb L J_ I i AHOa

odHo ae/eneue -2,5Mm.

FIG.14. Photographs of focal spot taken in X-ray light with a pinhole camera having a 0.2 mm aperture. Enlarge- ment 1 1. a - normal focusing; tungsten pin-type cathode; copper anode b - focusing with structure (three focal spots visible), tungsten pin-type cathode; aluminium anode; 30 mm thick

39 NEW HIGH-CURRENT RELATIVISTIC ELECTRON BEAM ACCELERATORS AT THE I.V. KURCHATOV INSTITUTE OF ATOMIC ENERGY

M V. BABYKIN, B.V BAEV, K.A. BAJGARIN, A.V. BARTOV, P.P GAVRIN, B.A. DEMIDOV, E.D. KOROP, V I MIZHIRITSKIJ, A.M. PASECHNIKOV, S S SOBOLEV, S.D FANCHENKO I.V. Kurchatov Institute of Atomic Energy, Moscow, USSR

Abstract

NEW HIGH-CURRENT RELATIVISTIC ELECTRON BEAM ACCELERATORS AT THE I.V. KURCHATOV INSTITUTE OF ATOMIC ENERGY. The technical characteristics and the results of experiments on the formation of electron pulses and self- fbcusing of the relativistic electron beams with the Angara-l and Kal'mar-l accelerators are described. A maximum voltage of 0 55 MV, short circuit currents of 0.4 MA and currents of focused electron beams of 0.12 MA are achieved.

1. INTRODUCTION

Increasing attention is being paid to the idea, first put forward by E.K. Zavojskij (USSR) [ 1] and F Winterberg (USA) [2], of a pulsed thermonuclear reactor based on the use of powerful relativistic electron beams. The suggestion made in 1972 by L.I Rudakov [3] of using a sheath for electron deceleration and gas-dynamic plasma compression and for slowing down particles as they fly apart, opened up the possibility of solving the problem with contemporary techniques for producing beams. Work on the physical and technical problems of a thermonuclear reactor with electron beam heating requires high-current pulsed electron accelerators giving a high current and energy density at the target At the I.V. Kurchatov Institute of Atomic Energy, where some devices had already been operating for several years, Angara-l started operation in 1975 and Kal'mar-l in 1976. In work with these new devices, particular attention has been paid to the problem of focusing the relativistic beam on a given point. The present paper reviews the technical characteristics of these devices, describes their various parts and presents the results of experiments on the formation of powerful electrical pulses and the self-focusing of beams of relativistic electrons.

2. ANGARA-1 AND BEAM PRODUCTION AND FOCUSING EXPERIMENTS

Angara-l Is intended for research on the processes involved in beam focusing, for generating the necessary specific power when a target is irradiated, and for investigation of the possibility of producing a pulsed thermonuclear reaction. The device is a three-step water line with a characteristic impedance of 1.5 ohm and an operating voltage of 1-1.5 MV. A diagram of the device is shown in Fig. 1. It is described in greater detail in Ref. [4]. The special features of the device are the use of gas-filled spark-gap switches to switch the low-impedance water line at megavolt potentials (especially the use of a multiple discharge in a spark-gap switch with annular electrodes for peaking the pulse front) and the use of a transformer line at the output, making it possible to vary the output impedance over a wide range (transformers with output impedances of 6, 2.2 and 0.83 ohm have been tested). The output power of the device may reach 10l -10 '2 W. Figure 2 shows, by way of illustration, an oscillogram of the voltage up to the peaking spark-gap switch and beyond. The pulse front up to the peaking spark-gap switch is 50-70 ns; beyond the peaking spark-gap switch it is shortened to 30 ns.

41 In adjusting the device, a flat graphite cathode 80 mm in diameter and 1 cm from the flat stainless steel anode was used. With a voltage of 1 MV and a current of up to 150 kA (output resistance of the line 6 ohm), self-focusing of a beam to a diameter of 5 mm in a diode was observed. Figure 3 shows a photograph of such a beam, taken in X-ray light with a pinhole camera. One shortcoming of this self-focusing regime is the spread (reaching several millimetres) of the focal spot relative to the centre of the cathode. To fix the place of beam transmission, focusing methods involving pins and projections of various shapes at the centre of the flat cathode were tried (Fig. 4). With a pin 2 mm in diameter, fairly good beam focusing (down to a diameter of 1 mm) was achieved, however, the low-resistance output of the device (0.83 ohm) was not matched sufficiently well with the diode impedance, which amounted to several ohm and reached 1 ohm only just before the diode was short-circuited by the plasma. To reduce the diode impedance, the pin was encircled by an annular projection (Fig. 4b) 8-12 mm in diameter. With this arrangement, the plasma in the diode first consists of individual filaments starting at the annular projection, then (during a period of 15-20 ns), the filaments collect around the pin (Fig. 5). After that one of the filaments becomes the main channel of the beam, so that with this focusing method the position of the channel failed to coincide with the centre of the diode by a distance of the order of the pin radius. Apparently, a similar current structure before focusing can occur in other devices. When the configuration of the projections was vaned it was found that, in the case of a flat-bottomed projection (Fig. 4c), the current channel sometimes stopped at the point where the flat bottom met the internal cone of the projection. As a result of this observation it was decided to do without a central pin and to use a cylindrical projection with a conical depression (Fig. 4d) This configuration has proved to be the most successful one, for it ensures the most stable current channel position - in the depression. The radius of curvature of the projection edge must be of the order of 0.1-0.2 mm. If the edge is very sharp, rapid short-circuiting of the diode by the plasma is observed; if it is too blunt, the plasma forms slowly and the current in the diode rises slowly. Under optimum conditions, the delay of the current relative to the voltage front is 20-30 ns (Fig 6). This time appears to be necessary for the plasma channel to form in the diode. Then, for 15-10 ns, the current rises to a con constant value of the order of 250 kA, which is observed for 30-40 ns. The existence of a current plateau indicates that at this time the inductive drop in voltage at the diode is zero and that the diode has a pure resistance. It is an interesting fact that current fluctuations with a frequency of up to 500 MHz are observed during this time (Fig. 6). The frequency of these fluctuations appears to depend on the thickness of the anode. with a thick anode (2 mm of stainless steel) it is 500 MHz, with an anode of copper or stainless steel foil 0.1 mm thick it is 200-300 MHz, for a thickness of 50 pm it is 120-150 MHz After the current limitation phase, the diode is short-circuited and the current rises to 600 kA, which corresponds to a short-circuit of the line at 0.5 MV. The current limitation and the appearance of oscillations with a frequency of up to 500 MHz probably indicates the development of a current instability in the plasma channel. The first to draw attention to this was M.V. Babykin. The current limitation in the diode channel may be due to the development of an ion-acoustic instability. According to the formula derived by L.I. Rudakov, the current in the channel should in this case be

M - I= 8500 - V/(MB) (A) vm which agrees closely with the values observed in the plasma focusing of a current in a diode. These calculations were based on the assumption that, while the ion-acoustic instability is developing, the plasma electrons are heated to an energy close to the applied voltage and carried to the anode as a result of current drift, producing bremsstrahlung. Figure 7 shows the resistance and power developed in the diode; they are calculated from the voltage and the current of the

42 diode. The diode resistance falls from 10-20 ohm at the start of the process to 1 ohm before the diode is short-circuited by the plasma The curve for the power developed at the anode is similar in shape to an X-ray oscillogram and reaches a maximum during the time when the curve for the current forms a plateau. The maximum power in the case under discussion is 4 X 1010 W and the total energy release before the diode is short-circuited is 2 kJ With better matching, the energy release reached 4-5 kJ. For estimating the diameter of the current at the anode, photographs (Fig. 8) of the anode were taken in X-ray light with a pinhole camera The diameter of the focused current at the anode was less than or of the order of 1 mm The current density was 30 MA/cm 2 and the power density at the anode 1013 W/cm 2. Figure 8b shows a photograph of the target, made of gold foil 15 pm thick and shaped like a spherical segment enclosing an angle of 120 ° This photograph shows that the beam has a diameter of ~ 1 mm and heats the top of the sphencal segment

3 KAL'MAR-1

3.1. Description of the device

Kal'mar-l, designed for the production of beams of electrons with an energy of 1 MeV and a current of 0.5 MA, is a double pulse-forming line (DFL) with an electrical length of 70 ns charged by a pulsed voltage generator with a capacity of 70 kJ. The generator (with IMP-100-0.1 capacitors) has 20 multiplier stages. The spark-gap switch of the first stage is of the trigatron type, the rest are uncontrolled. Thanks to the introduction of an inductance element for shunting the output of the second stage to earth, the scatter of the generator start-up delay (with an output voltage of more than 80% of the breakdown voltage) does not exceed Ar3 = 10-20 ns for r3 = 250 ns [5]. The DFL, the high-voltage diode and the high-voltage coaxial transformer are in a cylindrical stainless steel housing 4 m long and 1 m in diameter (Fig 9) which is sufficiently strong for vacuum pumping during filling with distilled water. The central DFL cylinder (3 in Fig. 9) is supported by stand-off insulators with resilient elements for damping the hydrodynamic shock which permit 3-5 mm axial displacement of the cylinder with subsequent return to the initial position.: In the experiments described below, the function of DFL spark-gap switch was performed by a water-filled.gap between the cylinder (3) and the regulatable shaft (1). Thanks to the precautions taken against hydrodynamic shock, the device can withstand an energy release of 20 kJ in the water-filled spark gap without sustaining mechanical damage Measurements show that the voltage drop in this spark gap reaches 20-25%. The housing has windows (2) for the initiation of a breakdown in the water by laser radiation, the possibility of doing this was investigated in a mock-up. Figure 10 shows the spark formed in the water by focused laser radiation (neodymium glass laser, 500 MW, 20 ns). The measurements in the mock-up (see Ref. [6]) showed that the impedance of the channel of a breakdown initiated in the water is several times smaller than in the case of auto-breakdown, and the scatter of the breakdown initiation delay is only a few nanoseconds under optimum conditions. To avoid breakdown of the pulsed voltage generator and the DFL in case of failure or of premature or delayed triggering of the DFL spark-gap switch, a non-linear resistor (several standard vilite resistors in the form of cylinders 100 mm in diameter and 60 mm high) is introduced into the charging circuit. For DFL charging to 1 MV, four vilite resistors in series were used; according to the data of Fig. 11 [7], this should give a logarithmic decrement of the oscillations of the generator-DFL system 6 = 1.3 with a loss of not more than 5% of the initial voltage.

43 An easily replaceable thin disc made of stainless steel, aluminium, lead or some other metal serves as the anode of the high-voltage diode; the cathode is also replaceable. Cathodes of various configurations were tned (plane, cone, ring, needle). The diode current is measured with a non- inductive resistor (1.8 X 103 ohm) inserted between the accelerator housing and the anode. The beam current is measured with a Faraday cylinder. The voltage at various points in the device is determined with capacitance-type sensors (14 in Fig. 9). The X-rays are registered by a plastic scintillator with a coaxial photoelectric cell (FEhK-16), an X-ray pinhole camera and thermoluminescent and chemical radiation detectors.

3.2. Experimental results

Figure 12 shows the behaviour of current and voltage in the high-voltage diode. The maximum voltage attained in the diode was 0.55 MV, the short-circuit current 0.4 MA and the current of the focused relativistic electron beam 0.12 MA. Figure 13 presents experimental data on the current in a flat-cathode diode as a function of the gap 6 between cathode and anode. It can be seen that, for low values of 6, the diode broke down and the modified current I was no different from the short-circuit current Isc At a certain gap value 6*, the current I was about half of Isc, as it must be in a regime where the impedances of the line and diode are matched. This conclusion is supported by the data from measurements of the voltage in the DFL (Fig. 14), according to which the amplitude of the reflected voltage wave is at a minimum when 6 = 6*. Figure 15a shows an X-ray image - obtained with a pinhole camera - of the focusing of a beam in a diode with a flat cathode It will be seen that the beam was focused to a diameter of 2 mm, while material was eroded from the corresponding place on the anode (Fig. 15b). Unfortunately as Fig 15b shows, the position of the beam focus on the anode changed sporadically from pulse to pulse of the accelerator, m addition, focusing was not observed for every pulse. The cathode configurations mentioned above were tested with a view to investigating the possibility of stabilizing the position of the relativistic electron beam focus. The best results were obtained with a conical metal cathode 7 cm in diameter, the anode having the same diameter The X-ray pinhole camera and the erosion of material from the anode showed that the conical cathode focuses the beam at the centre of the anode with 100% reproducibility. Figure 16a is a pinhole camera image where the electron beam focus diameter is not more than 1 mm, and the photograph in Fig. 16b shows the anode disc with material eroded exactly at the centre, indicating that the energy in the focus of the beam was not less than 3-4 kJ (corresponding to a current density of ~ 107 A/cm2 and a power density in the focus of ~ 5 X 1012 W/cm 2). The experiments showed that a conical metal cathode can withstand tens of accelerator start-ups at a beam current amplitude of 50-100 kA without being destroyed. The experimental curves for X-ray absorption in copper filters of variable thickness placed in front of the apertures of a nine-channel pinhole camera and m front of LiF thermoluminescent detectors agree reasonably well with the bremsstrahlung spectrum of electrons with an energy corresponding to the voltage in the diode.

4. DISCUSSION

With Angara-1 and Kal'mar-1 it is possible to conduct several important experiments showing that a thermonuclear reaction can be initiated with electron beams. In experiments with cathodes of various configurations, beams that focused on areas less than 1 mm in diameter were obtained. In contrast with previously known methods, focusing was achieved by simple means using projections of various configurations on a flat cathode (Angara-l) or conical cathodes (Kal'mar-l). Diode impedance may be selected by varying the angle of the cone or the configuration and size of the projections on the flat cathode. With the focusing methods developed it is possible to fix

44 the position of the focused beam accurately and to obtain a high density of current and of energy release. In Angara-l, current densities of up to 30 MA/cm 2 and power densities of up to 1013 W/cm 2 have been obtained; in Kal'mar-l, the corresponding values are 107 A/cm 2 and 5 X 1012 W/cm 2 . Such current and energy flux densities are sufficient for experiments on the irradiation of thermonuclear targets, for the study of plasma turbulences during the flow of powerful currents in plasma diodes, for investigation of the mechanisms involved in the interaction of a focused beam with thin metal foils, for the examination of plasma confinement and heating in thermonuclear targets under various conditions and for research mto the possibility of initiating a pulsed thermonuclear reaction. In conclusion, it is our pleasant duty to thank E.P. Velikhov and L.I. Rudakov for their constant interest and support.

REFERENCES

[1] BABYKIN, M.V., ZAVOJSKIJ, E.K., IVANOV, A.A., RUDAKOV, L.I., in Plasma Physics and Controlled Nuclear Fusion Research (Proc. Conf. Madison, 1971), IAEA, Vienna, Vol. 1, p. 635. [2] WINTERBERG, F., Phys. Rev. 174,1 (1968) 212. [3] RUDAKOV, L.I., SAMARSKIJ, A.A., Proc. 4th Europ. Conf. Controlled Fusion and Plasma Physics (Moscow, July-August 1973). [4] RUDAKOV, L.I., BABYKIN, M.V., Proc. 7th Europ. Conf. Controlled Fusion and Plasma Physics (Lausanne, 1-5 September 1975), Vol. II, p. 172. [5] DEMIDOV, B.A., IVKIN, M.V., Prib. Tekh. Ehksp. (1975), issue 3, 120. [6] DEMIDOV, B.A., IVKIN, M.V., PETROV, V.A., FANCHENKO, S.D., Prib. Tekh. Ehksp. (1974), issue 1, 120. [7] DEMIDOV, B.A., IVKIN, M.V., PETROV, V.A., FANCHENKO, S.D., Prib. Tekh. Ehksp. (1975), issue 3,37.

FIG.1. Diagram of Angara-1.

45 FIG 2 Oscillograms of the voltage in a water line up to a multichannel peaking spark-gap switch (front of the order of 60-70 ns) and after peaking (pulse front 30 ns)

FIG 3 Photograph of the self-focusing of a beam in a flat diode, taken in X-ray light using a pinhole camera with an aperture of 1 mm (beam diameter 4-5 mm).

46 i *10

i a) 6)

6) 2) J -L ' I a)

FIG 4. Projections used on a flat cathode in selecting the focusing regime. (a) pin 2 mm in diameter, 6 mm long, distance from anode 4 mm, (b) pin encircled by projection with conical end surface, (c) pin encircled by annular projection with flat bottom, (d) annularprojection with conical depression ensuring stable current channel position, (e) position of the flat cathode and the projection relative to the anode

FIG.5. Photographsof the tracks of individual currentfilaments on the anode when the potential difference has been removed from the diode 10-15 ns (left) and 20-25 ns (right) after the beginning of the process as a result of diode insulator breakdown The photograph on the right shows the current channels grouped around the central pin.

47 FIG.6. Oscillograms of the voltage in the diode (top), the diode currentand the X-ray pulse. The diode current forms a plateau after which it rises sharply as a result of short-circuitingby the plasma. The X-ray trace declines sharply the moment the diode is short-circuited Below is the calibrationsignal, with a period of 20 ns.

48 0 100 rO0W

4o O0o

FIG. 7. Results obtained by calculating the resistance of the diode and the power developed in it (lower curves) from the voltage (upper oscillogram) and current (other oscillogram) of the diode. The diode resistance changes from 10-15 ohm at the beginning of the process to 0. 7--1 ohm before the diode is short-circuited. The maximum power developed is 4 X 1010 W. The time dependence of power is similar to the oscillogram of the X-ray pulse (thirdfrom top)

49 FIG.8 Photographsof a sphericalsegment irradiatedby the beam and consisting of gold foil 15 pm thick; the photographs were taken in X-ray light with a pinhole camera. The pictures illustrate the various focusing regimes - sharp focusing, in which a spot about 1 mm in diameter at the top of the segment is irradiated,and relatively uniform irradiationof the entire surface of the sphericalsegment.

50 FIG 9 Diagram of Kal'mar-1 (1) Moveable shaft of uncontrolled DFL spark gap switch, (2) Window for laser heating, (3) CentralDFL cylinder; (4) Central cylinder holder, (5) Housing; (6) Coaxial transformer, (7) Water; (8) Cylindrical measurement resistors, (9) Plastic insulator; (10) Anode of diode, (11) Cathode of diode, (12) Partitioninsulator, (13) High-voltage input, (14) Capacitance-typesensor

FIG.10. Laser spark in water (focal length of lens 6 5 cm)

51 1,0

?qQ8,6 II 7

0,2

0 1 2 3 4 5 4ucno conpomuAneHu A, n

FIG 11. The logarithmic decrement (8) in the generator-DFL circuit and voltage losses as functions of the number of series-connected vilire resistorsfor a DFL charge of 0.5 MV

52 a

FIG 12. Current and voltage in a diode with a conical cathode (1) Regime with self-focusing of beam. (a), (b) and (c) are oscillograms of the voltage, currentand X-rays, in (d) T = 50 ns; (2) Short-circuitregime (b) and(a) are oscillograms of current and voltage.

53 7 KA

o o

10 0 _ 8p 8 o _ 0

_ o o °oe

o o

,I I~i X MM u , 5 10 6

FIG 13. The diode currentas a function of the gap 6 between cathode and anode for a voltage at the peaking system input of 500 kV, flat cathode.

54 FIG 14 Voltage in the DFL. (a) when diode impedance and the output resistance of the coaxial transformer correspond, i.e. 8 = *; (b) when 8 > 8 *, (c) short-circuit, (d) calibrationf= 1 MHz.

55 FIG.15. Self-focusing of the electron beam in a flat diode. (a) Pinhole camera image of X-rays from anode, (b) Erosion of the outer surface of the anode (stainless steel 1 mm thick), total number of start-ups of the device 8.

FIG.16. Focusing of an electron beam in a diode with a flat cathode: (a) Pinhole camera image of X-rays during one beam pulse, (b) Erosion of anode material duringone beam pulse.

56 LASER INTERACTIONS COMMERCIAL APPLICATION OF LASER FUSION*

L. A. Booth University of California Los Alamos Scientific Laboratory Los Alamos, New Mexico

ABSTRACT

The fundamentals of laser-induced fusion, some laser-fusion reactor concepts, and attendant means of utilizing the thermonuclear energy for commercial electric power generation are discussed. Theoretical fusion- pellet microexplosion energy release characteristics are described and the effects of pellet design options on pellet-microexplosion characteristics are discussed. The results of analyses to assess the engineering feasibility of reactor cavities for which protection of cavity components is provided either by suitable ablative materials or by diversion of plasmas by magnetic fields are presented. Two conceptual laser-fusion electric generating stations, based on different laser- fusion reactor concepts, are described.

I. INTRODUCTION The development of laser-fusion technology is progressing in an orderly manner with the time required to develop and construct high- power-level, short-pulse lasers pacing the program. Significant numbers of thermonuclear neutrons have been produced by fusion-pellet irradia- tions with laser power levels less than 1 TW. New and unprecedented lasers with power levels as large as 200 TW are scheduled for completion during the next five years. Achievement of the important milestone of scientific breakeven (thermonuclear energy release equal to laser energy input) is expected with this new generation of lasers. Although the technical feasibility of producing commercially useful thermonuclear energy releases from laser-induced fusion has not been demonstrated, theoretical predictions of fusion-pellet-microexplosion characteristics are being used in preliminary reactor design and evaluation studies.

*Work done under the auspices of the U.S. Energy Research and Develop- ment Administration, Contract Number W7405ENG36.

59 This paper describes some laser-fusion reactor (LFR) concepts, and attendant means of utilizing the thermonuclear energy for commercial electrical power generation. The conceptual LFRs discussed in this paper include a reaction cavity in which the thermonuclear energy is released from deuterium-tritium (D+T) reactions within a pellet, located at the center of the cavity with thermonuclear burn initiated by a laser pulse.

For (D+T)-burning plants, two essential requirements for a LFR concept are similar to those for a reactor concept based on magnetic confinement: o The need to produce tritium artificially because natural supplies are insufficient to support a large-scale power-generation industry, and o The need to convert the 14-MeV neutron energy into usable form.

Both needs are satisfied by providing a "blanket" of lithium which surrounds the reaction cavity. Tritium is generated in a major fraction of reactions between neutrons and lithium; and lithium, being a light element, also converts neutron kinetic energy to thermal energy by means of elastic-scattering reactions. Furthermore, additional thermal energy is produced by exoergic neutron reactions with the lithium. It is essential that at least as much tritium be generated as is burned and lost, and that as much as possible of the neutron energy be converted into high-grade thermal energy for ultimate conversion to a form useful to the direct consumer.

A characterizing LFR feature that differs significantly from magnetically confined fusion reactor concepts is the fact that fusion- pellet microexplosions represent substantial amounts of energy released on a very short time scale. The minimum energy release, determined by both physical and economic considerations, is probably about 100 MJ. Although the hydrodynamic blast created by the pellet microexplosion can be controlled with relative ease (because the energy is carried by a small mass of high energy particles), large stresses can result from high rates of energy deposition in the blankets and structural materials. A major design problem in containing this energy is posed by the need for a low-pressure cavity in which the pellet can be heated and compressed by a laser pulse without prohibitive laser-energy loss along the beam path, while, at the same time, maintaining a finite layer of blanket material that surrounds the cavity.

60 II. CHARACTERISTICS OF LASER FUSION

Pellet Design

In contrast to magnetically confined fusion where the (D+T) fuel would normally be injected into the reactor in gaseous form, laser- fusion fuel would be injected in "solid" form, i.e., as cryogenic-solid (D+T) spheres or as (D+T) gas encapsulated under pressure in more complex structures of high-Z material shells. The understanding of the physics of laser-induced fusion is in- complete so that definitive specification of neither the laser parameters nor the target design can be made with certainty. Sophisticated calculational techniques to analyze laser-induced fusion have been developed but suffer from lack of corroborating experimental data. 1 In this regard the situation is similar to that found in the controlled thermonuclear research programs in that progress must be based primarily on experimental investigations with the theory serving principally as a guide rather than the converse where experiments are used to confirm theoretical predictions. Theoretical energy-release forms from pellet microexplosions are described in Table I. For the bare (D+T) pellet, prompt x rays would be observed first. Next in time would follow the 14-MeV neutrons, then the plasma of pellet debris. For structured pellets, the energy release mechanisms observed just outside the expanding pellet will depend on the pellet yield and on the composition and mass of the structural con- tainer. The fractional energy release as x rays will be larger than for the bare pellet, but with softer spectra. However, a high-energy gamma- ray component appears due to (n,y) scattering reactions. Most of the 14-MeV neutrons escape the pellet with slight degradation in energy. Laser Requirements

The fundamental requirements on the laser system are established by the characteristics of fusion pellets. These requirements will vary to some extent, depending on fuel-pellet design and size. The basic pellet- determined requirements for the laser system are concerned with: (1) pulse intensity, (2) pulse duration, (3) wavelength, and (4) spatial and temporal pulse shape. A second set of criteria are those which are determined by the energy balance and economics in a laser-fusion electric generation station: (1) net laser efficiency, (2) pulse repetition rate, (3) costs (capital and operating), and (4) reliability and mean lifetime of components (especially power supplies and switches).

61 The most demanding requirement is the generation of high-energy pulses of a nanosecond or less duration which necessitates the achievement of the inverted population state nearly simultaneously throughout the lasing medium. Several types of laser systems are being studied in 5-7 laser-fusion programs throughout the world. These systems differ in the physical approach utilized to produce population inversions in the respective lasing media. In general, pulse shaping and power amplification are performed in separate laser stages. The initial stage is a low-power oscillator with modulators placed in a resonant cavity to produce a single, short (mode-locked) pulse with a controlled pulse shape. This initial pulse is amplified in passing through one or more amplifier stages.

Solid-state and liquid lasers are normally pumped with photons from flashlamps. Some gas lasers, e.g., CO2, are pumped with electrons from an electric discharge. Other gas lasers, e.g., HF, use exothermic chemical energy for pumping. The most common laser for current laser fusion research utilizes neodymium-doped glass as the lasing medium. Xenon flash lamps optically pump neodymium ions which are embedded in glass rods or disks. Laser pulses with energies in hundreds of joules and pulse durations of 109 to 10 11 s are obtained. Although it may be possible, in principle, to increase the energy level of the neodymium- glass system to that needed for successful pellet fusion, the efficiency (laser energy output to electrical energy input) of this system is fundamentally limited to about 0.1 to 0.2%. This limitation, along with inherent limitations on pulse repetition rate and glass damage from self focusing makes it a poor candidate for commercial power generation. The

CO2 laser, although having less favorable wavelength characteristics, is much more energetically efficient (potentially 5-7%) and is easily adaptable to the high repetition rate and continuously renewable lasing medium required for economic energy applications. Thus, for the present, the CO2 laser has been chosen as the basis of LFR concept studies.

Laser development is advancing rapidly, and it is impossible to predict the specific laser type, or types, that may ultimately be most advantageous for application in LFR systems. Lasing media now being evaluated experimentally include CO2 , HF, oxygen, excimers, and iodine with characteristicswith characteristics tabulated in Table II.

62 III. LASER FUSION REACTOR CONCEPTS

Conceptual designs of LFRs and electric generating stations are 9 10 being investigated at several laboratories in the US9 10 and in Europe.l Differences in projected fusion-pellet design and microexplosion energy-release characteristics between various investigators have resulted in different basic approaches to the design of reactor cavities and other generating station subsystems. There are economic incentives for maximizing pellet-microexplosion repetition rates. The feasibility studies of reactor cavity and blanket concepts discussed here are based on the use of fusion pellets consisting of solid spheres of (D+T) with a yield of 100 MJ. The calculated characteristics of the energy release mechanisms are those given in Table I. Although pellet designs for ultimate commercial application may differ substantially from that chosen for these studies, the pellet output characteristics will be sufficiently similar (i.e., the major fraction will still be 14-MeV neutrons) that LFR engineering concepts based on this pellet concept should be generally applicable to other reactor concepts.

Wetted Wall Reactor Concept

The wetted-wall LFR concept is shown in Fig. 1. The reaction chamber or reactor cavity is spherical and is surrounded by a blanket region consisting of liquid lithium and structural components. The cavity wall is formed by a porous refractory metal through which coolant lithium flows to form a protective coating on the inside surface. The protective layer of lithium absorbs the energy of the high-energy alpha particles, the pellet debris, and part of the x-ray energy. Part of the lithium layer is evaporated and ablated into the cavity by each pellet microexplosion and is subsequently exhausted through a supersonic nozzle into a condenser. The ablative layer is restored between pulses by radial inflow of lithium from the blanket region. The minimum required thickness of the protective lithium layer is determined by the amount of lithium that could be vaporized by each pellet microexplosion and by the desired protection of the cavity wall from surface heating by x rays. Analyses of lithium flow through the porous wall and along its inner surface indicate that 1 to 2 mm minimum- thickness lithium layers can be restored in less than 1 s.3 The minimum thickness of lithium on the interior of the cavity wall and the maximum allowable wall-temperature increase due to x-ray energy deposition

63 enable determination of the minimum permissible cavity diameter. The minimum cavity diameter for pure (D+T) 100-MJ microexplosions is X 3.4 m. The maximum amount of lithium that could be vaporized is X 1.25 kg per microexplosion, which corresponds to a layer on the inner cavity wall less than 0.1 mm thick. Analyses have also been made of cavity blowdown phenomena. Depending on the wavelength of the laser light utilized to implode and heat the pellets, it may be necessary to evacuate the cavity to a lithium density of % 1017 atoms/cm for efficient penetration by the laser beams. The time required to restore the cavity to this condition after a pellet microexplosion is % 0.8 s. From this and other considerations, it appears that 100-MJ pellet-microexplosion repetition rates of about one per second, corresponding to a thermal power level of 100 MW per cavity, will be practical for the wetted-wall reactor concept. Conceptual reactor designs include a tube through which pellets are injected pneumatically. Pellet guidance and tracking systems will also be required. To provide reasonably symmetric illumination of the pellet by laser light, eight laser-beam-transport tubes are arranged symmetrically around the reactor cavity. Blanket structures have not been designed in detail, however, analyses have been made of conceptual designs in which the liquid lithium is contained between concentric structural shells enclosing the reactor cavity. Designs that have minimum structural masses and that also have acceptable tritium breeding ratios include three structural shells in addition to the porous cavity wall. The porous cavity wall is supported by the innermost structural shell. The momentum from the ablation of lithium from the interior surface of the cavity wall is transmitted through the relatively incompressible lithium to other structural components. Structural shell thicknesses have been calculated to contain 100-MJ pellet microexplosions without exceeding fatigue stress limits for either niobium, molybdenum, or stainless steel at temperatures up to 1000 K. Because the energy deposition times are very short (< 106 s) compared to shell natural frequencies (% 10 3 s), the shells respond to the impulsive loads by ringing at essentially their natural frequencies, modified to the extent that they are hydrodynamically coupled to the liquid-lithium blankets. If the shell structure is to be stable, the ringing hoop stresses must be damped between successive pellet burns. Dynamic analyses indicate that adequate damping does occur and that the stresses are completely damped in less than 100 ms after pellet burn.l2

64 The lithium flow path envisioned for the wetted-wall reactor intro- duces the lithium at the outer surface of the porous cavity wall by means of structures concentric with the beam-transport tubes. The lithium then flows radially outward through the blanket. Uniform radial flow is achieved by including sufficient impedance to flow in the successive structural shells.

Magnetically Protected Cavity Wall Concept

Protection of reactor cavity walls from energy deposition and erosion by energetic charged particles by means of magnetic fields is an attractive conceptual alternative to ablative cavity liners. The essential features of a magnetically protected reactor concept are shown schematically in Fig. 2. The central portion of the cavity is cylindrical, with an impressed steady-state magnetic field (BZ) produced by a solenoid located concentric with and exterior to a lithium blanket region. The alpha particles and the ionized particles in the pellet debris resulting from fusion-pellet microexplosions are diverted by the magnetic fields to conical energy sinks in the ends of the cylindrical cavity. Cavity phenomena have been investigated for operation with low ambient gas pressure (< 101 atoms/cm3) and with the maximum cavity gas pressure (% 1017 atoms/cm3 ) through which intense laser beams can be transported efficiently. The high-energy alpha particles expand with an average initial radial velocity of X 9.8 x 106 m/s. Computer simulations for low ambient gas density show that the high-energy alpha particles act as single particles, going into gyro-orbits (radius % 1 m for B = 0.2 T) and spiraling out the ends to the conical energy sinks.1 During the time of flight (^ 5 x 10-7 s) of the bulk of the alpha-particle plasma to the conical energy sinks, the slower debris plasma is initially streaming at an average velocity of % 1.5 x 106 m/s. The debris plasma acts collectively; it excludes and then compresses the magnetic field between the plasma and cavity wall with pressure balance occuring at ' 2-m radius for BZ = 0.2 T. After several cycles of successive radial expansions and compressions of the debris plasma, it too will have ex- panded out the ends of the cylinder to the energy-sink regions. The cavity diameter (5 m) indicated in Fig. 2 was selected somewhat arbitrarily. Minimal cavity diameters will be constrained by allowable wall-surface temperature increases due to x-ray energy deposition. Cavity liners of materials with low atomic number (e.g., carbon) are

65 useful for decreasing metal-wall surface-temperature fluctuations. The geometry shown in Fig. 2 permits energy sinks to be designed with large surface areas. The surface area of each cone available for energy deposition by charged particles is more than ten times the cross sec- tional area of the cylindrical portion of the cavity. A high-temperature material such as a refractory metal carbide is envisioned for the energy- sink surface. Fringing of the magnetic field should permit tailoring the energy deposition density over the surfaces of the energy sinks. Liquid lithium might be used as a coolant and fertile material for the breeding of tritium in the annular blanket regions. Axial flow of lithium in the blanket annulus minimizes problems relating to pumping a conducting fluid across magnetic field lines. The solid angle subtended by the energy sinks is only X 10% of the 41 steradians through which the neutrons from pellet microexplosions expand. Preliminary estimates indicate that adequate tritium breeding ratios to sustain the fuel cycle can be obtained from nuclear reactions with lithium in the annular blanket regions alone. Thus, the conical energy sinks could be cooled by a fluid other than lithium, e.g., helium. There are several potential advantages of magnetic protection of cavity walls compared to other reactor concepts that have been considered. It is anticipated that thermonuclear-reactor component life- times will be severely limited by the rate at which damage occurs from products of fusion. Because power costs are dominated by capital investment, component replacement schedules, and duty factors, it is important to design simple, long-lived reactor cavities of minimum size with expendable components incorporated in a manner permitting rapid and convenient replacement. The conical energy sinks are readily accessible for replacement without disturbing the lithium blanket, the laser-beam optics, the solenoid, or the fuel injection system. Other major advantages of this concept are the possibility of achieving high pellet-fusion repetition rates and the elimination of involved procedures for removal of evaporated and/or ablated materials from the reactor cavity between successive pellet microexplosions. Also, the use of magnetic fields in this manner will eliminate streaming of charged particles through the laser-beam-transport tubes which might otherwise damage last optical surfaces. Computer simulations show that the effects of the magnetic field introduce a time spread over which the plasma reaches the energy sinks compared to free-streaming particles. This time spread may be helpful in reducing energy sink surface deteriora- tion by allowing time for conductive heat transfer.

66 Additional Reactor Concepts

A laser-fusion reactor concept, referred to as a suppressed-ablation 9 design, has been proposed that is similar to the wetted-wall design described above.

The diameter of the reactor cavity for the suppressed ablation concept is somewhat larger (u 4.4 m) than the diameter of the cavity in the wetted-wall design, and the cavity wall surface area is further increased by constructing it from pyramidal surfaces whose triangular bases form the first wall plane. The interior surface of the first wall

is protected by an X 300 pm thick layer of lithium that is pumped by capillary action from reservoirs. Each fusion-pellet microexplosion releases 7 MJ of thermonuclear energy. Because of increased cavity wall surface area, enhanced thermal conduction from the protective lithium layer to the bulk coolant, and lower pellet yield, lithium evaporation is diminished considerably. Thus, the time required after a pellet microexplosion to return the cavity to conditions permitting a subsequent pellet microexplosion is much shorter than for the wetted-wall design, and a pulse repetition rate of 10 microexplosions/s is thought possible. The reactor blanket is icosahedral with 12 laser beams that penetrate the blanket at the vertices of the icosahedron. The blanket is of modular construction and consists of 20 equilateral truncated triangular prisms. The blanket modules are constructed of niobium and can be extracted singly for replacement in the event of damage. The blanket coolant is liquid ltihium. The SATURN reactor conceptll represents an extension of some aspects of the suppressed-ablation design. The cavity and blanket are formed from polygonal shaped power and vacuum modules. Each power module, of which there are ' 1100, contains a blanket portion and a complete power conversion system (turbine and generator). The blanket portion is cooled by neon for energy conversion in a Brayton cycle. There are

X 70 vacuum modules with pumping ports in the blanket portions and pumps instead of power conversion systems. The cavity diameter is ' 20 m, and the inner surface of the cavity wall is not protected from x rays and charged particles. A pellet yield of 50 MJ and a pulse repetition frequency in the range of 10 to 100 Hz are proposed. A unique reactor cavity concept, called a lithium vortex reactor or BLASCON, has no cavity wall per se; rather a cavity is formed by a

67 vortex in a rotating pool of lithium in which fusion-pellet microexplosions take place. Rotational velocity is imparted to the circulating lithium by tangential injection at the periphery of the reactor pressure vessel. The lithium flows out of the spherical pressure vessel through a central port at the bottom. Bubbles of inert gas are injected into the lithium jets entering the vessel to provide an average void fraction of 2 or 3%. These bubbles serve to cushion the shock wave from the pellet microexplosion and thus reduce the stresses in the pressure vessel. Fusion pellets are injected into the lithium vortex through the top of the reactor vessel, and a single laser beam illuminates the pellet, also from the top. This concept has been proposed for fusion-pellet yields of X 1000 MJ and pulse rates of 2 microexplosions per second.

IV. ELECTRIC GENERATING STATION CONCEPTS

A simplified energy and mass flow diagram is shown in Fig. 3. Important considerations which lead to plant design choices include component reliability, redundancy of essential components, access to components for service and/or replacement, and minimization of hazards from radioactive materials to the environment and to operating personnel. A conceptual electric generating station design based on the wetted- 10 wall LFR concept is shown in Fig. 4. The reactors are located in a separate, annular building which encloses the laser system building. The number of reactors required for a given net power output depends on the efficiency of the energy conversion cycle and thus on the temperature of the reactor coolant. Pairs of adjacent LFRs are served by a common heat-transfer loop, a steam generator, and lithium-processing and tritium- removal systems. Each reactor is in a biologically shielded enclosure with penetrations for laser beams, liquid-metal coolant, and the introduction of fuel. The heat exchangers and lithium processing equipment for each pair of reactors are located in a biologically shielded enclosure adjacent to the reactor enclosures. Components containing tritium are designed to minimize component sizes and piping lengths. The laser system includes 16 separate main CO2 laser power amplifiers. Eight of these 16 lasers are fired simultaneously, and the eight laser beams are directed successively to respective reactor cavities by a rotating mirror. Each laser has a redundant partner to achieve high reliability and ease of maintenance.

68 The laser power supplies are located in the laser building above the main laser power amplifiers. Mechanical and structural isolation is provided for the laser system and the reactors and associated beam-transport and heat-transfer systems. Control rooms and other work areas are isolated from the reactor radioactive areas. Reactors and reactor components can be removed remotely through removable shield plugs and transferred to shielded work areas by a crane. Each reactor can be isolated from the system for service and/or replacement without affecting the operation of the remainder. Since the laser subsystem represents a significant fraction of a LFR generating station, it is economically advantageous to centralize components so that each laser system serves several reactors. Central- ized laser systems require fast beam switching from laser power amplifiers to selected beam ports. Laser-beam switching in preliminary generating station concepts is accomplished by rotating mirrors. The rotating mirror assembly consists of eight elliptical plane mirrors spaced uniformly about a rotating ellipse at 45 degrees to the beam direction toward the reactors. A stationary 45 degree mirror below the rotating mirror also consists of eight elliptical mirrors spaced around a circle. To achieve simultaneity of beam arrival at the fusion pellet within a small fraction of a nanosecond or less, the net path-length differences between various laser beams must be compensated. The most economical arrangement appears to be to adjust the path lengths between a master oscillator and the main laser power amplifiers. Arrangements for splitting the oscillator pulse into eight parts traveling different distances are easily devised.

An electric generating station concept based on the magnetically protected LFR is shown in Fig. 5. Four reactors with a thermal power output of % 1250 MW each are included in the station (compared with the wetted-wall reactor generating station concept which includes 20 reactors with a thermal power output of X 150 MY each). The major differences between this concept and the one based on the wetted-wall reactor design result from differences in the degree of modularization which lead to differences in the optimum number of redundant components and the potential advantages of centralizing components. The reactors, heat exchangers, lithium-tritium separators, control room, and energy conversion equipment are located on the first level of

69 the station. Hot-cell maintenance areas for periodic servicing of the magnetically-protected LFR energy-sink cones and other radioactive components are also on this level. Tracks are provided for movement of energy-sink cones between reactors and maintenance areas. Single-loop lithium heat-transfer systems are used between the reactors and the steam generators, and semipermeable membrane lithium-tritium separators are included in the lithium loops. Separate heat-exchanger and lithium- tritium separator systems are provided for each reactor. The pulse-forming networks are located on the second level and the main laser power amplifiers on the third level. There are 16 CO2 laser power amplifiers, 8 of which would be operated at one time to provide 8 laser beams for quasi-symmetric illumination of fusion pellets. Selec- tor mirrors are used to direct the laser beams from operating laser power amplifiers to the rotating mirror, also located on the third level. The required rotational velocity of the mirror is 10 revolutions per second. For the design laser-beam length, the laser beam focal spot travels only % 1 x 10 4 mm during a 1.O-ns pulse; thus, the focused beam will not move significantly off a millimeter-size target during the arrival -time of a laser pulse. A laser-power-amplifier and pulse- forming-network maintenance area is located on the third level which is serviced from ground level by a freight elevator. The front-end system, i.e., the oscillator and preamplifiers, is located on the top level. Differences in beam path length from the laser power amplifiers to the reactor cavity centers are compensated by corresponding differences in path lengths in the front end system (oscillator and preamplifiers) so that amplified laser pulses arrive at the cavity centers simultaneously. Each reactor can be isolated from the system for service without affecting the operation of the remainder.

V. SUMMARY AND CONCLUSIONS

The most critical unsatisfied technology requirements for laser fusion are those related to achieving significant fusion-pellet burn. These requirements include advances in laser technology and in fusion- pellet design and fabrication techniques. To date, laser-fusion experiments have yielded up to 107 neutrons with laser systems operating at a few tens of joules. These results, of course, have not indicated feasibility for commercial applications, but understanding of the fundamental physics of the laser-pellet interaction is being developed.

70 Within the next few years, 10 kJ laser systems will be operational and a clearer understanding of fundamentals will be gained. The major milestone of scientific breakeven, i.e., thermonuclear output equal to exceeding incident beam energy, is expected to require laser systems at powers exceeding 100 TW. Such a laser facility is planned for operation in the early 1980's. With the achievement of this milestone, the laser- fusion program would proceed from the research to the technology development phase, aimed at demonstrating the economic attractiveness of commercial exploitation in the late 1990's or early twenty-first century. The most critical parameter affecting the economics of a laser- fusion generating station is the product of laser efficiency and pellet gain. Obviously, this product must be greater than one for a net output of electricity and must be greater than two for commercial feasibility. Because laser efficiencies are likely to be less than 0.1, laser pellet gains must be greater than 20. Because it is felt that pellet gains greater than 100 are probably not achievable, the minimum laser efficiency of any proposed laser system must be greater than 0.02. Based on our current knowledge of the laser/pellet interaction, certain features of laser-fusion generating stations appear certain: o Conceptual LFRs are relatively small, compact systems and lend themselves naturally to the design of generating stations for a

range of power levels from % one hundred to several thousand mega- watts. Redundancy of essential components can be easily and economically incorporated in large power plants. o In a LFR, fusion pellet microexplosions must be contained in a manner that both prevents excessive damage to reactor components and permits recovery of the energy in a form suitable for utili- zation in an energy conversion cycle. Very-high-energy, short- pulse lasers are necessary for the compression and heating of fusion pellets to thermonuclear ignition conditions. The laser beams must be repetitively transported to and focused on pellets inside reactor cavities. o The fuel cycle that is receiving primary consideration is the deuterium-tritium cycle. Deuterium is easily and cheaply obtained from conventional sources; but it is expected that tritium will be produced, as needed, by reactions between fusion neutrons and lithium, which must be contained in blanket regions surrounding reactor cavities. Inner cavity walls must withstand pulses of x rays, 14-MeV neutrons, and energetic ionized particles that are released by the thermonuclear reactions. 71 Several LFR concepts are being evaluated to assess their feasibility, to define technology requirements, and to determine their prac- ticability for use in various applications. The two concepts that have been studied most extensively are known as the wetted-wall and the magnetically protected LFRs. These two fundamental approaches, together with variations, to the containment of fusion-pellet microexplosions and the recovery of thermonuclear energy for commercial use appear to be feasible and, moreover, to provide a basis for the conceptual design and evaluation of laser-fusion electric generating stations. While the direct production of electricity from LFRs in central generating stations is a principal objective of the Laser Fusion Pro- gram, there are other potential commercial applications that may prove to be no less important. Among such applications are the production of synthetic fuels, such as hydrogen, and providing high-temperature process heat that might be utilized in a variety of ways. Fusion neutrons can be used to breed Pu from 238U and 233U from Th. Systems designed for this purpose may be attractive compared to liquid-metal fast-breeder reactors. It is anticipated that many more significant applications of this nature will be discovered as laser fusion is developed and conventional fuels become more scarce.

REFERENCES 1. A. W. Ehler, D. V. Giovanielli, R. P. Godwin, G. H. McCall, R. L. Morse, S. D. Rockwood, "Evidence of Anomolously Reduced Thermal Conduction in Laser Produced Plasmas," Los Alamos Scientific Laboratory report LA-5611-MS (August 1975). 2. D. B. Henderson, Los Alamos Scientific Laboratory, unpublished work. 3. L. A. Booth (Compiler), "Central Station Power Generation by Laser Driven Fusion," Los Alamos Scientific Laboratory report LA-4858-MS, Vol. I (February 1972). 4. J. M. Williams, F. T. Finch, T. G. Frank, and J. S. Gilbert, "Engineering Design for Laser Controlled Thermonuclear Reactors," Proceedings of 5th Symp. on Engineering Problems of Fusion Research, Princeton, NJ (November 6-9, 1973). 5. T. J. Burgess, "Lasers for Fusion Systems," IEEE Trans. Plasma Science, PS-2, 26 (1973). 6. J. S. Gilbert, T. F. Stratton, and R. J. Jensen, "Potential Lasers for LCTR Applications," Proc. of the 1st Topical Meeting on the Technology of Controlled Nuclear Fusion, Am. Nuc. Soc. San Diego, CA (April 1974). 7. R. R. Buntzen and C. K. Rhodes, "Laser Systems for Laser Fusion," Proc. of 1st Topical Meeting on the Technology of Controlled Nuc- lear Fusion, Am. Nuc. Soc., San Diego, CA (April 1974).

72 8. O'Dean Judd, Los Alamos Scientific Laboratory, unpublished work. 9. J. Hovingh, J. Maniscalco, M. Peterson, and R. I. Werner, "The Preliminary Design of a Suprressed Ablation Laser-Induced Fusion Reactor," Proc. of 1st Topical Meeting on the Technology of Con- trolled Nuclear Fusion, Am. Nuc. Soc., San Diego, CA (April 1974). 10. J. Williams, T. Merson, F. Finch, F. Schilling and T. Frank, "A Conceptual Laser Controlled Thermonuclear Reactor Power Plant," Proc. of 1st Topical Meeting on the Technology of Controlled Nuclear Fusion, Am. Nuc. Soc., San Diego, CA (April 1974).

11. F. Bohn, H. Conrads, J. Darvas, S. Forster, "Some Design Aspects of Inertially Confined Fusion Reactors," presented at 5th Symp. on Engineering Problems of Fusion Research, Princeton, NJ (November 1973). 12. T. G. Frank and L. A. Booth, "Laser Fusion Reactor Materials Prob- lems Resulting from Fusion Microexplosion Emissions," presented at International Conference on Surface Effects in Controlled Fusion Devices, San Francisco, CA (February 1620, 1976). 13. J. J. Devaney, "Magnetically Protected First Wall for a Laser- Induced Thermonuclear Reaction," Los Alamos Scientific Laboratory report LA-5699-MS (August 1974). 14. D. A. Freiwald, D. O. Dickman, and J. C. Goldstein, "Computer Simulation of a DT Pellet Microexplosion in a Magnetically Pro- tected Laser Fusion Reactor," to be published in Bull. Am. Phys. Soc. (November 1975). 15. A. P. Fraas, "The BLASCON - An Exploding Pellet Fusion Reactor," Oak Ridge National Laboratory report TM-3231 (July 1971).

73 -4

I)A h

Cd 4 r-l >: : ·- ¢ o e o a,

.2) a)

1 t- 4UU

2 W cp O 0 i

0 0 0 - e- OO CC O O C 0Z ov.. u or-

X h O Ct 4i U

i U Z 4J4 O a4 -1 (a 3 42

X p tr~ Z U) .2) O ed o

0 0

H < o .2) O-3 0a o "O

oE ) 0a :4 L. h O1~~~~~~~~~~~~~-14~~~~~~~~. h

LU < -4O o o U- O; O .

CL>Dc: k. > &.

Z 42

t4· t o h

' '4 a.

2-. 2 0

74 -4 0.) ^ r-. 1*Ho\° I o LU) -1 1- (U Lf) l-4 0

- 4 --"

r< O l I-4 :j ve

Z 0 i-- CO o3 n.'>

LL o Lf) 00 '-l Cz: 4-M

U); t oI GIn v G-3) o- O 1 ti r - ¢ Bi > r~ LU) 1-4 < c; 0 1-4

-4 z LU [-- 0 P.

,-ox~-0 >> 4-) (L) t c-: h h I O 0 0 4J ()a> "->_ m oo -o 0 0

, C0 0M) X 0

td h

c 0 C> 0 0 i o)4. C7U CU ¢- [W X <£¢

75 Lithium Blanket Outer Shell

Structural Shell Porous Wall_

As, Inner Tube Structural Wall

Condenser

Fig. 1. Wetted-wall laser fusion reactor concept.

76 A- 0 cC 0 X= 4- X

O a ) -- -r E

C =3 > CL C) 0

04FJ

0o -r on a)ru -c * () de

ct) 0 - 4- CO' o C- 0

>.1

77 zo wI- rw

0.n T g) Z 4-

ag zZw .0 LL.> c

_ L n 0- - I- 0)I

cOp:~~~ 0

S aSI Z)""' I ar .~-I

c I $4 a3 _ W IA 0

W W ir * $

"Cd K A.t:~~~~~~t Id .. II CE 0)

XO~ 0:W~ i i ob

ww Wl- X r'n

78 0 0

0bo1:

-4k

bO 4c0

.-4

R, a>

*r-

79 0

n) si

o( ,4U)

Fi d) Inf A: U

0 - l

*c,

o a raC) CU

0 -H

bO LL

*l-

80 PA OVERVIEW OF THE PLANNING CN'SIDERPAIONS IN THE UNITED STATES INERTIAL CONFINEMENT FUSION PROGR'AM

DIVISION OF LASER FUSION U.S. ENERGY RESEARCH AID DEVELOP£ENT ADMINISTRAiTION WASHINGTON, D.C. 20545

Abstract

An overview of the United States research program in inertial confinement experiments is presented. The objective is to develop and demonstrate inertial confinement fusion as a proven energy technology.

IT IS A PLEASURE TO BE HERE TO DISCUSS THE U.S. RESEARCH PROGRAM IN INERTIAL CONFINEMENT EXPERIMENTS, I WELCOME THIS OPPORTUNIITY TO DISCUSS THE POTENTIAL OF INERTIAL CONFINEMENT FUSION AS A FUTURE, ESSENTIALLY INEXHAUSTIBLE SOURCE OF ENERGY. I WISH TO PRESENT HERE TODAY AN OVERVIEW OF THE U.S. PROGRAM FROM THE STANDPOINT OF THE PLANNING CONSIDERATIONS, GIVING PARTICULAR ATTENTION TO THE STATE OF THE TECHNOLOGY A;ND OUR PLANS FOR MOVING IT AHEAD TOWARD ULTIMATE CIVILIAN FUSION POWER. THE OBJECTIVE OF THE U.S, INERTIAL CONFINEMENT FUSION PROGRAM IS TO DEVELOP, DEMONSTRATE AND UTILIZE INERTIAL CONFINEMENT FUSION. PERHAPS AS EARLY AS THE 1990s IF THE COURSE OF RESEARCH IS FAVORABLE AND TECHNOLOGY DEVELOPMENT IS AGRESSIVELY PURSUED, INERTIAL CONFINEMENT FUSION COULD BE AVAILABLE AS A PROVEN AND DEMONSTRATED ENERGY TECHNOLOGY OFFERING DECISIVE ADVANTAGES OF FUEL AVAILABILITY, SCALE, CAPITAL COST, AND ENGINEERING PRACTICALITY OVER OTHER ENERGY OPTIONS. CALCULATIONS OF FUSION YIELDS FROM SMALL AMOUNTS OF DEUTERIUM-TRITIUM, MODELED WITH THE HELP OF COMPUTER CODES

81 INDICATE THAT PELLETS CONTAINING THIS FUEL COULD BE IMPLODED BY HIGH ENERGY DRIVERS, THE COLLAPSING MATERIAL OF THE SHELL SERVING TO COMPRESS THE FUEL TO SUFFICIENTLY HIGH DENSITIES TO CAUSE A SELF-PROPAGATING FUSION REACTION THROUGHOUT THE FUEL MASS. THE OBJECTIVE OF U.S. GOVERNMENT-FUNDED RESEARCH INTO THIS PROCESS SINCE THE MID-1960S HAS BEEN TO DETERMINE WHETHER A PRACTICAL IGNITION SOURCE COULD BE FOUND TO FOCUS ENOUGH ENERGY ON SUCH A PELLET IN A SUFFICIENTLY SHORT TIME TO CAUSE THE CONFINEMENT OF THE FUSION REACTION BY THE INERTIA OF THE IMPLODING PELLET. THE EXACT PERFORMANCE REQUIRED OF AN IGNITION SOURCE TO DELIVER STABLE, FOCUSABLE POWER IN WELL-FORMED PULSES TO THE FUEL PELLET, AND THE REQUIREMENTS FOR DIAGNOSTICS AND FOR EFFICIENT PELLET DESIGNS THAT WILL RELAX POWER REQUIREMENTS IN THE IGNITION SOURCE, HAVE LED TO A DISTRIBUTION OF PROGRAM EFFORTS INTO AREAS OF TECHNOLOGY DEVELOPMENT ALONGSIDE THE PRINCIPAL EFFORT IN TARGET INTERACTION EXPERIMENTATION. THE DEVELOPMENT OF IGNITION SOURCES OF INCREASING POWERS, WHICH REQUIRE NEW MATERIALS AND CONTROL DEVICES FROM BEAM FORMATION AND FOCUSING, DEMAND LARGE CAPITAL INVESTMENT AND LONG LEADTIMES. IGNITION SOURCES ARE ACCORDINGLY BEING DEVELOPED BOTH AS RESEARCH TOOLS FOR EXPERIMENTATION AND SIMULTANEOUSLY AS CANDIDATE DRIVERS FOR AN EXPERIMENTAL FUSION REACTOR. THE FULL UNDERSTANDING OF IMPLOSION PHYSICS THAT IS REQUIRED FOR PROOF OF SCIENTIFIC REASIBILITY IS ALSO NEEDED TO CHARACTERIZE FULLY THE VARIOUS CANDIDATE IGNITION SOURCES FOR USES IN LATER STAGES OF THE PROGRAM. INERTIAL CONFINEMENT FUSION RESEARCH TO DATE HAS PRODUCED PELLET IMPLOSIONS WITH OBSERVED THERMONUCLEAR NEUTRON PRODUCTION, THUS VERIFYING THEORETICAL PREDICTIONS IN A SATISFACTORY MANNER FOR VERY LOW POWER LEVELS. PELLET COMPRESSIONS AND NEUTRON YIELDS ARE NOW MANY TIMES LARGER THAN THOSE OBSERVED ONLY THREE TO FOUR YEARS

82 AGO. HOWEVER, VOLUMETRIC COMPRESSIONS TEN TIMES AS GREAT, AND PELLET ENERGY YIELDS SEVERAL ORDERS OF MAGNITUDE GREATER, ARE CALCULATED TO BE REQUIRED TO DEMONSTRATE THE SCIENTIFIC FEASIBILITY OF INERTIAL CONFINEMENT FUSION. SCIENTIFIC UNCERTAINTIES REMAIN ABOUT THE PRECISE NATURE OF THE PHYSICS OF BEAM ABSORPTION BY TARGET MATERIALS, AND THE MANNER IN WHICH IMPLOSION PHENOMENA OBSERVED AT LOW ENERGY LEVELS SCALE TO HIGHER POWERS. THE NEAR-TERM GOAL OF THE INERTIAL CONFINEMENT FUSION PROGRAM IS TO GAIN AN UNDERSTANDING OF THE NATURE OF LASER AND PARTICLE BEAM INTERACTIONS WITH MATTER, BY THE MID-L980S, IF PRESENT THEORETICAL CALCULATIONS HOLD, THE SCIENTIFIC FEASIBILITY OF FUSION ENERGY BY INERTIAL CONFINEMENT SHOULD HAVE BEEN DEMONSTRATED, IN THE MIDDLE TERM ONE OR MORE IGNITION SOURCES MUST BE DEVELOPED FOR RELIABLE OPERATIONS, A SUITABLE REACTOR MUST BE DESIGNED AND DEVELOPED, AND DEDICATED FACILITIES FOR MATERIALS TESTS. AS EARLY AS THE 1995-2000 TIME PERIOD IT MAY BE POSSIBLE TO BEGIN COMMERCIALIZATION OF INERTIAL CONFINEMENT FUSION TECHNOLOGY AS A SAFE, ENVIRONMENTALLY ACCEPTABLE, AND ECONOMICAL ENERGY SOURCE. THE PROGRAM IS EXPECTED TO EVOLVE THROUGH THE PHASES OF RESEARCH, DEVELOPMENT, AND DEMONSTRATION SHOWN IN FIGURE 1. AS CRITERIA FOR SCIENTIFIC FEASIBILITY ARE ESTABLISHED AND MET, PRIORITY WILL SHIFT TO THE TIMELY IDENTIFICATION OF THE MOST PRACTICAL OF SEVERAL UFSION IGNITION SOURCES DEVELOPED EITHER AS TOOLS FOR THE RESEARCH PHASE (ND:GLASS AND C02 LASERS AND ELECTRON BEAM ACCELERATORS) OR AS PROMISING HIGH-EFFICIENCY DRIVERS FOR REACTORS (SEVERAL CANDIDATE NEW LASERS, AND HEAVY IONS). STUDIES OF REACTOR TECHNOLOGY AND PARTICULARLY OF MATERIALS QUALIFICATION REQUIREMENTS, WHICH WILL HAVE MANY COMMON ELEMENTS WITH WORK UNDER OTHER PROGRAMS IN ERDA, WILL BE NECESSARY AS AN ONGOING EFFORT TO CONTRIBUTE TO THE SELECTION OF THE APPROPRIATE ENERGY SOURCE

83 TECHNOLOGY, REACTOR TECHNOLOGY DEVELOPMENT IS EXPECTED TO DOMINATE PROGRAM EFFORTS BY THE END OF THE 1980s. THE RELATIONSHIP OF THESE THREE PRINCIPAL ELEMENTS TO MAJOR PROGRAM TURNING POINTS APPEARS SCHEMATICALLY IN FIGURE 2, THE INTERACTION OF THEORETICAL CALCULATION, EXPERIMENTAL VERIFICATION, PELLET AND DRIVER TECHNOLOGY DEVELOPMENT, AND DETERMINATION OF REACTOR TECHNOLOGY REQUIREMENTS IS NECESSARY TO ENSURE THAT, AS THE PROGRAM MOVES OUT OF THE PRESENT PHASE DOMINATED BY APPLIED RESEARCH, EACH OF THESE ELEMENTS IS SUFFICIENTLY IN HAND THAT NO LONG DELAYS ARE IMPOSED BY LAGGING ELEMENTS IN THE PROGRAM, IN THE 1981-85 TIME PERIOD, IT SHOULD BE POSSIBLE TO ANSWER THE REMAINING QUESTIONS ABOUT THE PHYSICS OF INERTIAL CONFINEMENT FUSION, THE POWER SOURCE AND PELLET REQUIREMENTS THAT MUST BE MET THROUGH APPLICATIONS ENGINEERING, AND THE EFFORT THAT WILL BE REQUIRED TO DEVELOP REACTOR TECHNOLOGY SUITABLE FOR POWER PRODUCTION BASED ON INERTIAL CONFINEMENT. IF PROGRAM BALANCE IS MAINTAINED, AND THE PROGRESS OF RESEARCH HOLDS NO MAJOR ADVERSE SURPRISES, IT SHOULD BE PRACTICAL BOTH TO DEMONSTRATE SCIENTIFIC FEASIBILITY AND DELINEATE THE REQUIREMENTS FOR ENGINEERING DEVELOPMENT AND CONMERCIALIZATION WITH SOME DEGREE OF CONFIDENCE IN THIS TIME PERIOD, FIGURE 3 SHOWS THIS TIME PERIOD AS A MAJOR DECISION PHASE OF THE PROGRAM, NUCLEAR FUSION RESULTS IN RADIATION, MOST OF WHICH TAKES THE FORM OF HIGH-ENERGY NEUTRONS. ONCE CONTROLLED FUSION IS ACHIEVED BY INERTIAL CONFINEMENT, THE EFFICIENCY OF THE DRIVING POWER SOURCE AND THE FUEL PELLET YIELD WILL DETERMINE WHETHER ONE OR MORE OF SEVERAL POWER APPLICATIONS ARE FEASIBLE. PELLET ENERGY GAINS (ENERGY OUT VERSUS LIGHT ENERGY IN), NOW WELL UNDER ONE, SHOULD REACH UNITY BY 1981. WHEN HIGH-ENERGY LASER SYSTEMS WILL BECOME AVAILABLE FOR TARGET EXPERIMFNTS. BY THE MID-1980S, WHEN

84 THE DESIGN OF AN EXPERIMENTAL POWER REACTOR SHOULD BE WELL UNDER WAY OR COMPLETED, PELLET GAINS SHOULD EXCEED 100. LASER ENERGY EFFICIENCIES--THE RATIO OF LASER BEAM ENERGY TO POWER REQUIRED TO DRIVE THE LASER--NOW RANGE FROM BELOW 1 PERCENT FOR NEODYMIUM GLASS TO 1-2 PERCENT FOR C02. ALTERNATE LASER MEDIA NOW UNDER EXAMINATION MAY EVENTUALLY DEMONSTRATE 10-20 PERCENT EFFICIENCIES. ION BEAMS HAVE A THEORETICAL EFFICIENCY ON THE ORDER OF 30-50 PERCENT, AND ELECTRON BEAMS MAY BE CAPABLE OF 50 PERCENT OR MORE EFFICIENCY, DEPENDING ON ACTUAL PELLET GAINS ACHIEVED, EFFICIENCY REQUIREMENTS MAY BE RELAXED. A COMBINATION OF REAL AND THEORETICAL LASER AND ELECTRON AND ION BEAM MACHINE EFFICIENCIES, AND RELATIVELY MODEST PELLET GAINS, WOULD YIELI THRESHOLDS FOR VARIOUS APPLICATIONS AS SHOWN IN FIGURE 4.

NEGLECTED IN THIS HIGHLY TENTATIVE ANALYSIS ARE REJECT HEAT-- LARGELY A PRODUCT OF THE INEFFICIENCY OF THE POWER SOURCE--AS WELL AS THE POWER THAT MUST CIRCULATE TO DRIVE REPETITIVE OPERATION OF THE POWER SOURCE. PRELIMINARY ANALYSIS INDICATES THAT CIRCULATING POWER SHOULD NOT EXCEED 25 PERCENT OF PRODUCT POWER IF A DEMONSTRATION POWER PLANT IS TO BE COMPETITIVE WITH ALTERNATIVE POWER TECHNOLOGIES,

PRACTICAL EFFICIENCIES AS OPPOSED TO THEORETICAL EFFICIENCIES FOR VARIOUS DRIVERS ARE STILL UNDER STUDY. BOTH LASERS AND RELATIVISTIC ELECTRON AND ION BEAMS ARE STILL UNDER INTENSIVE DEVELOPMENT THAT COULD RESULT IN CLOSE APPROXIMATION TO THE THEORETICAL EFFICIENCIES. NEODYMIUM GLASS MIGHT APPROACH 1-2 PERCENT AND C02 UPWARDS OF 10 PERCENT AS THEORETICAL MAXIMA. COUPLED WITH HIGH PELLET YIELDS, THESE BASIC DRIVERS FOR RESEARCH COULD ALSO BECOME DRIVERS FOR POWER PRODUCTION OR AT LEAST FOR APPLICATIONS SUCH AS MATERIALS RESEARCH AND TESTING.

85 THE MOST DIFFICULT--AND PROBABLY THE MOST DESIRABLE TO MEET COMMERCIAL DEMAND--WOULD BE THE CONVERSION CYCLE OF HEAT TO ELECTRICAL POWER GENERATION, THIS CONVERSION CYCLE WOULD ALSO OFFER SIGNIFICANT ADVANTAGES OF SCALE IF INERTIAL CONFINEMENT TECHNOLOGY CAN BE DEMONSTRATED TO BE PRACTICAL AND EFFICIENT IN PLANTS OPERATING IN THE 500-1500 [W1E RANGE. SAFETY AND ENVIRON- MENTAL FACTORS ARE ALSO LIKELY TO BEMRET TO BEST ADVANTAGE BY THE ELECTRICAL POWER GENERATION CYCLE, ESPECIALLY IF REJECT HEAT CAN BE EFFICIENTLY RECYCLED THROUGH THE PLANT, THE HEAT CYCLE WOULD CAPTURE NEUTRON AND OTHER RADIATION OF THE FUSION REACTION IN A LITHIUM BLANKET, RESULTING IN HEAT TRANSFER THROUGH AN ELECTRICAL GENERATING AND POWER CONDITIONING UNIT TO DRIVE THE POWER SOURCE (C 25 PERCENT) AND SUPPLY A COMMERCIAL POWER GRID. THE POWER CONDITIONING EQUIPMENT WOULD BE THE CONVENTIONAL STEAM-HEAT CONVERSION PROCESS TO ELECTRICAL POWER OR POSSIBLY AN IMPROVED PROCESS DEVELOPED OVER THE NEXT QUARTER OF A CENTURY,

THE RADIOLYTIC CYCLE WOULD INVOLVE NEUTRON BOMBARDMENT OF HYDROGEN CONTAINED IN A SUPERSATURATED STEAM ENVELOPE TO PRODUCE NEGATIVELY CHARGED HYDROGEN ATOMS AND FREE OH+ RADICALS. THE LATTER WOULD BE CAPTURED BY BORON OR ANOTHER SUITABLE REAGENT WHILE THE HYDROGEN COULD BE USED DIRECTLY AS FUEL OR CONVERTED BY KNOWN PROCESSES TO METHANE (A SUBSTITUTE FOR NATURAL GAS) OR METHANOL (A SUBSTITUTE FOR GASOLINE). RADIOLYTIC SEPARATION OF HYDROGEN HAS NOT YET BEEN DEMONSTRATED TO BE EFFICIENT, AND A THEORETICAL UPPER LIMIT FOR THE CHEMICAL REDUCTION IS IMPOSED BY RECOMBINATION TO WATER MOLECULES.

THE BREEDER CYCLE WOULD EMPLOY NEUTRON BOMBARDMENT OF

THORIUM TO PRODUCE U233 OR U238/U 235 TO PRODUCE PLUTONIUM FOR

86 USE IN CONVENTIONAL FISSION REACTORS. BREEDER TECHNOLOGY, WHETHER DRIVEN BY INERTIAL CONFINEMENT FUSION OR FISSION BREEDER REACTORS, IS STILL UNDER DEVELOPMENT,

A HYBRID POWER SYSTEM BASED ON INERTIAL CONFINEMENT TECHNOLOGY MIGHT BE APPLIED NORMALLY TO EITHER THE RADIOLYTIC OR THE BREEDER CYCLE WHILE SERVING AS A STAND-BY OR PEAK DEMAND PERIOD COMPONENT OF AN ELECTRICAL POWER GRID. SYSTEM STUDIES HAVE CHARACTERIZED THE POWER APPLICATIONS INHERENT IN FUSION TECHNOLOGY SUFFICIENTLY WELL TO PERMIT CONFIDENT PROJECTIONS OR POWER APPLICATIONS ONCE THE SCIENTIFIC FEASIBILITY OF CONTROLLED

FUSION HAS BEEN DEMONSTRATED AND A TECHNOLOGY BASE FOR COMMERCIAL DEMONSTRATION HAS BEEN BUILT. THE THEORETICAL CALCULATIONS ON WHICH FIGURE 3 IS BASED HAVE A REASONABLE CONFIDENCE LEVEL THROUGH THE DEMONSTRATION STAGE, AT WHICH TIME ONGOING CALCULATIONS CONCERNING PREVAILING MARKETS, ECONOMIES OF SCALE, AND ENVIRONMENTAL AND SAFETY FACTORS CAN BE REFINED.

BY THE MID-1980Sj IMPLOSION PHYSICS UNDERSTANDING SHOULD HAVE PROGRESSED TO THE POINT AT WHICH CONCEPTUAL DESIGN OF AN EXPERIMENTAL POWER REACTOR IS UNDERWAY IN ONE OR MORE OF THE HIGH ENERGY LASER OR ELECTRON BEAM FACILITIES IN THE INERTIAL CONFINEMENT FUSION PROGRAM. A DEDICATED MATERIALS TESTING FACILITY, MAY BE REQUIRED BY THE CIVILIAN POWER PROGRAM.

THE TOTAL INVESTMENT REQUIRED TO DEVELOP AND COMMERCIALIZE INERTIAL CONFINEMENT FUSION MAY BE AS HIGH AS $10 BILLION OVER AND ABOVE RESEARCH FUNDED TO DATE, SIGNIFICANT PRIVATE INVESTMENT IN TERMS OF THAT TOTAL CANNOT BE EXPECTED BEFORE AN OPERATING FACILITY HAS DEMONSTRATED RELIABLE, ECONOMICJ SAFE AND ENVIRONMENTALLY ACCEPTABLE OPERATION, ON A SCALE AND WITH OPERATING CHARACTERISTICS THAT MEET A PREVAILING MARKET DEMAND. CRITICAL

87 DECISIONS MUST BE MADE AT SOME POINTS IN THE PROGRAM ABOUT THE RATE OF DEVELOPMENT THAT SHOULD BE PURSUED, ONCE THE FEASIBILITY OF DEVELOPMENT TO COMMERCIAL APPLICATION HAS BEEN ESTABLISHED. BECAUSE ALTERNATIVE TECHNOLOGIESJ NOTABLY MAGNETIC CONFINEMENT FUSION, HOLD LIKE PROMISE AND DEMAND COMPARABLE INVESTMENTSj THE CRITICAL DECISIONS MAY BE INFLUENCED AS MUCH BY CONDITIONS EXOGENOUS TO THE PROGRAM AS BY THE POSSIBILITIES INHERENT IN THE DEMONSTRATED FEASIBILITY OF INERTIAL CONFINEMENT FUSION, THE OPTIMAL STRATEGY FOR PROGRAM DEVELOPMENT, CONSISTENT WITH AN OVERALL NATIONAL ENERGY RESEARCH, DEVELOPMENT AND DEMONSTRATION PLAN, APPEARS TO BE THE COMPLETION OF PHASE ONE OF THE PROGRAM AS SOON AS POSSIBLE, THE DEMONSTRATION OF FEASIBILITY WOULD QUALIFY INERTIAL CONFINEMENT FUSION AS A PROVEN CANDIDATE TECHNOLOGY FOR DEVELOPMENT TO MEET LONG-TERM NATIONAL ENERGY NEEDS. SHOULD THE NATIONAL ENERGY STRATEGY FAVOR PROCEEDING WITH THE SECOND PHASE OF ENGINEERING DEVELOPMENT, LARGE INVESTMENTS WOULD BE REQUIRED TO INDUSTRIALIZE THE PELLET TECHNOLOGY, SCALE ONE OR MORE INPUT POWER SOURCE SYSTEMS TO HIGH ENERGY OPERATION WITH RELIABILITY AND EFFICIENCY, AND DESIGN AND DEVELOP THE REACTOR TECHNOLOGY REQUIRED FOR PRE-COMMERCIAL DEMONSTRATION. THE DEVELOPMENT AND DEMONSTRATION PHASES OF THE PROGRAM WOULD BE EXTREMELY COSTLY, BUT A DECISION TO ENTER UPON COMMERCIALIZATION WOULD IMPLY SELECTION OF INERTIAL CONFINEMENT FROM AMONG AVAILABLE LONG-TERM ENERGY OPTIONS AND ADEQUATE INSTITUTIONAL PREPARATION TO ENSURE REASONABLY RAPID ADOPTION OF THE DEMONSTRATED TECHNOLOGY BY THE PRIVATE SECTOR, THE PRESENT APPROACH TO IMPLEMENT AN AGGRESSIVE DEVELOPMENT STRATEGY IS BASED ON A TENTATIVE PHYSICS UNDERSTANDING DRAWN FROM PARTIAL EXPERIMENTAL VERIFICATION OF THEORETICAL CALCULATIONS OF IMPLOSION PHYSICS AND BEAM-TARGET INTERACTION

88 PHENOMENA. SINCE PARTIAL VERIFICATION HAS BEEN OBTAINED AT VERY LOW ENERGY LEVELS RELATIVE TO PROJECTED FUSION CONDITIONS, THE FOLLOWING APPROACH HAS BEEN ADOPTED TO ATTAIN A FULL UNDERSTANDING OF THE FUSION PROCESS AS SOON AS POSSIBLE, OPERATE AS SOON AS PRACTICAL IN THE FUSION REGIME (WHERE PELLET GAIN RANGES FROM 0.1 TO 1) IN ORDER TO VERIFY SCALING, TEST CODES, AND IDENTIFY AND RESOLVE UNKNOWN PROBLEMS THAT MAY ARISE AT HIGH ENERGIES. CONDUCT BASIC EXPERIMENTS ON PRESENT-DAY POWER SOURCES TO DEVELOP A COMPLETE IMPLOSION PHYSICS UNDERSTANDING. HIGH TECHNOLOGY, LARGE-SCALE EXPERIMENTS ARE BASED ON THE CAPABILITIES OF THE ERDA LABORATORIES TO MODEL MATHEMATICALLY THE PHYSICAL PROCESSES OF INERTIAL CONFINEMENT FUSION. DEVELOP ALTERNATIVE IGNITION SOURCES SUFFICIENTLY FAR TO SELECT AN EXPERIMENTAL REACTOR DRIVER. PROVIDE ECONOMICAL PELLETS FOR CURRENT FACILITIES AND DEVELOP THE TECHNOLOGY TO PROVIDE PELLETS FOR ADVANCED ENGINEERING FACILITIES WITH HIGH REPETITION RATE REQUIREMENTS. BUILD A TECHNOLOGY BASE FOR THE ENGINEERING DEVELOPMENT PHASE OF INERTIAL CONFINEMENT FUSION BY SUPPORTING FUSION-RELATED RESEARCH IN UNIVERSITIES AND INDUSTRY BOTH INDEPENDENTLY FROM AND COLLECTIVELY WITH THE NATIONAL LABORATORIES. THE PROGRAM, CARRIED ALL THE WAY THROUGH TO CIVILIAN APPLICATIONS, WILL BE FACILITY LIMITED BECAUSE OF SCIENTIFIC REQUIREMENTS AND RESOURCE LIMITATIONS. THE PRESENT RESEARCH TOOLS FOR EXPERIMENTATION--LASERS AND ACCELERATORS--ARE ALSO CANDIDATE POWER SOURCES FOR DEVELOPMENT AS FUSION DRIVERS. HOWEVER, ADVANCED LASER DEVELOPMENT, INCLUDING THE IDENTIFICATION OF NEW LASER MEDIA, MUST PROGRESS RAPIDLY IF IGNITION SOURCE DETERMINATION IS TO KEEP PACE WITH SCIENTIFIC UNDERSTANDING.

89 NARROWING OF CANDIDATE IGNITION SOURCES WILL OCCUR BEFORE THE PROGRAM ENTERS THE ENGINEERING DEVELOPMENT PHASE, KNOWN ENGINEERING PROBLEMS BEYOND SCIENTIFIC FEASIBILITY DEMONSTRATION ARE FORMIDABLE; HENCE, A DEDICATED MATERIALS TEST FACILITY APPEARS TO BE NECESSARY, BUT MIGHT OPERATE IN PARALLEL WITH THE FIRST EXPERIMENTAL POWER REACTOR AND ALSO SERVE TO MEET THE REQUIREMENTS OF OTHER PROGRAMS IN ERDA, THE PRESENT APPROACH IS TO LIMIT PLANNING FOR LARGE FACILITIES TO ONLY THOSE SOURCE CANDIDATES WITH THE STRONGEST POTENTIAL TO LEAD DIRECTLY TO POWER APPLICATIONS IN THE NATIONAL INTEREST. THIS APPROACH IMPLIES A CONSERVATIVE PROJECTION OF NEEDS FOR MAJOR FACILITIES IN ORDER TO RETAIN FLESIBILITY AND AVOID TOO EARLY A COMMITMENT TO LARGE, COSTLY FACILITIES WITH LONG PROCUREMENT LEADTIMES, FOR PLANNING PURPOSES ALTERNATIVE TIME SCHEDULES FOR MAJOR DECISIONS IN THE PROGRAM CAN BE IDENTIFIED AS SHOWN IN FIGURE 5. THE ALTERNATIVES DIFFER MAINLY IN A HIGHER DEGREE OF RISK AND GREATER PEAKS IN INVESTMENT ASSOCIATED WITH PARALLEL DEVELOPMENT OF FACILITIES, AS OPPOSED TO SERIAL DEVELOPMENT IN WHICH THE DESIGN OF EACH FACILITY OCCURS AFTER THE PRECEDING FACILITY HAS REACHED THE MILESTONES REQUIRED FOR FULL SCIENTIFIC JUSTIFICATION OF THE SUCCEEDING INVESTMENT. OVERALL COSTS ARE SUFFICIENTLY TENTATIVE THAT NO COMPELLING ARGUMENT CAN BE MADE FOR A PREFERENCE AMONG THE ALTERNATIVE TIME SCHEDULES ON THE GROUNDS OF OVERALL COSTS. THE GREATER RISKS OF ACCELERATED DEVELOPMENT WOULD BE OFFSET BY EARLIER RESOLUTION OF THE SCIENTIFIC UNCERTAINTIES AT WHICH THE PROGRAM IS NOW DIRECTED. USING COMPUTATION TECHNIQUES IT IS POSSIBLE TO CALCULATE WITH SATISFYING ACCURACY THE CONDITIONS REQUIRED FOR INERTIAL CONFINEMENT FUSION IMPLOSIONS TO OCCUR. THERE REMAIN UNCERTAINTIES ABOUT THE PHYSICAL PROCESSES INVOLVED, AND MUCH EXPERIMENTATION

90 REMAINS TO BE DONE TO VERIFY THE CHARACTERISTICS OF THE IRRADIATION SOURCE AND PELLET REQUIRED FOR ENERGY APPLICATIONS. SCIENTIFIC FEASIBILITY WILL BE DEMONSTRATED BY VERIFICATION ON A LABORATORY SCALE OF A THOROUGH UNDERSTANDING OF THE PHYSICS OF INERTIAL CONFINEMENT FUSION, THE TENTATIVE CRITERIA FOR DEMONSTRATION ARE: O DEMONSTRATE ABILITY TO DIAGNOSE THE KEY PHENOMENA O VERIFY ABILITY TO CALCULATE TARGET PERFORMANCE O DETERMINE TARGET DESIGN CRITERIA O DEMONSTRATE FOCUSING AND COUPLING OF THE INCIDENT BEAM O DEMONSTRATE ABILITY TO MEASURE SOURCE CONVERSION PARAMETERS,

MAJOR UNCERTAINTIES ARE THE PRECISE NATURE OF THE ABSORPTION MECHANISM AND THE DEGREE TO WHICH ABSORPTION IS DEPENDENT ON THE WAVELENGTH OF THE INCIDENT BEAM, THE BURN HISTORY OF THE PELLET, WHICH IN TURN DETERMINES THE DEGREE TO WHICH THE ENERGY PULSE MUST BE TAILORED TO CAUSE MAXIMUM COMPRESSION BEFORE THE FUEL MASS BEGINS TO HEAT, MUST BE CAREFULLY MODELED ON THE BASIS OF REPEATED EXPERIMENTS. THE PRODUCTION AND GROWTH OF INSTABILITIES, WHICH COULD LARGELY DEFEAT THE ABSORPTION PROCESS AT HIGH ENERGIES, CANNOT BE SATISFACTORILY PREDICTED BY THEORETICAL TOOLS NOW AVAILABLE. HENCE, OUR ABILITY TO DESIGN HYDRODYNAMICALLY STABLE PELLETS MUST BE TESTED THROUGH EXPERIMENTATION ON THE LARGE MACHINES NOW PLANNED OR UNDER DEVELOPMENT, THEORETICAL CODES WILL BE NORMALIZED AT SUCCESSIVE ENERGY LEVELS UNTIL THE SCALING CHARACTERISTICS OF THE REACTION PHENOMENA WITH ENERGIES ARE DETERMINED CONCLUSIVELY,

PELLET DESIGN IMPROVEMENTS, BASED ON THE ITERATIVE CYCLE SHOWN IN FIGURE 6 RESULT IN THE RELAXATION OF THE POWER

91 REQUIREMENTS FOR ACHIEVING A GIVEN AMOUNT OF COMPRESSION. EVEN MACHINES OF LOW POWERS CONTINUE TO GIVE USEFUL EXPERIMENTAL RESULTS IN THE STUDY OF BEAM ABSORPTION, X-RAY SPECTRA, AND PULSE DYNAMICS (WAVELENGTH AND PULSE SHAPE) AND ASSIST DIAGNOSTICS DEVELOPMENT. IN GENERAL, A GIVEN EXPERIMENT CAN BE PERFORMED MORE SIMPLY AND CHEAPLY ON A SMALLER MACHINEs LEAVING THE LARGE MACHINES THAT REPRESENT THE STATE-OF-THE-ART FOR KEY EXPERIMENTS. PELLET DESIGNS CAN BE PROVEN AT LOWER POWERS BEFORE BEING USED ON THE LARGEST MACHINES AVAILABLEJ THEREBY RELAXING THE COST AND LABOR CONSTRAINTS ON PRESENT PELLET CAPABILITIES. HENCE, AN IMPORTANT USE OF OPERATING FUNDS IN THE PROGRAM DURING THE RESEARCH PHASE IS TO OPERATE SMALLER MACHINES, SOME OF WHICH ARE LOCATED OUTSIDE THE ERDA LABORATORIES, IN SUPPORT OF THE LARGE-SCALE EXPERIMENTAL PROGRAM ON THE NEWEST AND LARGEST MACHINES, THE PRODUCTION RATE AND COST OF PELLETS BECOME MATTERS OF CONCERN ALREADY IN THE RESEARCH PHASE OF THE PROGRAM. WHILE BARE SHELLS CAN BE PRODUCED IN LARGE QUANTITIES, QUALITY CONTROL, FILLING, AND FINISHING ARE STILL SLOW AND LABOR INTENSIVE. THESE PROCESSES REQUIRE AUTOMATION TO MEET REQUIREMENTS FOR THE EXTENSIVE EXPERIMENTATION CALLED FOR BY THE PROGRAM STRATEGY IN THE NEXT FEW YEARS. AS MORE COMPLEX TARGETS ARE REQUIREDJ AUTOMATION CAN OFFSET BOTH THE COST AND PRODUCTIVITY EFFECTS AND ENSURE AN ADEQUATE SUPPLY OF PELLETS WITHIN THE BUDGETARY CONSTRAINTS OF THE PROGRAM. THE GOAL OF THIS ACTIVITY IS TO PUT PELLET PRODUCTION ON AN INDUSTRIAL BASIS BEFORE THE END OF THE RESEARCH PHASE SO THAT PELLET FABRICATION TECHNOLOGY DOES NOT BECOME A PACING ITEM IN THE PROGRAM. IGNITION SOURCES FOR THE RESEARCH PHASE OF THE PROGRAM ARE NEEDED BOTH TO EXPLORE AT RELATIVELY LOW POWER LEVELS BEAM-PLASMA INTERACTIONS, AND TO BE SCALED TO HIGH ENERGIES IN ORDER TO LEARN

92 THE IMPLOSION PHYSICS PARTICULAR TO EACH KIND OF IGNITION SOURCE. IN ORDER TO DEMONSTRATE AS EARLY AS POSSIBLE THE FEASIBILITY OF LASER OR PARTICLE BEAM DRIVE FUSION IMPLOSIONS, THE PRESENT PROGRAM APPROACH CONTEMPLATES THAT THE MOST ADVANCED AND TRACTABLE OF THESE IGNITION SOURCES SHOULD BE DEVELOPED TO THE POWER LEVEL REQUIRED FOR A LABORATORY-SCALE VERIFICATION OF THEORETICAL PREDICTIONS, THE REMAINING IGNITION SOURCES ARE TO BE DEVELOPED SUFFICIENTLY FAR TO CHARACTERIZE THEM AS FEASIBLE FUSION DRIVERS,

THE LATTER INCLUDE C02 AND OTHER CANDIDATE LASERS FOR DRIVING AN EXPERIMENTAL POWER REACTOR, AS WELL AS ELECTRON BEAM AND ION TECHNOLOGY, TARGET EXPERIMENTS ON NEODYMIUM GLASS LASER SYSTEMS AT LLL, TOGETHER WITH RECENT ADVANCES IN COMPUTER-CONTROLLED BEAM POINTING AND FOCUSING AND IN THE PRODUCTION OF NEW GLASSES THAT WILL PERMIT HIGHER POWER AMPLIFICATION THAN DOES THE SILICATE GLASS USED TO DATE, INDICATE THE FEASIBILITY OF AN UPGRADE OF THE SHIVA LASER SYSTEM AT LLL. A STUDY PERFORMED IN FY 1976 INDICATES THAT THE SYSTEM COULD BE UPGRADED TO 100-200 TERAWATTS BY INCREASING THE 20 BEAMS OF SHIVA I TO 32 BEAMS, THE NEW GLASSES--FLUOROSILICATES, FLUOROPHOSPHATES, AND FLUOROBERYLLATES-- WOULD APPEAR TO APPROXIMATELY DOUBLE THE POWER THROUGH THE SYSTEM. THE COMBINATION OF THESE TECHNICAL ADVANCES IN AN UPGRADED SYSTEM OFFERS THE POSSIBILITY OF PELLET FUSION YIELDS SUFFICIENTLY HIGH TO TEST THE FEASIBILITY OF INERTIAL CONFINEMENT ON A LABORATORY SCALE IN THE EARLY 1980S. RECENT DEVELOPMENTS IN THEORETICAL CALCULATION AND EXPERI- MENTATION INDICATE THAT SOME OF THE MAJOR PROBLEMS THAT APPEARED TO BE INHERENT IN THE LONG WAVELENGTH OF CARBON DIOXIDE GAS MAY BE TRACTABLE OR AVOIDABLE. DEVELOPMENT OF THE TWO-BEAM AND EIGHT-BEAM SYSTEM AT LASL, AND PROTOTYPE DEVELOPMENT AND TESTING

93 FOR THE 100 KILOJOULE HEGLF SYSTEM, WILL PERMIT EXPLORATION OF IMPLOSION PHYSICS AT LONG WAVELENGTHS AND POSSIBLY DEMONSTRATE

THE SUITABILITY OF CO2 AS A LASER MEDIUM FOR AN EXPERIMENTAL POWER REACTOR, AT ITS PROJECTED ENERGY LEVEL--100 KILOJOULES-- HEGLF WOULD APPEAR TO BE BOTH AN INCREASINGLY VIABLE OPTION TO THE SHIVA SYSTEM UPGRADE FOR IMPLOSION PHYSICS RESEARCH, SEVERAL VIABLE CANDIDATE LASER MEDIA HAVE BEEN IDENTIFIED, WHICH HAVE CHARACTERISTICS OF EFFICIENCY AND WAVELENGTH DESIRABLE FOR AN IMPLOSION FUSION DRIVER, ADVANCED LASER TECHNIQUES, INCLUDING NEW PUMPING AND PULSE FORMATION TECHNIQUES, ALSO NEED TO BE DEVELOPED, ALONG WITH THE TECHNOLOGY REQUIRED TO GIVE A FUSION DRIVER A SUFFICIENTLY HIGH REPETITION RATE FOR MATERIALS TESTING AND REACTOR OPERATION. ACCELERATOR DEVELOPMENT AND EXPERIMENTATION ARE REQUIRED TO DETERMINE THE FEASIBILITY OF PARTICLE BEAM-DRIVEN FUSION IMPLOSION WITH THE HIGH EFFICIENCIES THEORETICALLY AVAILABLE FROM ELECTRON AND ION GENERATION. THE RELATIVELY WELL-DEVELOPED STATE OF ACCELERATOR TECHNOLOGY, AND RECENT RESULTS THAT INDICATE THE POSSIBILITY OF GENERATING INTENSE ION BEAMS OR STORING IONS IN A RING SYSTEM, MAKE PARTICLE BEAM TECHNOLOGY A LARGELY INDEPENDENT AND RELATIVELY INEXPENSIVE ALTERNATIVE TO LASER TECHNOLOGY, IN WHICH THE ATTAINMENT OF HIGH ENERGIES AND MACHINE EFFICIENCY REMAIN MAJOR DEVELOPMENT PROBLEMS. IN THE ENGINEERING DEVELOPMENT PHASE OF THE PROGRAM, OVERALL FACILITY COST AND EARLY AVAILABILITY MAY DETERMINE THE POWER SOURCES--AND THE FEASIBILITY--OF DEDICATED FACILITIES ON WHICH TO PURSUE THE MATERIALS RESEARCH AND TESTING THAT WILL BE REQUIRED TO SUPPORT THE CONCEPTUAL DESIGN OF AN OPERATING TEST FACILITY FOR CIVILIAN POWER, A CRITICAL DECISION WILL BE WHETHER ONE OR MORE OF THE HIGH ENERGY LASER OR PARTICLE BEAM MACHINES UNDER

94 DEVELOPMENT FOR THE RESEARCH PHASE OF THE PROGRAM CAN BE RETROFITTED OR UPGRADED FOR THESE PURPOSES. EFFICIENCY WOULD BE UNIMPORTANT FOR EITHER OF THESE USESj AND REPETITION RATE WOULD APPEAR TO BE IMPORTANT ONLY FOR THE MATERIALS TEST FACILITY. FOR AN EXPERIMENTAL POWER REACTOR THE POWER SOURCE WOULD HAVE TO BE SUFFICIENTLY WELL DEVELOPED TO BE CAPABLE OF SUSTAINED OPERATION, WHICH IMPLIES CONSIDERABLE ADVANCE IN LASER OR PARTICLE BEAM ACCELERATOR TECHNOLOGY AS WELL AS A REPETITION RATE AND ENOUGH EFFICIENCY TO ENSURE A NET ENERGY GAIN FROM PELLETS. THE SIZE AND CHARACTER OF HIGH-GAIN PELLETS DEVELOPED IN THE RESEARCH PHASE WILL DETERMINE SOME OF THE CHARACTERISTICS REQUIRED IN THE INPUT POWER SOURCE SYSTEMi WHILE THE RELATIVE COST, RELIABILITYJ AND ENERGY LEVEL OF THE VARIOUS CANDIDATE SOURCES MAY BE SECONDARY CONSIDERATIONS IN THE RESEARCH AND ENGINEERING DEVELOPMENT PHASES OF THE PROGRAM. THE TIME REQUIRED TO DEVELOP A RELIABLE LASER EMPLOYING A NEW MEDIUM AND NEW EXCITATION TECHNIQUES, FOR EXAMPLE, MAY MAKE IT NECESSARY TO DEVELOP A DRIVER SPECIFICALLY FOR ENGINEERING PURPOSES, PARTICULARLY IF A DEDICATED MATERIAL TEST FACILITY IS REQUIRED WELL BEFORE THE FIRST EXPERIMENTAL TEST REACTOR IS DESIGNED, HENCEj THE INPUT POWER SOURCE SELECTION SHOWN IN FIGURE 2 MAY NOT BE A SINGLE CHOICE, BUT RATHER A SERIES OF CHOICES BASED ON PELLET REQUIREMENTS AND THE STATE-OF-THE-ART AT THE TIME WHEN ENGINEERING FACILITIES MUST BE BUILT TO MEET PROGRAM GOALS.

SYSTEM STUDIES ARE THE PRINCIPAL TOOL IN THE RESEARCH PHASE OF THE PROGRAM WITH WHICH TO DEFINE PROBLEMS THAT ARE COMMON TO ERDA REACTOR DESIGN AND DEVELOPMENT FOR SEVERAL ENERGY OPTIONS AND PROBLEMS THAT ARE UNIQUE TO INERTIAL CONFINEMENT FUSION. THE OPTIMAL STRATEGY FOR ERDA WOULD APPEAR TO BE TO MAINTAIN

95 CLOSE COORDINATION OF REACTOR DEVELOPMENT EFFORTS AMONG SEVERAL PROGRAMS, TO EMPLOY EXISTING AND UPGRADED INDUSTRIAL CAPACITIES IN THIS FIELD, AND TO ENSURE SUFFICIENT CONSULTATION WITH UTILITIES, AND WITH OTHER GOVERNMENT AGENCIES CONCERNED WITH LICENSINGs SAFETY, AND ENVIRONMENTAL CONSIDERATIONS, SO THAT THIS AREA OF THE PROGRAM EFFORT IS DIRECTED EXPLICITLY TOWARD OPERATING REQUIREMENTS FOR COMMERCIALIZATION. THE INERTIAL CONFINEMENT FUSION PROGRAM IS A RESEARCH EFFORT REQUIRING A BROAD BASE OF RESEARCH IDEAS, SYSTEM STUDIES, BASIC EXPERIMENTS AND CALCULATIONS, AND DESIGN AND FABRICATION OF DIAGNOSTIC AND RESEARCH TOOLS. PROGRAM FUNDS ARE ALLOTTED TO BROAD AREAS WHERE QUESTIONS OR PROBLEMS HAVE BEEN IDENTIFIED FOR RESOLUTION, THIS ALLOTMENT MUST BE SUFFICIENTLY FLEXIBLE TO PERMIT SOME DUPLICATION OF EFFORT WHILE AVOIDING WASTE OF LIMITED RESOURCES. BECAUSE NO CRITICAL PATH HAS BEEN IDENTIFIED TO ATTAIN THE PROGRAM OBJECTIVE, AND UNCERTAINTIES REMAIN ABOUT THE DEGREE TO WHICH PERCEIVED PROBLEMS CAN BE OVERCOME OR AVOIDED WITHOUT SIGNIFICANT REDIRECTION OF PROGRAM EFFORTS, THE LINKAGE BETWEEN PROGRAM GOALS AND NEAR-TERM RESEARCH TASKS RESTS ON A TENTATIVE UNDERSTANDING OF THE REQUIREMENTS TO ATTAIN FUSION CONDITIONS THROUGH PELLET IMPLOSIONS THAT IS SUBJECT TO CONTINUING REVIEW AND EVALUATION. A LARGE SHARE OF TOTAL PROGRAM FUNDS IS COMMITTED TO THE CONSTRUCTION AND OPERATION OF LARGE RESEARCH FACILITIES. THE REMAINDER IS DIRECTED TOWARD THE SUPPORT OF RESEARCH EFFORTS IN UNIVERSITIES AND INDUSTRIAL FIRMS WHENEVER THE UNIQUE CAPABILITIES OF THE ERDA LABORATORIES ARE NOT REQUIRED FOR SPECIFIC TASKS AND A SUITABLE CONTRACTOR CAN BE FOUND. TO FOSTER AND EFFICIENTLY USE A RESEARCH BASE IN UNIVERSITIES AND INDUSTRY REQUIRES MORE STABLE FUNDING--REASONABLY PREDICTABLE AMOUNTS OVER TWO OR MORE YEARS--THAN IS POSSIBLE IF

96 THIS AREA OF THE PROGRAM IS TREATED AS ENTIRELY ANCILLARY TO THE CORE PROGRAM. HENCEJ FORWARD PLANNING FOR RESEARCH IN THE CORE PROGRAM IS REQUIRED, BOTH TO RELATE CURRENT ACTIVITIES TO THE PROGRAM GOALS AND TO PROVIDE A FRAMEWORK EVALUATING AND SELECTING OUTSIDE RESEARCH PROPOSALS FOR SUPPORT, THE VERY LARGE RESEARCH FACILITIES DEDICATED TO THE INERTIAL CONFINEMENT FUSION PROGRAM ARE LOCATED AT THREE OF THE ERDA LABORATORIES, THE LABORATORIES POSSESS THE THEORETICAL. AND DATA MANAGEMENT AND COMPUTATION ABILITIES REQUIRED TO MAKE THE CALCULATIONS BY WHICH RESEARCH TASKS ARE DEFINED. THE LABORATORIES ALSO POSSESS STABLE RESEARCH TEAMS AND EXPERIENCE IN THE MANAGEMENT OF LARGE-SCALE EXPERIMENTS, WITH ADEQUATE PROVISION FOR THE EXCHANGE OF RESEARCH IDEAS AND RESULTS WITH

OUTSIDE RESEARCH ESTABLISHMENTS, AND FOR TECHNOLOGY TRANSFER TO INDUSTRY, THE ERDA LABORATORIES CAN SERVE AS THE LOCUS OF A CORE PROGRAM DIRECTED TOWARD THE DEMONSTRATION OF SCIENTIFIC FEASIBILITY, IT IS THE POLICY OF ERDA TO LIMIT ENERGY-RELATED RESEARCH IN THESE CORE LABORATORIES TO THE LARGE-SCALE EXPERIMENTATION AND MATHEMATICAL MODELING OF PHYSICAL PROCESSES THAT CANNOT BE PERFORMED OUTSIDE THOSE LABORATORIES. BASIC RESEARCH IN LASER-MATTER INTERACTIONS, AND DEVELOPMENT OF BOTH IGNITION SOURCES AND DIAGNOSTICS, IS PERFORMED TO THE EXTENT PRACTICAL OUTSIDE THE ERDA LABORATORIES, AT OTHER ERDA AND OTHER GOVERNMENT LABORATORIES, IN UNIVERSITIES, AND IN INDUSTRY,

MUCH OF THE TECHNOLOGY REQUIRED TO MEET PROGRAM GOALS IN THE RESEARCH PHASE AND LATER IS COMMON TO LARGE TARGET IRRADIATION SYSTEMS, PELLET FABRICATION REQUIREMENTS; AND REACTOR SYSTEMS WITHOUT REGARD TO THE SPECIFIC REQUIREMENTS OF EACH IGNITION

97 SOURCE, THE LONG-TERM PLAN IS PREDICATED ON THE PARTICIPATION OF INDUSTRY IN SEVERAL AREAS OF DEVELOPMENT: O LASER AND ACCELERATOR DESIGN AND CONSTRUCTION O AUTOMATED, HIGH-VOLUME FABRICATION OF PELLETS FOR HIGH REPETITION RATE IGNITION SYSTEMS O RAPID PELLET HANDLING SYSTEMS FOR REPETITIVE OPERATION OF EXPERIMENTAL, AND LATER DEMONSTRATION, REACTORS O POWER SUPPLIES AND POWER CONDITIONING TECHNOLOGY REQUIRED IN COMMON WITH OTHER ENERGY SYSTEMS.

THE FIRST TWO OF THESE AREAS REQUIRE EARLY ATTENTION EVEN IN THE RESEARCH PHASE OF THE PROGRAM, THE REMAINING ONES, AND ADDITIONAL AREAS OF INDUSTRIAL PARTICIPATION EXPECTED TO BE IDENTIFIED AS THE PROGRAM EMPHASIS SHIFTS TO TECHNOLOGY AND ENGINEERING DEVELOPMENT, SHOULD BE ADDRESSED THROUGH SYSTEM STUDIES, PERFORMED BY OR IN COOPERATION WITH INDUSTRY, WELL BEFORE THE MID-1980S. .PROGRAM ACTIVITIES IN PHASE I OF THE PROGRAM (FIGURE 1) CAN BE BROADLY CHARACTERIZED AS BASIC AND APPLIED RESEARCH AND TECHNOLOGY DEVELOPMENT (FIGURE 7). HOWEVER, THE PHASES SHOWN IN FIGURE 1 ARE ILLUSTRATIVE OF PROGRAM EMPHASIS RATHER THAN PRESCRIPTIVE. WELL BEFORE THE RESEARCH PHASE IS COMPLETED BY THE DEMONSTRATION OF SCIENTIFIC FEASIBILITY, SOME PROGRAM ACTIVITIES SHOULD HAVE CONTRIBUTED TO DEFINING THE TASKS IN THE ENGINEERING DEVELOPMENT PHASE OF THE PROGRAM, PELLET FABRICATION AND IGNITION SOURCE DEVELOPMENT FOR LARGE RESEARCH FACILITIES, AND REACTOR SYSTEM STUDIESj SHOULD LEAD DIRECTLY TOWARD THE CONCEPTUAL DESIGN OF AN EXPERIMENTAL POWER REACTOR. HENCE, THE TRANSITION FROM TECHNOLOGY TO ENGINEERING DEVELOPMENT IN THIS PROGRAM IS NOT EXPECTED TO BE SHARPLY DEFINED.

98 THE LONG-TERM PROGRAM PLAN SERVES AS THE FRAMEWORK FOR ACTIVITIES DIRECTED TOWARD THE ATTAINMENT OF PROGRAM GOALS. THE ANALYSIS FOR THE PROGRAM PLAN IS PERFORMED BY THE PROGRAM DIVISION IN CONSULTATION WITH THE MAJOR PROGRAM PARTICIPANTS. THE PROGRAM PLAN DEFINES THE OBJECTIVES OF THE PROGRAM AND THE' STRATEGY TO ATTAIN THE OBJECTIVES BY ROLL-BACK PLANNING--I.E., THE PROGRAM GOALS, DOWN TO THE NEAR TERM, ARE DEFINED BY THE OBSTACLES THAT NEED TO BE REMOVED AND THE UNCERTAINTIES TO BE RESOLVED IN ORDER TO DEMONSTRATE THE SCIENTIFIC FEASIBILITY OF INERTIAL CONFINEMENT FUSION AND APPLY THE OPTIMUM TECHNIQUES DEVELOPED TO ENERGY REQUIREMENTS, THIS APPROACH IS ILLUSTRATED IN FEGURE 2 IN WHICH THE LOGICAL DEVELOPMENT OF PROGRAM GOALS FLOWS IN THE OPPOSITE DIRECTION TO THE DEVELOPMENT OF A FULL IMPLOSION PHYSICS UNDERSTANDING AND THE ESTABLISHMENT OF DECISION CRITERIA TO PROCEED TO THE LATER STAGES OF IMPLEMENTATION-- EXPERIMENTAL, AND EVENTUALLY DEMONSTRATION, POWER REACTORS, THE NEAR TERM GOALS, TO ACHIEVE SCIENTIFIC FEASIBILITY DEMONSTRATION, ARE ESTABLISHED BY CONSULTANTS AND A STANDING ADVISORY GROUP AS REQUIRED, AND BY THE INJECTION OF NEW IDEAS FROM A VARIETY OF SOURCES--RESEARCH IDEAS AND PROPOSALS, CONTACTS WITH PARTICIPANTS IN RELATED PROGRAMS ELSEWHERE IN ERDA AND IN OTHER GOVERNMENT AGENCIES, AND CONTACTS WITH FOREIGN RESEARCHERS IN LASER AND PARTICLE BEAM FUSION AND PLASMA PHYSICS GENERALLY,

99 27 WL C) U

- i--2:0I - 2

5- I--0 oC Q: ' a L I-- < 2- _JS: LU 1C..) >O?Q LUD WUtLU

LLI Q 0 _ - l- r- L FHO : -4 l-t(_ CD =3 L O c LU - U $- O C-,-

S3: LU

LU5-Li-CJ LUz HwL

_j - CD ClLUJ CQ Cn r- u o

>- - I WLU QZ Cn0 >- (1LU-I 0C CO GO , , ,

-_ CoZ U Z' -O 2:: z C)0 U 2 o LtLU V)LU FH H

LLC- ag-L % u

LLH i uci C'C

Lt o ._ -1

- _ o ._0 LL. UrL..) F -i ,-110 i? V)tZ UJZ CYQ CC l-0 4_-C 2<3 i 2: C)Q 3. _ J Z H

i__F W L O CD *

H H '-X'-E Z C OU

Ltv _- L ) S -LUt- 2: LU 0- cl:LU <~~~~~~~~~~~~~aU U)C> C) 4

0(J- 0. Q I- UJ

100 c;C: t-

I)Jc1(J S

5g3

C2.

O_ U I- g t£

VG C; UJ · e ·

......

:u 1< cc 5 ..

. . .* 0: .

WCC cL ( C -

<& [r: C) ~- ^-5 -, 50 5 20 0 6;00* £ p." W~Yli<· i 0 1- c tM .~ 2 0 0 N $ .B.

m§ 'v 3 ft

c- W

2C cc

t 0 0S - 3 S a cLX Q. 79 UJ S; .

2 0,

c c

3 .--..1C[ O*I- o

* * e

101 0 g <1

LU-%/*~ ~ ~ ~ ~~ ~~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ rL L< - Cr- 0I~C I -- I °'~~~~~~~~~~~~~~ I // / /~~~~ ~'i~~~-u/*>Q/ /f uA/- 2W~~~ /:./,5gi ~,///

0 ~ 3. C~~~~~) 0~£ - 1 , 0 o S~ K2I- L, --i >'~^LU ~ . 1 <~ l:'' c ,'' / *// //a /,~' ',~0I 0 0 1 ~ ~ ~ ~ ~ ~ . 0r5"oU 14 0 0,~~ ,_ ?~ /.,;@~~~~~~~~~~~~~~~~I

Lu~' L. j '? ( C Z : a£. .l.~.. ¥ A. ,. .' V1 / § §.4 o I£ <"X |g^ -o,-I '< 4'/-~

~~oI- LL.*.^i2£i. coto S ^ ------"V „ ______,____ r-/ 12? p 'L'jl I' ] I i , ~~~~~~~~~~~~~~~~~~~~~~~~~~illI I ~]' ' J Ju .... I > § / //^% ^' I ^~~~~~~~~~~~~~L 4u t1~~

~~8 ', C) '//. /~~~~~~~~~~~~~~~~~~~~~~~~~~~ Wa~~

~~z a- ^^ ^B ~~~~~~~~~~~~~~~~~~~~~~~~~~I- §w- -_j^ ! ! v 'i~~

~~y. 4>' o fs §"{>= s I S? - ^ 5 !§ g S BS | ^ ?! ^ .£LU~~

~~5 s 5 I I t> y ^ s s § ^ - ^ ^r I"~~

~~4 I,-i/i -i uj uo- ~ u uj"r< i- W~o- - Sti- .^.E ~~~o 5 ; a 1.^4 cIK -j|i cccc £ - ^ § K0 §§ "ittaC § ^ ~~~~~~~~~~ -c 2: 0 _j c:u 0 -j . 0 o. AM c.. < w v) Li U- 44 -~~

~~102 0 .l Cd 0) -4 ,toc r-4 *,r aP )> 03 cd J) t~o ,4 *rO bO 0) c) * CO a) .- 03 -Put -p 0) a) 0a~~

~~o 10 00 ^1 ot0 .i~~

~~0 bO P $-4h 0o *rl UT uC 0) $0 1 P Cl)'r~cli a-03 0 r Cl >P C(L - W 0 q- n~~

~~Ir IJ III I II.0~~

~~I- c a CY aO 4i H u H 0~~

~~1-1 cr r- _+ Pq . r o g 0 t-i tnC CL g 0 ,-i P4 &4 0~~

~~PL p2 4J- 0 I-{ CO g 00 00 00 v- v000 0C-0 roo P £nU I0 10 -10 r- o - r o a00 - - v TA i-1 r-i~~

~~k& 01~~

~~0 0~~

-l CM4 0 00000 0 O r-i n n V1 M U · . . * *

~~a) rl 0 -V Pm m 0)0 03t pr U O 03 0) > 40 0CO 0 cd (4 CO o PI 00 &.CM ..O CO 3 -IP: Z R 0e Z p4g 0 h-1~~

~~103 (0~~

~~C) N~~

~~N 0 )~~

~~0 0 0 N1~~

~~C) W3.0 ^ 0 0)~~

~~z 0 -: 1-1~~

~~CO£0~~

~~U . I-~~

~~LLi N C, 0 -Jul~~

~~0 LLi p»CO~~

~~Cc C,, >- >- F-, Ww W - Q Lt Cl- LU h--cn I- LJ Q~~

~~LU 0. I- _ < W cci-i w rD o:> ^ -1 (3 UL vW 0 U Ur~~

~~104 0)~~

~~C1)~~

~~x l r ct~~

~~a.- a c, ^ r c: ^ 4--- 0 W0~~

~~oa~~~~a C UJ30 ^- *^ . U <~~~~y . 0. v 0.S~~

~~I \~~

~~H - I C~~

*C)i 0 -*" < 0

~~105 3: C)~~

~~U)' 0. :: ft 0 0 CD.)~~

~~CO _1 I- ; LUW W - /~~

~~,C) -4' W I- L 1, "% . ,.~~

~~,' * - /~~

~~0 t.t.j .44i~~

~~C.)~~

~~ULLU1-Li CL - >U~~

~~-O ** Lli E 1~~

~~o3 C u - L - - 00 0 0 U C,,O 106' C,,.- *0 ~O~~

~~_. C2 U 4-4*-~~

~~< 0 U C n C - X - (X10-. 6 0 J I*o- I -0 ,o0. 4Q> ' CLL -r 4J C) 0~~

~~c v- uJ) ,D'--0 u-~~

~~_ C4~~

~~<106 I * RECENT LASER-DRIVEN-IMPLOSION MEASUREMENTS AT KMS FUSION* F. J. Mayer~~

~~KMS FUSION, INC. Ann Arbor, Michigan 48106~~

~~Abstract~~

For efficient thermonuclear burn of laser-driven implosions of deuterium-tritium (DI) many physical phenomena must be understood. Experiments are described which give information on laser-target coupling (absorptivity), electron and i-ray preheat, hydrodynamic stability, and hydrodynamic energy transfer to the fuel.

~~INTRODUCTION~~

~~To understand laser-driven implosions, of deuterium-tritium (DT)~~

~~at densities and temperatures required for efficient thermonuclear~~

~~burn, many physical phenomena must be experimentally investigated.~~

Among the more important are laser-target coupling (absorptivity), electron and X-ray preheat, hydrodynamic stability, and hydrodynamic energy transfer to the fuel. This report describes experiments pertaining to target absorptivity scaling with Z (atomic number), laser intensity, and laser pulse length. Experiments using deliberately introduced asymmetries, both spatial and temporal, bear on the issues of preheat and implosion

~~symmetry. Measurements of DT alphas and DD protons from the compressed target core indicate shell preheat and hydrodynamic energy transfer efficiency.~~

~~'Finally, a new target interaction diagnostic technique makes use of large-~~

~~diameter plastic bubbles to collect activated material and to measure~~

* This research was supported by the United States Energy Research and Development Administration under Contract E(ll-l)-2709.

~~107 target-absorbed energy. The KMSF laser system using the newly developed~~

~~"plasma spatial filter" is also briefly described.~~

~~THE LASER~~

~~The KMSF Nd:glass laser system is shown in Figure 1. The laser pulse~~

~~originates in a mode-locked Nd:YAG oscillator wherein individual 40 psec~~

~~pulses are temporally assembled by the KMSF-designed pulse stacker (l) (2) Figure 2 shows a streak-camera recording ) of four 40-psec pulses separated~~

~~by 60-psec intervals. The laser system is a combined rod and disc ampli-~~

~~fier system, the last eighteen modules being high-power disc amplifiers~~

~~having a 10-cm clear aperture. A polarizer splits the beam into two com-~~

~~ponents after the first six high-power amplifiers (indicated in the figure~~

~~by boxes) and the two beams are further amplified by a circulating-double- (3) pass technique wherein each beam is split by a polarizer and, using~~

~~Faraday rotators, is allowed to circulate in opposite directions through the~~

~~final six amplifiers and is then recombined at the same polarizer back~~

~~into one beam.-- In this configuration, the laser beam emerges with mixed~~

~~polarization.~~

~~In the rod amplifier system (before the 64-mm-rod stage) a new device called a "plasma spatial filter" (PSF) has been incorporated().~~

The PSF focuses (f/3.6 lens) the laser beam onto a planar CH2 target, and uses the plasma reflected energy to form the main laser pulse. It has been found that this device produces a less coherent laser pulse and a

~~considerably smoother laser beam time-integrated spatial distribution.~~

~~Figure 3 shows comparisons of the time-averaged beam spatial profile with~~

~~and without the PSF. NotiCe the average high-intensity "hot spots" have~~

~~been strongly reduced with -the PSF.~~

~~108 TARGETS~~

~~The targets used in these experiments were mostly spherical glass- shell targets ( filled with DT gas. Targets ranged in diameter from~~

30 to 200 micrometers and were filled with from 10 to 100 atmospheres of DT. Wall thicknesses ranged from 0.4 to 2.0 micrometers. For testing the scaling of absorptivity with Z (average -atomic number), glass shells coated with lithium, boron, polyvinyl alcohol, and lithium borohydride were used. Plastic spherical shells of polymethyl-methacrylate (PMMA) and polyvinyl alcohol (PVA) were also used.

~~ILLUMINATION~~

~~The spherical-shell targets were illuminated by the KMSF-designed~~

~~lens-ellipsoidal-mirror combination(6 ) that provides near-normal incidence~~

~~illumination over 70% of the target surface (see insert in Figure 4).~~

~~Normal-incidence illumination is necessary to minimize refractive losses~~

~~and so increase the amount of absorbed laser energy.~~

~~DIAGNOSTICS~~

~~Figure 4 shows the distribution of the various diagnostic instruments~~

~~located in the direct-view diagnostic region (DVDR). Most of the diag-~~

~~nostics have been previously described((7). New instrumentation includes~~

~~an infrared radiometer (#8 in Figure 4) and an alpha-proton detector~~

~~(#7 in Figure 4) which are described below. Absorption measurements~~

~~are made using secondary-electron-corrected charge collectors for ion~~

~~expansion energy and thermoluminescent dosimeters ) for X-ray energy.~~

~~Neutron total-yield measurements have been made using activation (9) counters( and plastic scintillation detectors. Time-of-flight~~

~~measurements with the plastic scintillation detectors have identified~~

~~the neutrons as 14.1 MeV neutrons from the D-T reaction.~~

~~109 IMPLOSION MEASUREMENTS~~

~~The DT-gas compression is inferred from X-ray pinhole photographs~~

(Figure 5). Multiple pinholes are covered with different thicknesses of beryllium foil. The foil filtration technique( ) is used to estimate the electron temperature from the spatially resolved regions of the target implosion.

The X-ray photograph is a time-integrated record of the X-rays emitted by a target as a consequence of the laser irradiation, from which the history of the implosion can be inferred, showing where heating and compression were greatest . Two regions are commonly observed for which X-ray emission is strong. The first is near the initial location of the glass shell, where rapid absorption of energy from the laser results in a hot, dense layer approximately at the original diameter of the pellet.

As the implosion progresses to the kinetic phase, a portion of the original shell moves rapidly inward acting as a pusher, entrapping the fuel, until it is stopped by the back pressure of the highly compressed and heated

~~DT gas. The second region where X-ray emission is strong occurs when the pusher is stopped and its directed kinetic energy becomes thermal energy.~~

During the kinetic phase, as a result of lower temperature in the pusher and the relatively short dwell time at a given radius, emission is weaker, and there is a decrease in film density on the X-ray photographs. Two circular regions of higher film density are observed (Figure 5), one representing the original pellet diameter and the other the final diameter of the compressed pellet (or an upper limit to the final diameter, depending on the resolution of the X-ray camera). The measurement ofithe diameter of this stilT-intact pusher, which acts as a tamper at peak compression, compared to the initial shell diameter, permits calculation of the volume reduction and hence determines the compression of the DT fuel.

110 The amount of X-ray emission observed in the central region of the pinhole photographs depends on the fuel and tamper constituents. The radiative process for the DT fuel is free-free bremsstrahlung, while for the higher-Z tamper radiation will also be produced by free-bound and bound-bound transitions. For glass-shell targets, in which the average atomic number of the shell is 10 times that of the fuel, the tamper produces more X-ray emission than the fuel.

~~EXPERIMENTS ON TARGET ABSORPTIVITY~~

~~Laser light is absorbed classically in a fully ionized plasma by the~~

~~inverse bremsstrahlung process. The ratio of the absorbed to incident flux~~

~~is given by~~

~~gabs/Qo = - e~~

~~with A defined by~~

~~A = Kgd KB L (2)~~

~~wherein L is the electron density scale length and KB the absorption~~

~~coefficient. For inverse bremsstrahlung,~~

~~KB X Z/k2 3/2 (3)~~

~~where Z is the ionic charge, 0 the electron temperature, and X the laser wavelength. For A much less than unity (which corresponds to the~~

~~experimental observations) e, the absorbed energy fraction, is given by~~

~~E -= abs/o A = KBabs&O~~~~~~~~~~~~~~~4 L % ZL/X2 3/2 (~~

~~111 The energy flux F transported by the electrons cannot exceed~~

~~F = n C e 3/2 (5)~~

~~where ne, Ce are the electron density and electron thermal speed, respectively. Energy conservation requires that~~

~~F = bs (6)~~

~~Thus~~

~~ZL 2 63/~~

~~ea (ZL o/A2)1/3 (7)~~

~~Consequently,~~

~~£ = abs/0o " (ZL/X20)) (8)~~

From Equation (8) it is seen that the fractional energy absorption scales as (Z/fo) 2 when the electron density scale length, L, is constant in time. If the electron density scale length is assumed to vary on the

~~2 hydrodynamic time scale, i.e., L = C t e 0 t then the following scaling e results,~~

~~[ -zt ]2/5(~~

~~2T o~~

~~F= Oabs (Zt) 1/5 (10) ~0 j-iA---- r I1~~

~~112 Experiments were performed with various low-Z-shell targets, in particular~~

~~PMMA and PVA plastic shells, at incident laser power ranging from 0.2~~

~~to 0.5 terawatts and pulse lengths from 30 to 200 psec. For typical~~

~~60-pm-diameter glass-shell targets ({ = 10) fractional absorptivity was~~

~~^ 17%; plastic-shell targets (Z z 3.5) showed fractional absorptivity of X 12%, somewhat closer to the Z½ scaling of Equation (8).~~

~~Another series of experiments was performed on glass-shell targets~~

~~over a range of different laser intensities. Figure 6 shows the frac-~~

~~tional energy absorption versus on-target intensity along with the best power-law fit (exponent -1) close to that of Equation (8).~~

~~Figure 7 shows the results of another 'series of experiments on~~

~~glass targets,where the pulse length was varied at constant laser intensity.~~

~~Absorptivity was found to be nearly independent of pulse length, also in agreement with Equation (8). Plotted in Figure 7 are also the time-depen-~~

~~3 / 5 dent scale.length case (e . t ) and an essentially similar hydrodynamic~~

~~computer simulation (labelled TRHYD).~~

All of the above results are consistent with the assumption of inverse-bremsstrahlung absorption with an electron density scale length not varying with time on the scale of a few hundred picoseconds. This result may be consistent with self-steepened density profiles(l2)

~~IMPLOSION SYMMETRY EXPERIMENTS~~

~~A series of experiments was performed in which the laser pulse in.~~

~~one channel of the two-channel illumination system was deliberately~~

~~delayed 58 psec by inserting glass slabs (35 mm thick) into the channel.~~

~~The delays resulted in characteristic off-center target implosions as~~

~~seen in Figure 8. In these experiments, glass-shell targets of nominally~~

~~144-micrometer diameter and 1.2-micrometer wall thickness were imploded~~

~~with approximately 70 joules and a pulse length of 240 picoseconds.~~

~~113 The off-center implosions produced a reduction in neutron yield~~

~~(the yields were X 2 x 10 ) compared with the experiments with no delay,~~

~~as would be expected. From such time-delayed experiments, assuming that~~

~~the shell-wall acceleration time plus the introduced delay time is short~~

~~compared with the collapse time, it is possible to deduce the average wall~~

~~implosion velocity. Implosion veloLcites, so determined, varied from about~~

3 x 107 cm/sec to 6 x 107 cm/sec. These experimentally observed implosion velocities are larger than the implosion velocities obtained from hydrodynamic computer code simulations of purely ablatively-driven shells when the simulations are adjusted so that the postulated absorbed laser'energy is equal to the experimentally measured absorption. Computer-code calculations show increased wall velocities when some form of shell preheat is (13) added ) (either energetic electrons or X-radiation); hence it is believed that some substantial shell preheating occurs in these implosion experiments.

~~Sensitivity of the implosion symmetry to illumination symmetry was studied in another series of experiments using 100-micrometer-diameter,~~

~~1-micrometer-wall-thickness glass-shell targets. The illumination pattern on the pellet surface from the ellipsoidal mirrors when the paraxial rays~~

~~coincide with the pellet center covers a solid angle subtending 144° about the optical axis. The marginal rays (at 720 with respect to the optical~~

~~axis) are approximately 40% more intense than the paraxial rays. Between~~

~~70° and 90° from the optical axis, the pellet receives no laser flux; hence displacements of the paraxial focal positions from the target center~~

~~are required to fill the unilluminated equatorial band. In Figure 9, X-ray~~

~~pinhole photographs of implosion symmetry illustrate the effect of the paraxial~~

~~focal shift. Target shot .2442 had no introduced paraxial focal shift, i.e., both beams were focused at the target cneter, Note the prolate-spheroidal~~

~~shape of the imploded target center, indicating less equatorial heating than polar heating. For target shot 2460, the lens paraxial focal positions~~

~~114 were moved toward each other approximately 10 micrometers, measured from the~~

~~target center. Notice that the central implosion structure is more~~

~~spherical, indicating more balance between equatorial and polar heating.~~

~~For target shot 2456, the lens paraxial focal positions were moved~~

~~approximately 20 micrometers and an oblate spheroidal implosion resulted~~

~~indicating more equatorial heating than polar heating. As would be~~

~~expected, the experiment with the most symmetric central implosion~~

~~structure produced the largest neutron yield (7 x 10 ).~~

~~MEASUREMENTS OF ALPHA PARTICLES AND PROTONS~~

~~Figure 10 shows a schematic of the alpha-particle-and-proton detector(4)~~

~~which makes use of magnetic deflection 5) and time-of-flight particle~~

~~identification. The detector uses a thin plastic scintillator foil and~~

~~a fast photomultiplier tube 154 cm from the target. Figure 11 shows some~~

~~typical oscilloscope traces from this detector displaying a scattered-X-ray~~

~~pulse (far left - used as a timing fiducial), a proton pulse (center) and~~

~~an alpha-particle pulse (right), followed by some energetic "fast" ion~~

~~signals (far right). The measurements made from such traces give the proton~~

~~and alpha particle yields, the mean proton and alpha energies, and the alpha-~~

~~particle energy spread. Table I is a summary of some recent experiments~~

~~on various target types. Indicated in the table are fill pressures of~~

~~deuterium and tritium, laser energy, pulse length, mean proton energy E ,~~

~~mean alpha energy E , alpha energy spread AE , number of alphas detected N ,~~

~~and the number of protons detected Np (detector solid angle = 8.4 x 10- 5 sr).~~

~~The alpha particles lose energy in traversing the fuel and the~~

~~compressed glass tamper. The fractional residual alpha energy has been~~

~~calculated for various values of the tamper fpTdr (density-radius product)~~

~~and tamper electron temperature. The contribution to the alpha energy~~

~~loss from DT fuel is estimated to be small compared to the loss in the~~

115 tamper. Figure 12 shows the results of these calculations. Measurements of the tamper electron temperature made by foil filtration result in electron temperatures of between 1 and 3 keV. From Table I it is seen that'fractional residual energies are typically X 3% to 10%, the indicated

-3 -3 2 IPTdr is roughly between 2 x 10- and 5 x 10 3 g/cm . Hydrodynamic computer- simulation calculations have been performed which indicate that the higher tamper electron temperatures (% 2 to 3 keV) and relatively low values of fPTdr may be due to either hydrodynamic "burn through", i.e., strong electron conduction preheating, or some other shell preheat mechanism, either energetic electrons or X-radiation.

The alpha energy spread is due, in part, to ion thermal broadening. Additional broadening of the alpha spectrum is due to time and space-dependent fpTdr and electron temperature, doppler shift, range straggling, etc. However, an upper bound to the fuel ion temperature is found by assuming all the broadening is ion thermal broadening. The alpha spectrum is broadened (l6 ) as

~~AE (keV) = 177 0Q~~

~~where 0. is the fuel ion temperature in keV. For example, shot D of~~

~~Table I has AE = 430 keV, giving an upper limit ion temperature of~~

~~0i = 5.9 keVT~~

~~The ratio of yield of DT alpha particles to the yield of DD protons is a direct measure of the space-and-time-averaged fuel ion temperature~~

~~(assuming a maxwellian ion distribution obtains in the core). The yield ratio is given by,~~

~~N nT DT Np nD DD N 2D - cv>DD~~

~~116 where the nD, nT are the deuterium and tritium fuel densities and DT,~~

DD are the maxwellian-averaged reaction rates for the DT reaction and the proton branch of the DD reaction respectively. Figure 13 shows a plot of this ratio for a 6:4 deuterium-to-tritium ratio. As seen in Table I, the experimentally observed ratio is rather small. To increase the statis- tical sample, a series of similar experiments using 50-to-60-micrometer diameter, 0.7-to-0.8-micrometer-wall-thickness targets were assumed identical. For these experiments a total of 1420 alphas and 10 protons were recorded, giving a ratio of 142 and an ion temperature from

~~Figure 13 of 6. = 2.6 keV. Poor proton statistics however make the +2.6 error large 09 keV.~~

~~PLASTIC BUBBLE DIAGNOSTIC~~

A new laser fusion diagnostic technique (1 7) has recently been developed which makes use of large ( 4 cm) diameter, thin wall ( 10 micrometers) polyvinyl acetate plastic bubbles to capture neutron-activated tamper material. Figure 14 is a photograph of the large plastic bubble,

*and Figure 15 shows the energy-flow characteristics of the plastic bubble.

Experiments have shown that less than 0.5% of the laser energy is deposited in the bubble wall during irradiation of the laser target mounted at the center of the bubble. Both soft X-radiation, Ufast" corona- produced ions, and slower tamper ions are effectively absorbed in the plastic-bubble wall, as is seen in the comparisons in Figure 16. The somewhat more energetic X-radiation,above the cut-off of 2 mils of beryllium foil,is attenuated only a factor of about 3, so that X-ray pinhole photographs of the target implosion may be taken through the bubble wall.

~~117 The nuclear reaction of interest is the neutron activation reaction,~~

~~n + Si 2 8 1Al + p 1414 13~~

~~tw = 2.3 min.~~

~~4S i 2 8~~

~~0- (2.85 MeV) Y (1.78 MeV)~~

28 The Al produced is trapped in the bubble wall which is then subjected to radioactive analysis in a gamma-ray spectrometer. The number of Al nuclei produced during the neutron generation period is given by,

~~00 00 N Jf dt P drVt) drdt o0 rf~~

~~NA Nn nsi o(l4 MeV)to~~

~~density, nsi where Nn is the total neutron yield, pT is the tamper mass is the silicon number density, a (14 MeV) is the (n,p) reaction cross-~~

~~t is the effective tamper thickness. Substituting section for Si , and 0 the various quantities, it is found that~~

~~NA . 4 x 10- N fTd~~

In present experiments, the neutron yield is on the order of 10 and the JPTdr estimated above from the alpha particle energy loss is about -3 2 3 x 10- g/cm , giving'the number of activated nuclei as

~~NA 120~~

~~118 which,with relatively low y-ray counting efficiency,has poor statistics.~~

~~This experiment has not yet been completed.~~

~~Of possibly more immediate value is the diagnostic use of the plastic~~

~~bubble as an absorbed energy calorimeter . Since effectively all~~

~~of the target absorbed energy is trapped by the bubble wall, a measurement~~

~~of the wall temperature rise allows estimation of the absorbed energy.~~

~~A noncontact method using a far-infrared radiometer (#8 in Figure 4) has been developed which measures the temperature increase. Figure 17 shows~~

~~the radiometer output signal from a target implosion inside the plastic~~

~~bubble.~~

~~The energy delivered to the bubble increases its temperature by an~~

~~amount E o AT = T - T 0 mBv~~

~~where Eo is the deposited energy and mB and Cv are the bubble mass~~

~~and specific heat. The thermal radiative cooling is much faster than~~

~~conduction cooling, so that the temperature decay is given by~~

~~dT = eff (T 4 4 -- (T -T) dt mBC -~~

~~where To is the ambient temperature, a is the Stephan-Boltzmann constant,~~

~~A is the bubble area and £ef f is the effective bubble emissivity.~~

~~The solution to the above cooling equation for small initial tem-~~

~~perature increase AT(O) is,~~

~~AT(t) = AT(O) ( +-3/2 AT(O) et/T_ 3/2 AT(O) T T o o with the time constant~~

~~CvmB~~

~~heffAT0~~

~~119 Figure 18 shows an internally-heated blackbody calibration source whose temperature is measured by thermocouples and is used to calibrate the infrared radiometer.~~

~~For shot 2711 of Figure 17, the time constant was measured to be 6.2 sec., and with a bubble mass of 75 milligrams, gave a temperature increase of about~~

21° C, which corresponds to a deposited energy of about 3.4 joules. A check of the accuracy of the absorbed energy was obtained from the experimental energy balance. For this experiment, the incident laser energy was

~~13.7 joules, the refracted plus reflected energy was 10.1 joules, leaving~~

~~3.6 joules as absorbed energy, in good agreement with the radiometer measurement.~~

~~CONCLUSIONS~~

The laser-driven implosion experiments 'discussed above support the following conclusions: (1) absorptivity scaling is consistent with the inverse bremsstrahlung process; (2) measured implosion velocities suggest the possibility of some shell preheating; and (3) implosion symmetry is sensitive to illumination symmetry.: Alpha particle and proton measurements have produced information on the central implosion conditions, such as fuel-ibn temperature and tamper electron density. The newly developed plastic bubble will provide a direct measure of tamper fpTdr and another calibrated absorbed-energy measurement.

~~ACKNOWLEDGEMEINT~~

~~The author would like to express appreciation to~~

~~the experimental and theoretical groups at KMSF~~

~~and particularly to R.R. Johnson and R.R. Goforth~~

~~for many'helpful discussions.~~

~~120 REFERENCES~~

~~1. C.E. Thomas and L.D. Siebert, Appl. Opt. 15, 462 (1976).~~

~~2. T.A. Leonard, G. Charatis, M. Gredell and J. Ash, Proceedings of 12th International Congress on High Speed Photography, Toronto, Canada, August 1976 (to be published).~~

~~3. K.A. Brueckner, S. Jorna and K. Moncur, Appl. Opt. 13, 2183 (1974).~~

~~4. K. Moncur, KMSF internal report number U436 (to be published).~~

~~5. D.E. Solomon and T.M. Henderson, J. Physics D: Applied Physics 8, L85 (1975).~~

~~6. C.E. Thomas, Appl. Opt. 14, 1267 (1975).~~

7. G. Charatis, J. Downward, R. Goforth, B. Guscott, T. Henderson, I. Hildum, R. Johnson, K. Moncur, T. Leonard, F. Mayer, S. Segall, L. Siebert, D. Solomon and C. Thomas, Plasma Physics and Controlled Nuclear Fusion Research (Proc. Int. Conf., Tokyo, 1974) IAEA, Vienna (1975).

~~8. F.J. Mayer, G.R. Montry, and E. Benn, Advances in X-Ray Analysis, Vol. 18, p. 169, Plenum Press, New York, 1975.~~

~~9. R.J. Lanter and D.E. Bannerman, LASL Report No. LA-3498-MS July 16, 1966; F.J. Mayer and H. Brysk, Nucl. Inst. Meth. 125, 323 (1975).~~

~~10. F.C. Jahoda, E.M. Little, W.E. Quinn, G.A. Sawyer, and T.F. Stratton, Phys. Rev. 119; 843 (1960).~~

~~11. P.M. Campbell, G. Charatis, and G.R. Montry, Phys. Rev. Lett. 34, 74 (1975).~~

~~12. D.W. Forslund, J.M. Kindel, K. Lee, E.L. Lindman and R.L. Morse, Phys. Rev. A 11, 679 (1975). K.G. Estabrook, E.J. Valeo, W.L. Kruer, Phys. Fluids 18, 1151 (1975).~~

~~13. G.S. Fraley and R.J. Mason, Phys. Rev. Lett. 35, 520 (1975).~~

~~14. R.R. Goforth, F.J. Mayer, H. Brysk and R.A. Cover, KMSF internal report number U477 (to be published in J. Appl. Phys.).~~

~~15. V.W. Slivinsky, H.G. Ahlstrom, K.S. Tirsell, J. Larsen, S. Glaros, G. Zimmerman and H. Shay, Phys. Rev. Lett. 16, 1803 (1975).~~

~~16. H. Brysk, Plasma Phys. 15, 611 (1973).~~

~~17. F.J. Mayer and W.B. Rensel, J. Appl. Phys. 47, 1491 (1976).~~

~~·18. F.J. Mayer, L.D. Siebert and J. Simpson, KMSF internal report number U437 (to be published).~~

~~121 CM cM CM ! I I 1 I~~

~~3i R M iP oRrI s IJ< 'o odCS o d o ld~~

~~ol, n ,- , ,~~

~~u.x Nf-I - 4UI$ 00 00 L 000 0 W~~

~~.1UJ I-~~

~~_. l C A SS C C C -4 D .~~

*<'2 ,o Vmh10ooooclh flr *a' d 0 01-^oe 10C1c1c

~~^s?^ ^~` ^ ^ ' ^^liiFlf~ ^ y~~

~~C,, U) u1 w r + C. w * 0D -*o~~

~~k o~~

~~122 d 0}~~

~~P C 0 mcd~~

~~a *rI O +u~~

~~C 02 02.- 4 '-~~

~~b4U -1 r d~~

~~h *4 ica ,~~

~~h2 a P 0a0202~~

~~ow a0.~~

~~(Df: o~~

~~1d 1C~~

~~- L E _ C4. w t -f Y W~~

~~123 CD~~

~~aJ) rd O~~

~~0u r-I Q~~

~~ww W o~~

~~k w cn^~~~~~~

~~Figure 2 Streak camera record of four stacked pulses spaced at 60 pisoseconds. Three images at decreasing exposure level are recorded.~~

~~124 HIGH POWER LASER BEAM INTENSITY PATTERNS~~

~~WITHOUT FILTER: WITH PLASMA SPATIAL FILTER:~~

~~OUTPUT POWER: 240GW :81.42 OUTPUT POWER: 247 GW 8=: i46~~

~~OUTPUT POWER:s 337W*' B~W.99 OUTPUT POWER: 3:37GW Bet199~~

~~Figure 3 Comparison of laser beam intensity patterns with and without the plasma spatial filter.~~

~~125 P;~~

~~0o a,~~

~~C) 0~~

~~bO *rl~~

~~U(2 .Ho 0~~

~~.'3~~

*rlao

~~0 .r Pa, 4-Cd .1 -1 .H 'd a,~~

~~ca p~~

*i-aE r4B

~~126 %t H av a a)~~

~~o *H 0~~

~~'i 1-1 ()~~

~~.0 H~~

~~dIo 0 CU 0+i~~

~~1 A H( cal 3 Q) 0 z a O H o .° O *H c- ) 0o E d a§n 4 *5S E H .H C) iSO Or C O > c)l~ h a) ±, 0 RU ?~~

~~UL-'~~

~~T1 *rl~~

~~I0 1 O 1 Iu C d 0 _ _ L O~~

~~127 ++) -~~

~~. .A TCAw Ne.p .~~

~~(8»0<)~~~~~~~~~~ 4)..~~~~~~~~~~~~~~~-+Q C~~

~~.4 +Q x g < +- I U; U) +C0 I++ I / ~ +,4.~ + I,,~ -. n a,a + 0~~

/0 r4 0*

~~5-k i- a) +~~ ~~~~~~~~~~~~~~~~E 0 . s-~~

~~+ · w a, ·1~~~~~~~~~~~d o~'. a) tt~ Iii I I Iii I I~ ~ F: * .~~~ ~~mwo~~~n..-.-~gu\eoz S~~~~~~~~~~~~~~ 9.4 ~ ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - ~~~

~~128 0 N (0 U aC.) o U) *H U) o3 0 h1 H 0 O O *R csi · , >, o3~~

~~( a) 4-)g w N -3r-\ 0 a) 0 *H0 tJ a) +P 0 u C) *d H4 C +~~

~~0 C) 4J h o a) 0 d o0~~

~~> * 0~~

~~0 -P +P *Hi 4- *H l4- ,- o ? a)~~

~~c0 4' 0 rA1 o(% pi Pt bO OJ a) 0 *H a) h~~

~~-P 0 o ph h 0- c*rl U a cr *H~~

~~a) 0 C b a~~

~~129 1 _ _~~

~~16.6 um~~

~~2ro 1-25.3 2.3 u~~

~~Target Shot 2369 144 um Diameter~~

~~Figure 8 Off-centered implosion X-ray pinhole photograph and micro- densitometer scan produced by introduced temporal delay in laser beam entering from right-hand side of the picture.~~

~~130 99 pmiamneter Target Shot 2442~~

~~Figure 9 X-ray pinhole photographs of implosions with introduced paraxial focal displacements. Upper figure: no focal shift; central figure: 10 pm focal shift; lower figure: 20 pm focal shift.~~

~~131 QaII m~~

7§

~~o0 4-~~

~~-p0~~

~~1\ i0 o cj~~

~~a E >-r p) Lu ,'d ,a0 o 0 E a a: c,0 a Er~~

~~a a,rI l. -1- rjl0 4t Co~~

*.4 4o

~~P.,~~

~~W 0~~

~~132 O~~

~~o 0 d~~

~~-I so o,0~~

-4

~~I C % i~~

~~I) . rd~~

~~0 H~~

~~H a a CM a f 4 N~~~" O h~~

~~133 o uu o H~~

~~C) i 0 ed rn H cdA N H ro*O~~

~~crN O *H - -Pp E a) m~~

~~L. 0c 5Hc 4 F- ( U a)~~

~~-1 H~~

~~V 00~~

~~10 0+W'~~

~~bD p'H~~

~~134 400 I m - I | U I- l-~ 5 h I I I I~~

~~I -^ -~6:4 DT 1 4~~

~~C ;~~

~~200,~~

~~2O - - - * ^ - - * -; -- * K- -^ - - ^~~

~~I a 3 4 5 6 7 e(keV)~~

~~Figure 13 Ratio of the yield DT alpha particles to DD protons for a 6:4 deuterium to tritium mixture ratio as a function of ion temperature 8. in keV.~~

~~135 cr +3 0) 0 bO~~

~~a)Cd~~

~~I.-~~

~~0 po~~

~~0 *d 0 -t.~~

~~P.,~~

~~136 CaL~~

~~1.06um LASER ENERGY .AvWUV HARD X-RAYS n NEUTRONS '"*~ SOFT X-RAYS a ALPHA PARTICLES o ooo TAMPER IONS~~

~~Figure 15 Schematic of energy flow through bubble wall.~~

~~137 I2 00 1N C5O 2 z W s W -E m- 3 o0~~

~~U)cn ., LLI LL LU LL. I- = 0 A 50 X w WLL. zo -~ g z 61 tg- x c2 C) IdU)~~

~~0 ?i 0 m mw ,0 0~~

~~(0 i-- 0 H 0 M m ^C.ffl ho C) C) cd 00~~

~~0 o -l, WLU~~

~~r-l ^! '0 0 .,H~~

~~U 0 '10 Cc g I 0 b, 8 bo .A W~~

~~138 1 0 H0 0. E o.tw~~~~~S9 OU11 e 0~~

~~w 4J~~

~~.r) -) 4 0H~~

~~(AW* f.rO C.O~~

~~o 0 0 0 0 0 0 4. 0 N C0 4. NV N (AW) indino 8OO1313~~

~~139 ; ?1'~~

~~:I ·!·i I: i=ii,· i · i- .. i' ii:I :-111 : -i· ·:;·· iiri:-;i -I 'i i i ·- i I~~

~~i j -aii r-: 1 m . -i~: i *i -i-~~

~~-I _:i ii :,i·-li - -li ii~~

~~Figure 18 Photograph showing internally-heated blackbody calibration ball and bubble target.~~

~~140 High Power Glass Laser System "Gekko" for Fusion Research~~

~~C. Yamanaka, T. Yamanaka, Y. Kato, T. Sasaki, K. Yoshida and Y. Mizumoto~~

~~Institute of Laser Engineering Osaka University Osaka, Japan~~

~~Abstract~~

The performances of various glass laser systems are being examined.at the Institute of Laser Engineering, Osaka University. Comparisons between various glasses are made including the nonlinear refractive index, stimulated emission cross section and optical quality. The two-beam laser systems "Gekko I" and "Gekko II"' are used in target compression experiments and a four-beam system "Gekko IV" is under construction.

~~1. Introduction~~

Recently great progress has been made in the development of high-power pulsed laser systems. Among them, glass laser is the most developed system in terms of the focusable power on a target and the ability to control the temporal pulse shape in picosecond time scale. In order to achieve scientific breakeven by laser- fusion, laser system will have to be operated at the best possible performance. Although basic understandings related to high- intensity glass laser systems seem to have been made, there are various technical problems to be solved. In this talk, we would like to describe the glass laser fas- cilities at the Institute of Laser Engineering, Osaka University and describe some of our efforts to improve the performance of glass laser systems. Fig. 1 shows our glass laser systems for laser-plasma experiments. The first one "Gekko I" at Nagoya University is being used to study basic laser-plasma interactions. Some of the recent results obtained by this laser will be presented in another talk

141 2) at this conference. The second laser "Gekko II" is a two-beam glass laser system with output energy of 150J per beam in 3ns. This laser is used for symmetric irradiation and target compression experiments. These lasers are equipped with silicate glasses. We are currently constructing a new laser-fascility using phosphate laser-glasses for high density compression experiment in picosecond time scale. It is called "Gekko IV". It is a four-beam system with final aperture of 110mm, and it is expected to give 150J per beam in 200ps. It is scheduled to be completed soon.

2. Gekko II 2) As for the 2-beam laser system,"Gekko II", fig. 2 shows the present setup of the whole system. This laser was originally designed and has been used for amplification of nanosecond pulses. However it is being modified for picosecond pulse amplification for target compression experiments. The oscillator comprises a Q-switched glass laser and a mode-locked YAG laser. After pulse shaping the laser pulse is amplified by 9-stage amplifiers. Two dye cells are placed to prevent the strong amplification of spontaneous emission. Amplifica- tion of reflected and transmitted laser light from a target is suppressed by four Faraday rotators. At present, two aspheric lenses of F number 1.1 are used for target irradiation. Use of reflection optics developed originally by KMS for more uniform irradiation is being planned. Fig. 3 lists the optical components to control the beam quality. In order to suppress the prepulse noise, double Pockels cell and two dye cells are used. Proper choice of transmittance of the dye cell is very important to suppress the parasitic oscillation and still maintaining the reasonable transmittance to the main pulse. Target is isolated by 4 Faraday rotators. Main problem of the Faraday rotator is the optical damage of thin film polarizers. This damage threshold is usually -4J/cm 2 for nanosecond laser light. This damage threshold actually limits the output flux of nanosecond laser systems. Great care has to be taken to reduce Fresnel diffractions in the laser system in order to avoid surface as well as bulk damage of various optical components. 4) We are using glass soft apertures recently developed by HOYA 2 OPTICS. ItsOPTICS. damageIts threshold is about 5J/cm . Since it is an

142 absorption type filter, transmittance property does not change even if local breakdown is produced inside the glass. Fig. 4 shows the diffraction property of this soft aperture. As is shown it produces diffraction rings when it is uniformly illuminated. However in our laser system it is illuminated by a Gaussian beam with a beam radius approximately equal to that of the soft aperture. In this case, we found that the glass soft aperture is very effective to reduce the Fresnel diffracion for the propagation distance of 20m. Fig. 5 shows the laser system seen from the target chamber. We can obtain stable output energy of 150J in 3ns with typical repetition period of 20 min. The energy of prepulse noise is kept less than lOmJ on a target. The focal spot size of the laser beam by an F 1.1 aspheric lens is 100lm. This laser is currently used for target irradiation and compression experiments.

~~3. Gekko IV~~

As for the four-beam glass laser system "Gekko IV", which is under construction and scheduled to be completed soon, this laser uses phosphate laser glasses, and is designed to generate high-intensity picosecond light pulses. Recent studies at various laboratories have shown that, for picosecond light amplification, maximum intensity is limited by the beam breakup effect due to the nonlinear refractive index of laser glasses. This beam breakup effect can be characterized by 6) the breakup integral "B", defined by the first equation of fig. 6,

~~B=k I(Z)dZ. no 5 By integration, Imax is shown to be proportinal to the product of~~

B and the figure of merit of the laser glass MEno0 /n2 . For reasonably uniform target irradiation by laser light, B-value has to be kept less than 4. In this case, output intensity is usually limited to 3'4GW/cm 2 for silicate glasses. In order to increase the maximum intensity, we should use glasses of small nonlinear refractive index and also increase the gain coefficient of amplifiers to reduce the total length of glasses. Gain coefficient can be increased by using a laser glass of high stimulated emission cross section, a. For large amplifiers, stored energy is usually limited by parasitic oscillation inside the amplifier and pumping capability.

143 The table in fig. 7 compares various laser glasses. The last column compares the relative figure of merit of these glasses. 7) From this comparison fluoro-beryllate glasses are expected to produce more than 3 times higher intensity than present silicate glasses. However, it has serious difficulties in production and fabrications. At the present time phosphate glass seems to be the best laser glass. In comparison to silicate glass, phosphate glass has smaller nonlinear refractive index, larger stimulated emission cross section, and very good optical quality. The phosphate glass 4) LHG-7, developed this year by HOYA, has zero temperature coefficient of optical length ( d(nl)/dt=0 ). However it should be noted that fluorescence peak wave-lengths are different for silicate ( 1.062pm ) and phosphate ( 1.054pm ) glasses. This requires the the development of a suitable oscillator whose wave-length matches to the gain peak of phosphate amplifiers. We have experimentally studied the lasing and amplification properties of phosphate glasses. First, mode-locking property was tested using a saturable dye as a passive mode locker. Fig. 8 shows the mode-locked output of a phosphate glass. We found that we can obtain more stable mode-locking with phosphate glasses than with silicate glasses. Pulse width could be controlled up to 200ps with intracavity etalon plates. However mode-locking stability is not yet sufficient as an oscillator of a large amplifier system. But the stability of mode-locked glass laser can be greatly improved by using an acousto-optic loss modulator inside the 8) passively mode-locked cavity as was shown recently by S. Kishida of Nippon Electric Company. Another approach will be to operate YAG oscillator at 1.052pm by suppressing other lines as demon- stratef by J. McMahon of Naval Research Laboratory, U.S.A. We think we can construct a very stable oscillator for phosphate glass system by further developing these approaches. Next we tested the amplification properties. Fig. 9 shows the gain of rod amplifiers of 15mm diameter and 30cm long at differnt wave-lengths. The gain coefficient of phosphate glass at optimum wave-length is 2 times larger than that of silicate glass. Similar results were reported for disc amplifier from NRL. These experimental results support the prediction that we can build superior amplifier system with phosphate glasses. Parasitic suppression is very important to obtain high gain coefficient with large amplifiers. For rod amplifiers, it can be

144 achieved by applying anti-reflection coatings on the ends of amplifier rods and also by index-matching liquids circulated arround the rods. For disc amplifiers, the black solder glass for phos- 4) phate, BSDL-7, seems to be satisfactory since the refractive index ratio is 1.02. Fig. 10 shows the staging of Gekko IV as is planned. We will adopt rod amplifiers up to 8cm and will test both a rod amplifier and a disc amplifier above that diameter. We expect the output intensity of 750GW per beam at 200ps with the final aperture of 110mm in diameter. The development of this laser system will reveal various interesting properties of high gain amplifier system with low non- linear refractive index, and will provide important parameters for future design of more powerful laser systems.

~~REFERENCES~~

~~1) C. Yamanaka, in "Laser Interaction and Related Plasma Phenomena", Vol.2, p.481, ed. by H.J. Schwarz and H. Hora, Plenum, New York, 1972.~~

~~2) K. Yoshida, M. Hohashi, T. Sasaki, and C. Yamanaka, Tech. Rep. of Osaka Univ. 26, 127 (1976).~~

~~3) C.E. Thomas, Appl. Optics 14, 1267 (1975).~~

~~4) T. Izumitani, M. Tsuru, Y. Asahara, Y. Kato, and C. Yamanaka, Paper T2, IX IQEC, Amsterdam (1976).~~

~~5) V.I. Bespanov and V.I. Talanov, Zh. Eksperim. Teor. FiZ. Pis'ma Red. 3 471 (1966) [JETP Lett. 3 307 (1966)].~~

~~6) Lawrence Livermore Laboratory, Laser Program Annual Report-1974 (UCRL-50021-74), p.178.~~

~~7) M.J. Weber, C.B. Layne, R.A. Soroyan, and D. Milam, Paper V6, IX IQEC, Amsterdam (1976).~~

~~8) S. Kishida and T. Yamane, Paper C2, IX IQEC, Amsterdam (1976).~~

~~9) J.M. McMahon, Paper V5, IX IQEC, Amsterdam (1976).~~

~~145 e o o *i o (00 0 *-A 0 4J .- o I M *d *H & U 11 ( ( O 'dOHu N n Nrd4) c(U0 : ft~~

~~O AH )H U~~

~~4 C3rO C) Cn~~

~~Z0 PO~~

~~(- m4O o> Pon Pc:(~~

~~H (0 0 1 m~~

~~Hh > ~0 ~ ~ ~ ~~ot~~

~~rA E- (0 - i 0a. 0o s - f *d*d (M cm N. 4- hhi h) hi U) H 00 0, 0 P5 Pi ErOM0 1lLn An Ln H M rr rH H X~~

~~rzU)M 3 -4)S1, oc'tiC>Oro Ca)ftC Na)OtCo~~

~~Ui H 0> CM O r-O HU to< 0 U 4 H4J 4- _ 4J- E q d -4(~~

~~U 0) 0 *H H (> U'C*) *H- 0 < *X U )~~

~~arCz IH J r4~~

~~H4-i n) H o o o *H f ra1~~

~~H > H H H 0 0 0~~

~~2 K K~~

~~146 GEKKO-II~~

~~I AMPI PREAMP~~

~~V VI VI1 vim~~

~~0 30: SA POL R=50%01 DC POL.~~

~~Setup of Gekko II~~

~~147 GEKKO-II CONTROL OF BEAM QUALITY~~

~~0 Oscillator/Pulse Selection Q-Switched Glass Oscillator/3ns Mode-Locked YAG Oscillator/100ps-300ps~~

* Prepulse Noise Suppression -5 Double Pockels Cell T=105 against Noise Dye Cell (2 in series) T=0.10 against Noise T= 0.65 for Main Pulse (per stage)

* Fresnel Diffraction Glass Soft Aperture (2 in series)

* Isolation of the Target Faraday Rotator (4 in series) Protection 0.01 per Stage

~~148 FRESNEL DIFFRACTIONS (Uniform Illumination)~~

~~APERTURE DIFFRACTION SHAPES PATTERNS (MEASURED)~~

~~HARD APERTURE F=1.9 ( 2a=5mm)~~

~~a 0 a~~

~~SOFT F=5.7 APERTURE (2a=8.7mm)~~

~~0 Radial Distance~~

~~149 "Gekko II" seen from the target chamber~~

~~FlI5.~~

~~150 PICOSEC PULSE AMPLIFICATION~~

* BREAK-UP INTEGRAL

~~n2 B = k- I(Z)dZ~~

~~MAXIMUM INTENSITY * MAXIMUM INTENSITY~~

~~n gB nO (AN)B 0 M-AN-B max n 2 n2 n 2~~

* TO INCREASE I FOR GIVEN B-VALUE, max

~~1) Small n 2 2) Large g - Large a - Parasitic Suppression - Sufficient Pumping~~

~~151 00 0 1-~~

~~r-4 rlC r~~

~~M a- LA c LA in n1 r> ,- I H H O O o~~

~~0ato H rn (U U) in n o< In o u) i- rl rr- r-l~~

~~o H H P<, i t (° 0 H 0 ~~~

~~a t- rl vS *g r-i Ln~~

~~V1 ( _ _~~

~~m A_ P4 hi _ M ]'- >4 Ej O Pi p~~

~~o U P4 PI P H m] O0 0O MA00 H A . ,0 U] P4 DPrz~~

~~152 Mode-Locked Output of a Phosphate Oscillator ( LHG-5 ) ( 5 successive shots )~~

~~153 Amplification Measurement (15xx 300 )~~

~~13 .6 40- 12 'E / PPhosphate U 1 (at 1.054pm )H 11' ~ cn I n Z 0 20- lu Silicate 0 1 /d (at 1.064pm) 8 -/ - ---- Phosphate .~-e_,?^^P Ct (at 1.064pm) 5 I I I i 2 3 5 Pumping Energy (KJ)~~

~~154 E a) U) (L~~

~~U) 0 cn~~

~~I c0 (.D < E ~1 W Cll 0 ,Q( ^ o~ o LL~~

~~1< U)~~

~~1-)~~

~~155 High Power CO 2 Laser System for Plasma Research (Lekko I)~~

~~S. Nakai, M. Matoba, H. Fujita H. Nishimura, N. Banjyoya, H. Daido T. Seki, K. Iba and C. Yamanaka~~

~~Institute of Laser Engineering Osaka University Osaka, Japan~~

~~Abstract~~

High power CO 2 laser system of the output energy 200 J in nsec was constructed. Engineering problems concerning with the high power generation were investigated; generation of short pulse, oscillation in double band with multi-line, conditions of efficient pumping, system for high energy extraction, isolation of amplifier stages, uni-guide to protect optical components.

~~1. Introduction~~

In this paper, the developments of high power E-beam CO 2 laser system in Osaka University (Lekko I) are reported. The laser system Lekko 1 generates a short pulse output energy of 200 J in nsec. It consists of oscillator, Ge mode locker, GaAs optical shutter combined with Ge polarizer to select out one pulse from mode locked pulse train or to slice out a pulse of few nsec-duration from a normal Q switched pulse, preamplifier, E-beam sustained main amplifiers No.l, No. 2 and No. 3. The schematic diagram of the system is shown in Fig. 1.

2. Oscillator Three type oscillators were developed. The first one is so called TEA lasers of two different kinds. The second one is the double band oscillator of 10.6 pm and 9.6 pm. The third is a high pressure laser of 15 atmosphere.

~~2.1 TEA oscillator and pulse shaper The Peason-Lamberton type1 ) and a ladder discharge U-preionized~~

TEA laser operated stably in TEM0 0 mode with a simple structure. By using GaAs optical shutter combined with a LTSG, arbitrary pulse length longer than 2 nsec could be cut out from normal Q switched

157 pulse. More shorter pulse can be obtained to select out one pulse from mode locked pulse train. The stable mode locking was achieved by two methods. Active mode locking was obtained using Ge acousto- optical modulator of driving frequency 20 MHz. Passive mode-locking was obtained using P type saturable Ge plate.

~~2.2 High pressure oscillator There is another method for the generation of shorter pulse.~~

15 atmospheric pressure CO 2 laser was developed successfully using 2)3) E-beam sustained discharge.2) High pressure CO 2 laser is very interesting to get short pulse due to wide spectrum band, and also to keep the fast relaxation between the lasing level and energy stored levels by sufficient collisions for efficient energy extraction in short pulse amplification. The dimension of excitation volume is 20 cm long and 3 cm wide, and anode-cathode spacing of pumping discharge is adjustable from 1 to 3 cm. As the laser window, KCl plate of 5 cm thick was used. The typical performance is shown in Fig. 2. Stable discharge up to 15 atmosphere with arbitrary gas mixture was achieved. The limitation of pressure came from mechanical strength of discharge chamber and foil window. Specific electric field strength of discharge E/p and energy input to laser gas can be independently controlled. Small signal gain coefficients for the probe laser line of P(20) are plotted in Fig. 3. They are shown as a function of discharge pumping input for several laser gas pressures. Small signal gain increased with increasing pumping and showed a tendency of saturation. Using 36 % reflectivity output mirror, which is polished Ge plate, 2.5 Joule output in 80 nsec pulse was obtained. The domi- nant line of oscillation was R(16) in 10.6 pm band. Pumping energy was 200 J/£.atm and laser efficiency was 2.2 % to input energy. Small signal gain was also derived from the threshold measurements of oscillation. In Fig. 4, the comparison of small signal gain in P(20) and R(16) is shown. The gain of R(16) was higher than P(20). It is a reasonable result because the overlapping of rotational lines in R-branch is easier than P-branch due to close spaeing. The pressure dependence of small signal gain gives the pressure broadening coefficients. The experimental results for several input energy and calculated value from broadening coefficients of 3.5 GHz/atm are shown in Fig. 5 by the broken and solid line respectively. They show a good coincidence. The pressure broaden-

158 ing coefficients of about 3.5 GHz/atm is consistent with average value 4)-7) of recent results by absorption measurement. This laser is now under investigation of mode-locking oscillation to generate ultra short pulse.

2.3 Double band and multi-line oscillator As for the efficient extraction of energy from amplifier stages by short pulse, a multi-line, double band laser is useful. Simul- taneous laser oscillation in 10.6 pm and 9.6 pm band has been achieved by forming double cavity on the same gain media 8) using a dispersive element like gratihg-in the cavity. With a absorption cell in the cavity similar simultaneous oscillation has been observed. These techniques are very important in high power,high efficiency laser of short pulse for laser fusion.

3. Amplifiers 3.1 Amplifier system E-beam controlled amplifier No. 1 was designed to use a hot cathode electron emitter to achieve the long electron beam pulse as to preserve the flexibility of the performance for the several testing purposes. The dimension of active medium is 5 cm x 5 cm x 80 cm. Performance characteristics such as small signal gain (N5 %/cm), gain distribution, gain history relative to the pumping current has been investigated. Amp. No. 2 has a cold cathode with thin blades of Titanium foil of 30 pm thick. The volume of pumped region is 10 cm x 10 cm x 100 cm. The performance characteristics of this amplifier were also investigated. As for the Amp. No. 3, the dimension of the pumped region is 20 x 20 x 200 cm. Cold cathode E-beam gun is used.

3.2 Energy extraction Energy extraction experiments have been performed on the amplifier chain. In Fig. 6, the energy extraction is shown using the input pulses of various time durations, 3 nsec, 12 nsec and 37 nsec. From these data, we can calculate the saturation parameter as a function of pulse length. The result is shown in Fig. 7. Saturation parameter decreases with decrease of the pulse length. In longer pulse length saturation parameter had the tendency to be constant. These characteristics of saturation parameter are explained by the finite relaxation rate

~~159 of rotational levels.99) For our experimental condition, the saturation energy of single rotational level and all rotational levels are respectively,~~

~~E 9.4 mJ/cm 2~~

Es2 E s -k(JT)- 139 mJ/cm where, k(J0) is the partition fraction of the transition level J0 = 19. Using measured small signal gain 5 %/cm, and active volume and available data about saturation, the capability of each stage of our system is shown in Table 1. The upper corresponds to single rotational transition and the lower corresponds to all energy extraction. In Table 2, performance of each stage are shown. When multi- line oscillator is used, output energy of 500 J in nsec is expected.

3.3 Pumping by PFN The use of a pulse forming network (PFN) in E-beam controlled CO2 laser pumping improve the utility efficiency of capacitor. The impedance of laser gas is controlled by the accelerating voltage of the electron beam to match the PFN line impedance. In the matching condition, all of the stored energy in the PFN is delivered to the laser gas at constant electric field strength which is optimum in laser pumping. We have designed and tested a PFN power source for pumping of Amp. No. 1 of Lekko system, the active size of which is 5 cm x 5 cm in cross section and 80 cm in length. The electron gun is hot cathode and the gun voltage is varied from 160 KV to 240 KV to adjust the discharge impedance for matching with that of PFN. The discharge current and voltage wave forms of matched and mis- matched condition are shown in Fig. 8. In Fig. 9 the variation of impedance and input energy as a function of gun voltage are shown. It is seen that at the matched condition maximum energy input is obtained with constant source energy. At optimum condition 100 % transfer of source energy is achieved.

~~4. System engineering~~

~~In the amplifier chain with high gain medium, the isolator between each amplifier is one of the important components. Para-~~

160 sitic oscillation wastes inversion population and degrades the contrast ratio of main pulse. We are developing the saturable absorbers of gas cell, film, plate of various material to suppress the parasitic oscillation. Another serious limitation for high power laser system is optical damage in laser components. The laser beam diameter should be designed for the energy density lower than the damage threshold. It is important to get homogeneous field intensity across beam radius and to prevent the reflected light to be amplified by uni- guide for protection against damage.

~~5. Conclusion~~

High power CO 2 laser system of 200 J output energy in nsec single pulse was constructed. Double and multi-line oscillator and high pressure oscillator were developed for the purpose of effective extraction of stored energy in short pulse amplification. Saturation parameter was measured in nsec pulse length. Measured values were explained by the finite relaxation rate of rotational levels. E-beam controlled amplifiers were pumped uniformly up to 300 J/Z.atm. Technical problems such as isolator to suppress the parasitic oscillation, uni-guide, damage threshold of optical components were investigated.

~~REFERENCES~~

~~1) P. R. Pearson and H. M. Lanberton; IEEE J. Quant. Elec. QE-8 1972 p. 145.~~

~~2) N. G. Basov, E. M. Belenov, V. A. Danilychev and A. F. Suchkov; Kvantovaya Electron (Moscow) No. 3 1971 p. 121 [Sov. J. Quant. Electron 1, 1971 p. 306.]~~

~~3) N. W. Harris, F. O'Neill and W. T. Whitney; Appl. Phys. Lett. Vol. 25 1974 p. 148.~~

~~4) E. T. Gerry and D. A. Leonard; Appl. Phys. Lett. Vol. 8 1966 p. 227.~~

~~5) T. K. McCubbin et al.; Appl. Phys. Lett. Vol. 8 1966 p. 118~~

~~6) U. P. Oppenheim and A. D. Devier; J. Opt. Soc. America Vol. 58 1968 p. 585~~

~~7) R. R. Patty et al.; Appl. Opt. Vol. 7 1968 p. 2241~~

~~161 8) M. Keller, M. Matoba, S. Nakai and C. Yamanaka; Japan J. Appl. Phys. Vol. 4 1975 p. 423~~

~~9) G. T. Schappert; Appl. Phys. Lett., Vol. 23 1973 p. 319~~

~~Table 1 Limitation of laser output by single line and multi- line amplification~~

~~Amp.1 Amp.2 Amp.3~~

~~E 0 alS 1 5 40 S (J)~~

~~E lt1S 15 74 578 s (J)~~

~~162 II - l 1~~

~~i 1 i (1) 1 - 8 E o- 0 C\( 0 oo 0 0 0 l 1 o0 LO Q d 0 0 i 0 X in L)in LO 0C) 1 r-1 (D 0 QS O -u0 t.4 LC) 4-1 LC) 0 U- 1 U) u 00 12 0 X .,- 1 l~~

~~1 (0~~

~~i U i i 0 LO j QE 1 ,-4 x x in 1 (0 0 1 E-4 Xx LO l~~

~~163 0_\L. .C::~E .. - --- u a~~~~ -c_ > on. ¢, /~~~~~~~~~~~~~~~~.s/'..O ~..CI- u_ .0 d :>.-s.u~ /~~~~~~5/1^i Y_ n _ 0-; _z 5·^:~~

~~B n 10 ytEC) \ s.I~~~~~~~~~~~ Co 0 C ,,~~

~~.2 1- C£Li ^i%52 ~E.o § E : G:0 ~_ :nn ^J ; _c.EU 0 o =~~

~~?- :- L j j f.0~~

~~- 0-· .------( ,--4"~~

~~) i-ii0 -"-} *u0 ~·bcn3, --^--~~

~~b *3"g - a £--~S~g ^\ S~~~~Fre-C Amp. iiLit, 0 ^% - *< - - 3 i i71 ' ,i~~ro m.LLI~~

~~Pro. Amp. 1,11,1111~~

~~164 Gun Voltage 160 KV 1.5 Ps~~

~~- -- Beam Current~~

~~Discharge Current 1:^zj-~~

~~Small Signal Gain --- / CO2:N 2 :He -_- =1:1:3 (9 atm) -- 1ps/div E P=3.7 KVicm. atm~~

~~Fig. 2 Performance of high pressure CO2 laser~~

~~165 0 0 4 cl 0 3 O U Cl 2 r) 1 E 0 0 50 100 150 Energy Input (J/l-atm)~~

~~Fig. 3 Gain characteristics of high pressure CO2 laser~~

~~166 r-~~

~~aE U '0 4 o- /i1 c- 3/ (9 // P(20) a2~~

~~an CCO 2 :N2 : Hei=l :1:3 -i 1 9 atm E U) 0 L 0 50 100 Energy Input (J/l-atm)~~

~~Fig. 4 Gain characteristics of R(16) and P(20) in high pressure~~

~~CO 2~~

~~167 00 a- 0 4 c 0 3~~

~~en 2 OI (n 1~~

~~-f 0 5 7 9 11 13 15 Pressure (atm)~~

~~Fig. 5 Gain overlapping by pressure increase~~

~~168 E~~

~~00 E c~~

~~o oOa\II -6I E 4-> 00.~~

~~0 .c- C1 c IC~~

~~o so C 4J 4 D\ - 4.,~~

~~0 o 0 0 C) ( wfl) Ino /'.Jau3~~

~~169 200~~

~~- U Es -0 -- )~~

~~100 WWI)~~

~~CO:Ni^H 11-1 3 p= 1.5atrnI. nsec rE TR=0.106 0 50 'p (nsec)~~

~~Fig. 7 Saturation parameter as a function of pulse width~~

~~170 Impedance Matchfing~~

~~Gun 240 KV PFN 60 KV Discharge I. 3 KA/div Discharge V. 0 KVldiv~~

~~Gun 180 KV PFN 60 KV Discharge I. 1.5 KA/div Discharge V. 20 KV/div Mismatching 8.8Q Sweep 2 Ps/div~~

~~Fig. 8 Laser pumping by PEN source~~

~~171 (f) A6Jou3 0 0 0ok 0 0 CO CN 0 T r I I 1 0 / '4T 4' 0 X 4)- COl . 0 *r X (0 u~~

~~0) 0 X _·I 0a O 0 E n--0 c n--- 0 0 C-~~

~~0 ouOpdu (W) ecuopedwiI~~

~~172 Super-Compression and Its Stability of Multi-Structured Pellet~~

~~Keishiro Niu~~

~~Department of Energy Sciences Tokyo Institute of Technology Ohokayama, Meguro-Ku, Tokyo Japan~~

~~Abstract~~

~~To compress the multi-structured pellet, an optimal time dependence of the input laser power is derived. A criterion for the Rayleigh-Taylor instability is also given for this compression method.~~

§1. Introduction The compression ratio of the medium through one shock wave is limitted in a finite value of order unity. In order to compress the Deuterium-Tritium fuel to more than 1000 times of the solid density at the pellet center, nearly adiabatic successive compressions are preferable. In §2, we propose one-dimensional analytical model for super-compressions of multi-structured pellets. In §3, a criterion for the stable implosion with respect to the Rayleigh- Taylor instability is examined for our super-compression model.

173 §2. Super-Compression of Multi-Structured Pellet A pellet is assumed to consist of a D-T fuel (region f in Fig. 1) and a high Z material (region h). Among the velocity V of the ablation surface A (whose Mach number is

~~N), the fluid velocity ub and the sound velocity Cb just out of the ablation surface, we assume the following relation~~

~~V - ub = aCb. (1)~~

~~Here a is a constant and a=l leads eq.(l) to the Chapman- Jouguet condition1 3). If we limit ourselves to one-dimensional plane configuration to simplify the analysis, we obtain the relation~~

E = 21 - -1 (2) 2(y -1)N e y+1/ where E is the laser power absorbed in the ablation surface, Y the ratio of the specific heats, p the pressure, C the sound velocity. The suffix a refers to the values just before the ablation surface. Accompanied by the acceleration of the ablation surface A from N to N+AN', two weak shocks

~~Sa and Sb (whose Mach number are 1+ M i and 1+ Mb, respectively) are generated as shown in Fig.2. As shown in Fig.3, 'shock~~

Sa propagates in the region h to the discontinuity surface B, and a part of it transmits into the region f as a weak shock St and a part of it comes back to the ablation surface as a rarefaction wave Ra. We represent the rarefaction wave as a line and characterize it by its Mach number 1-AMra as a conventional way. The changes in Mach numbers of the shock wave and the rarefaction wave in the region

174 h due to the interaction of them can be neglected in comparison with their absolute values. The velocities arid the pressures must be equal in two sides of the discontinuity surface B. Thus we obtain a relation

~~AMi = AM + AM (3)~~

As shown in Fig.4, the rarefaction wave Ra accelerates the ablation surface from N to N+AN" and transmits as a rarefaction Rb whose Mach number is 1-AMrb. If we combine AN' and AN", we obtain a relation between AN=AN'+AN" and AMi,

~~AN 2B(y-l)G-4 f(y +1)-y$(2-G) _- = ----- AM. , (4) N CY+1)(y+l) l where~~

~~2(y2-1)NE $ = 1 - / , = 2' (5) YPaCa(1-N2)~~

~~AMt 2 G = - = (6) AMi /Pho/Pfo + 1~~

In eq.(6) ? is the density, the suffices h and f refer to the values in region h and f respectively, and the suffix o indicates the initial values. Let us denote the number of shock waveslaunched from the ablation surface A per unit time by n. In the limiting case that n- , AMi+O, nAMi-A, eq.(4) reduces to

~~2B (y-l) G- 4/(¥y+1)-yB (2-G) N C exp At , (7)~~

~~whereis thet time after the pellet is irradiated. where t is the time after the pellet is irradiated.~~

175 On the other hand, we consider the condition that all the shock waves in region f converge at the pellet center at the same time O' shown in Fig.5. Omitting the detailed derivation, we can give the propagation time T of a shock wave which is launched from the ablation surface A at t and reaches the pellet center at t+T as follows,

~~T - exp (-2GAt) . (8)~~

~~Combining eqs.(2), (7) and (8), we can derive an optimal time-tailoring form of the laser power E,~~

~~E T , ) where 2G{YB- /B (y+l-y¥)} +4/ (y+l)-y3 Q = ------(10) (y+l)BG~~

~~The values of g are plotted versus PhO/PfO in Fig. 5 for a=0.5, 1 and 2.~~

§3. Criterion for Stable Implosion with Respect to Rayleigh- Taylor Instability at Ablation Surface The laser power E given by eq.(9) generates a peculiar acceleration of the ablation surface. Bodner4)derives the following relation for the growth rate ym of the Rayleigh- Taylor instability at the ablation surface,

~~qc3+(q 2l+Q)o2+(2qQ-q+2£q 3 )o~~

~~+(1-q) (1-Q)+cq 4 = 0 . (11) where~~

kur Ym nkpa Pb , , - Q - , - (12) vkg pag Pa 176 In eq.(12), ur=V-ub is the ablation velocity, k the wave number of the Rayleigh-Taylor instability, g=--dV thethe accelration of the ablation surface and n some unknown constant which is the same quantity as B defined in Ref.(4). In view of eq.(ll), we can easily get the sufficient condition for stability

~~q2 > 1 and Q > 1. (13)~~

~~Consequently the range of k for stability is~~

~~{ g/Ur2 k > larger of (14) yg/nCa-a~~

~~By use of the results in the preceding section, eq.(14) reducesto ( 2(y-1) 4yG-8(2-G)/o I --- exp At. (15) k > larger of atsC (y+l) 2y(y-l) 4G --- 2 -c exp t . (16) In(+1) tsCo y+1~~

In eqs.(15) and (16) t s is the time at O' in Fig.5. The exponent in eq.(16) is always larger than that in eq.(15). Accordingly the sufficient stability criterion is only eq.(16), when t becomes large enough. Here we set that the minimum k is equal to 1/R, where R is the distance from the ablation surface to the center.

~~1 t+T -4G - =R = V dt= Ca0ts exp -At. (17) km m~I+ t y+1~~

~~Substituting eq.(17) into eq.(16), we obtain~~

~~2(y-1) 1 > 2 (18) n (Y+l)~~

~~177 Since n is defined as~~

~~1 (dE/dt)/E (dCa/dt)/C a- (dN/di)/N _ ------==1+ 1 n (dpa/dt)/Pa (dPa/dt)/Pa (2-G) = 1 + (19) ayG by Ref.(4), eq.(18) reduces to~~

~~PhO 11 2 =( )(20)~~

~~Pf0 3 with y=5/3. Thus the pellet which has~~

2 Pho ( -) 13 (21) PfO 3 becomes stable for the Rayleigh-Taylor instability, provided a=l. Equation [17] shows that the pellet radius converges to zero with t'o. If eq.[17] is slightly modified to satisfy that many shock waves converge not to the pellet center but to the radius 6, then

~~expk [[*4 y 4 A G$]] (22) CotS+a exp[y+ AGt] and the stability condition becomes~~

~~1 2y(y-1) (23)~~

~~1+(6/Cot 5 ) exp[4-TGAt] n(+l) 2~~

~~Equation (23) cannot be satisfied whent-*. In order to realize a stable implosion, we turn off the laser light at t=tO. Let us define the compression ratio + as~~

~~178 final density of fuel 4 ( = : = exp [ -AGto] initial densith of fuel Y+1 then 6, the deviation of the shock-converging surface from the pellet center, must satisfy the following condition,~~

~~Cot s RO-6 RO R6 ~ (24) where RO is the initial pellet radius. Relation (24) will be satisfied when 6<0. Thus we conclude that the stable implosion requires the correct time-tailoring form of the~~

~~laser power and especially the correct estimate of t s.~~

Refferences 1) S. Mikoshiba and B. Ahlborn: Phys. of Fluids 17 (1974) 1198. 2) L. D. Landau and E. M. Lifshitz: Fluid Mechanics (Pergamon, New York, 1959). 3) R. Courant and K. O. Friedrichs: Supersonic Flow and Shock Waves (Interscience, New York, 1948). 4) S. E. Bodner: Phys. Rev. Letters 33 (1974) 761.

~~179 - -~~

~~p V '-4 I !~~

~~(f)~~

~~Ua '- r 0 B A~~

~~Fig.l: Density distribution in the pellet.~~

~~180 ablation t surface 1 A~~

~~shock shock S wave Sb wave a~~

~~r 0 Fig.2: Interaction among the generated weak shock waves and the ablation surface.~~

~~discontinuity t surface B rarefaction shock wave t wave Ra~~

~~shock Si wave~~

~~Fig.3: Interaction among the weak shock waves, the rarefaction wave and the discontinuity surface.~~

~~181 t ablation surface A rarefaction wave / Rb ./~~

~~rarefaction wave Ra r 0~~

~~Fig.4: Interaction among the rarefaction waves and the~~

~~ablation surface.~~

~~182 t~~

~~U /R~~

~~0 A~~

Fig.5: r-t diagram of the ablation surface (solid line A-F), the successive shock waves (solid line A-B, B-O', A'-B', ...), the rarefaction waves (double chain line B-B'-E). Dotted lines denote the pathes of particles.

~~183 15~~

~~~~1 ~~~c0 =05~~

~~c =I~~

~~5--~~

~~I I ,illii I I Ill ll 1 5 10 50 100 Ph0o/Pf~~

~~Fig.6: Exponent Z of the tailored pulse (eq.(9)) is plotted vesus the density ratio for a=0.5, 1 and 2.~~

~~184 NON-LINEAR PROCESSES IN A DENSE PLASMA~~

~~A. SAMARSKIJ, S.P. KURDYUMOV, N.V. ZMITRENKO and A.P. MIKHAILOV Institute of Applied Mathematics, Academy of Sciences of the USSR Moscow, 1976~~

~~Abstract~~

NON-LINEAR PROCESSES IN A DENSE PLASMA. The paper presents the main results of the study of phenomena which accompany the development of strongly non-stationary processes (so-called peaking regimes) in a continuous medium whose description includes quasi-linear transfer equations.

~~INTRODUCTION~~

A study is made of quasilinear processes of transfer in a continuous medium (the coefficients of heat conductivity, conductivity, etc. depend on temperature, magnetic field strength and other parameters) under conditions where peaking regimes occur. Peaking regimes mean strongly non- stationary processes with parameter values rising with time in accordance with a law under which they become infinite within a finite time. An example of such an externally fixed boundary regime is the law of growth of a laser flux Go G = -- (Fig. 1), which is used for the near-adiabatic compression of the target nucleus in (tf - )2 laser-induced controlled nuclear fusion [ 1]. Peaking regimes may also occur in a medium owing to its non-linear properties - for example, as a result of the action in the medium of non-linear volume distributed sources. The study of peaking regimes [2-6] has shown that in combination with the non-linear properties of the medium they cause a number of new, paradoxical features characterizing the quasilinear processes of the transfer of heat, the magnetic field, etc. Strongly non-stationary processes give rise to metastable localization of the heat, of the magnetic field and of other quantities on particular spatial scales or mass sections (in a compressible medium). Localization is the internal reason for the decay of the medium into thermal, magnetic and other structures. These structures (i.e. inhomogeneities of the temperature, the magnetic field, etc.) are open thermodynamic systems which sustain themselves by converting ordered hydrodynamic motion or the magnetic-field energy or thermonuclear (chemical) energy into heat. The processes of generation, self-sustainment, multiplication and complex interaction of structures in a plasma have been observed repeatedly in physical experiments (see, for example, Refs [7-9]), in the numerical modelling of non-linear MHD and RMHD processes, and in a number of theoretical studies (see, for example, Refs [10-15]). The development of structures in a peaking regime is governed by a number of general, fairly simple principles which can tentatively be called the laws of thermodynamics for strongly non-stationary processes. Structures can co-exist and combine - thereby complicating the organization of a medium - only when certain conditions are fulfilled (the principle of superposition of non-linear systems). By way of example, we shall analyse some very simple problems using the equation of non-linear heat conductivity in a stationary medium. This enables us to show clearly the laws of the processes being studied and to formulate the principal conclusions.

185 The metastable localization of heat is considered mainly m the study of heat propagation into a cold half-space. The temperature (or heat flux) at the boundary vanes in the peaking regime, and this simulates the action of a peaking laser pulse on the medium. The parameters of a DT plasma heated in a peaking regime are evaluated. The development of the structures and the laws of the thermodynamics of strongly non- stationary processes are analysed mainly in connection with the burning of a medium with non-linear heat conductivity and volume heat sources. We give an example of the complex combining of different structures in a plasma with allowance for gas-dynamic motion, heat transfer and other processes.

~~METASTABLE LOCALIZATION OF HEAT~~

~~1. In the simplest case, the process of heat propagation in a medium with non-linear heat conductivity is described by the equation~~

~~a ar (K(T) (1)~~

Here T(r,t) is temperature, t time and 0 < r <+ oo the coordinate of K(T) = KOTU, a > 0 being the coefficient of heat conductivity. The solution of different boundary problems and Cauchy problems for Eq. (1) showed that heat penetrates a cold medium (K(T) = 0) with a finite velocity [16-18]. The finite velocity of propagation is a direct consequence of the dependence of the heat conductivity coefficient on temperature. With ordinary boundary regimes which grow with time but not in the peaking regime, heat propagates in the form of a heat wave with a fimte front whose coordinate increases with time. The effective (skin) depth of heating of the medium by the heat wave also increases. Below we consider the class of boundary peaking regimes (S or LS regimes) whose action causes metastable heat localization, the region with non-zero temperature not changing during a finite time interval. The temperature and the quantity of heat inside the region of localization may increase to arbitrarily large values. The effective depth of heating of matter by the heat wave (half-width) remains constant (S regime) or decreases with time (LS regime).

~~2. At the boundary of a non-heated medium,~~

~~T(r, to) = 0 (2) let the temperature rise in accordance with the law~~

~~T(0, t) = T (tf-t)n , n < 0, To = const. > 0 (2')~~

For to < t < tf, the law (2') simulates the growth of temperature m the peaking regime with the instant of focusing (the temperature becoming infinite) t = tf. When to = - c, the problem (1), (2), (2') is self-similar. Of course, the actual process begins from some finite instant t = to = - oo. If the initial data for t = to are not self-similar (for example, zero values), some time is needed for the process to become estabhshed and for its laws to become clear.

~~3. When n <- 1/a, there is no localization. Heat propagates into the non-heated medium in the form of a wave with growing half-width; the wave front has a finite velocity (HS regime).~~

186 In this sense, the solution is in no way different from those known earlier [16-17]. The boundary regime forms a "convex" profile of a heat wave moving deep into a space. When n = - 1/a, to = - oo, the problem (1), (2), (2') has an analytical solution in the form of a standing heat wave (S regime), which was first studied in Ref. [19]:

T(r, t) = fT(tf-t r) - / (1-r/r o)21a, rrp (3) where r,, is the depth of heating of matter by the standing wave, determined by the properties of the matter and by the intensity of the boundary regime (by the constant To):

~~rp = (2 KoT g (a + 2)/a)l/2 (3A)~~

The solution (3) demonstrates the effect of metastable localization of heat in the S regime. The heat wave front is immobile and the half-width constant in spite of the fact that, as the instant of focusing approaches, the temperature and the quantity of heat in the region 0 < r < rp tend to infinity. When the S regime starts at instant t = to, to * - oo, and at zero initial values, it needs a certain time to become established. Propagating in the HS regime, the heat wave first attains a depth r = rp and then stops, and the S regime proceeds to materialize nght up to instant t = tf. In the case n > - -, heat propagates in the LS regime Figure 2 shows temperature profiles at different instants obtained by numerical calculation of the problem (1,),(2), (2'). The LS regime of heat penetration is accompanied by a reduction in the half-width of heating with time. The crosses indicate half-widths, which decrease after the regime is established The next equal portions of heat arriving at increasingly short intervals of time are localized in the zone diminishing with time near the heating boundary. ' -

4. The examples given here illustrate the peculiar inertia of heat in a medium with nonlinear - * heat conductivity.. We formulate the pnncipal results relating to the effect of metastable localizationi of heat [4, 6]: Localization is caused by the specific "concave" nature of the temperature profile in a heat wave. Localization is characteristic of a particular class of boundary peaking regimes. Any boundary peaking regime which does not change with time faster than the boundary S regime (i.e majonzed by the S regime) creates a localizated heat wave profile. The temperature and the quantity of heat in the zone of localization can increase indefinitely. The effective depth of heating is constant or decreases. - In the peaking regime, the process of heat propagation as the instant of focusing approaches (at a developed asymptotic stage) is determined only by the boundary law and does not depend on the initial values. In particular, localization occurs effectively also for a non-zero initial temperature background. The conclusions are based on rigorous mathematical theorems and on a study of analytical, self-similar and numerical solutions of the corresponding problems. The results are generalized for a number of cases where heat propagation is more complex - for example, for the multi- dimensional case.

5. The boundary regimes leading to heat localization make it possible in principle to concentrate any quantity of energy in a specified region of space and contain it for a finite time. It is of interest therefore to estimate the parameter values which a plasma attains as a result of heating by a strong laser pulse in a peaking regime giving rise to localization - for example, in the S regime.

187 Let us assume that a laser pulse is heating a completely ionized DT plasma (o = 2.5), that the laser radiation is being absorbed at the boundary and that gas-dynamic motion, the self- radiation of the plasma and other processes can be neglected [20]. At instant to = 0 a plane layer of matter of area s = r2 adjacent to the vacuum begins to be heated. With the above assumptions we can make estimates on the basis of the analytical solution (3). Establishment of the S regime for a = 2.5 is ensured when the condition tf/e Ž 103 is fulfilled, where tf is the focusing time, tf- e the time of conclusion of the heating process and e the time of maintenance of temperature TM (tf- e) attained by instant t = tf - e. The thermal flux at the boundary in the S regime vanes in accordance with the law

~~a+1 W(0, t) = W o (tf - t) a =Wo(tf-t)-1, 4 (3B)~~

When tf/e/ 103, the ratio of the maximum flux WM(tf - e) to the initial flux WH(O) is WM /WH - 10- 110 s. Half of the pulse energy is spent in time 5 e s tf/200 (Fig. 1). The parameters of the pulse are connected with those of the plasma by the following relations: /Watt /p\05 WMIm t 6.7' 1l30'T . 2 (keV) e-°S(nsec)' (-)

~~r, (cm) 1,110-2 T (keV) e (nsec) .) (3C)~~

~~2Watt PO\14/9 Eo(J) 1.5.10-2.e 8 2 3/9(nsec).WM/9(Watt-9 (-P cm 2 -/ \ p/~~

Here TM (tf -e) is the maximum temperature, r, the depth of heating, EO the total energy of the laser pulse, p the plasma density in g/cm3 and po - 0.2 g/cm3 the density of the condensed DT mixture. The influence of gas dynamics is assessed on the basis of the ratio r,/rg.d., where rg.d. is the depth of penetration of the rarefaction wave calculated from the speed of sound at the boundary:

~~/ p\o.- rp/rg.d.- 2 10- 3 e-0 5 (nsec)' T M75(keV)'--) (3D) \PO/~~

~~We give a table of estimates for plasma density p = 10- 2 P, P _ 2' 10-3g/cm3, tf/e = 103: (4)~~

~~No. TM (keV) WM(cm- rp (cm) e (ns) Eo (J)~~

~~1 1 1014 8 X 10 -3 5 X 10 -3 10-1~~

~~2 3 1015 3.4 X 10- 2 6 X 10-3 10~~

~~3 8 1016 1.1 X 10 - 1 5.2 X 10 - 3 103~~

The relation r0p/rg.d. ~ 1 is valid for cases 1-3. Since the plasma density and the confinement times are small, it is mainly the electrons which are heated. In analysing the possibility of obtaining high ion temperatures and an appreciable neutron yield, it is necessary to take into account gas-dynamic motion and other processes.

188 Localization can be detected experimentally in a temperature and time range other than that in estimates 1-3. This follows from the similarity relations, which are confirmed by numencal calculations. When the scale of the characteristic times increases by a factor a, the temperature scale diminishes by a factor a- 1/a.

~~DEVELOPMENT OF THERMAL STRUCTURES~~

1. Peaking regimes can occur in a medium with non-linear heat conductivity and without boundary regimes as a consequence of the action of non-linear volume sources of heat [5, 6]. Let us in fact consider, in the region - oo < r < + oo, the Cauchy problem for the quasilinear equation of heat conductivity with a source (the problem of burning)

~~atT a T(KT + o TT, q0 >)> (5) for t > to with the initial conditions~~

~~T(r, t o) = To(r) (6)~~

The initial perturbation of temperature To(r), which initiates burning, is specified over a finite interval or over a number of finite intervals or throughout the space. For values of /> 1 there is a solution of the problem (5), (6) in the form

~~T(r, t) = qn(tf-t) n f(a), (7) = r[ Koqgna (tf-t)m] - , where~~

~~1 +a-B/ n =(1 p)-, m = (1 + no) 2 2(1-n) and f (A)is the solution of the equation~~

~~- nf + mf' = (fa f') + f, f'-, (8) where fO f' = 0 for r = + oo. The function (7) is the solution of the problem (5) with initial data in a special form:~~

~~n To(r) = q (tf- 0t ) f( o) -Ta(r, to) (9) where~~

~~o(r) = r [ Ko q (tf- to)m ]- 1 (9A)~~

~~The value of tf (instant of focusing), which determines the duration of the existence of the solution (tf- to), is found from the initial data:~~

~~189 ffi-1 (0) f/-l (O)^= to + 1 tf = t+ 0 Tt- (0, to) (10)~~

The temperature in the solution (7) grows as t -e tf in the peaking regime Peaking regimes also result from the solution of the problem (5) with initial data in another special form. To(r) = T(to) = const. This is the solution of the problem of homothermic heating and has the form

n T(r, t) = T(t) = {(/-1) q0 (tf -t) (11) where tf is determined on the basis of expression (10) with f(O) = (3- 1)n A study of the stability of the solution (11) in relation to small perturbations (6 T = A(t) exp i (cot - r)) shows that, when p < a + 1, it is stable for perturbations of any wavelength and, when / > a + 1, it is unstable for perturbations of any wavelength, the perturbations growing m accordance with the law (tf - t)fn When P = a + 1, homothermic burning is unstable for perturbations with wavelengths greater than the critical wavelength (X > Xc = 2r/a + 1 /Kl q ), the perturbations growing in accordance with the law (tf - t)n ' , where n, = fn (1 -(Xc/X) 2 ).

2. A study of the problem (5), (9) leads to the conclusion that there are three regimes of burning of the medium: (1) HS regime (for 1 a + 1). From expression (7) it follows that the point of the temperature profile with fixed E (wave of state) moves in the direction of decreasing Iri (reduction in the effective width of the region of burning). At the same time, the heat wave front is at infinity rp = + The study of the temperature profile for solutions of the form (7) is based on a study of Eq (8). With the substitution of x = fa+ 1 it is transformed into an equation of motion of a material point in a field of forces.

~~x"=mix +l'-(o+ l)nx +l+x + l1 (12)~~

If m - 0, the field of forces (12) is not conservative In the HS regime a friction force (m <0) and in the LS regime a building-up force (negative fnction, m > 0) act on the point. In the case of the S regime (m = 0), the field of forces is conservative - its potential has the form ao+ , (+ 1)2 0+2 0+ 1 _+2 V(x)= x - -- x - . It attains a minimum V =---- a at point 2 ao(o+2) o+l 2(o+2) x = X0=o Xa o = andX = becomes zero for x = xix = 0O and x = xx- =[ 2) a+ EquationE(8)i (8) in this case has the first integral ( ) 1 (x') 2 + V(x) = Eo (13) 2 where the constant Vo < Eo < 0 means the total oscillation energy. When Eo < 0, oscillations occur around the homothermlc background (x = Xo), when Eo decreases, their amplitude and period decrease. When Eo = Vo, there occurs a harmonic oscillation with an infinitely small

~~190 amplitude and period (A H =-y For Eo = 0 the oscillation is of the maximum possible amplitude (x = x 2) and period (A T = - 2a + ). In this case Eq (12) is integrated and gives~~

~~f() =xa+l () 2 (a + ) sin 2 a (14) a(a+2) A^T /1 where 0 is the integration constant. The dimensional period length is~~

~~2wr LT = a2a-+l /Ko/qo (15) CT~~

~~The solutions obtained may appear at the developed stage of formation of a homothermic burning instability, the dimensions of the regions of burning (Ar) being a spectrum of lengths'~~

~~Xc < - K/q < Ar < LT (15A)~~

A numerical estimate of the two periods of the solution (14) is presented in Fig. 3. Each layer of length LT burs m such a way that the temperature has the profile (14) and rises in accordance with expression (7). Here to = 0. If Eo > 0, there are no solutions satisfying the condition f" f' = 0 for f > 0

3 The above-mentioned regimes are asymptotic. When an arbitrary perturbation T(r, to) = To(r) at the finite instant t = to is specified, they become established at a stage where much more heat is being released than existed at the initial instant t = to. The process of release is distinguished by several features. First of all, Ta(O) -- 0 as to - - oo. The problem, which begins with infinitely small perturbations, is self-similar and is considered for time t E (- 0, tf). Equation (5) permits a shift in time, and in the self-similar problem the instant t = tf can be accommodated in t = 0. Then, t G (- oo, 0) [2-6]. In the case where the special (self-similar) initial data of expression (9) are specified, the instant tf is determined on the basis of expression (10). For the sake of definiteness, let us assume that to = 0. Then, tf ~ (qo TOa1 (0, 0)) - . If the initial data are not self-similar (do not have the form of expression (9)), tf can no longer be determined on the basis of expression (10). From dimensional considerations

tf = r [qoT or (16) where Tom = max To(r) and r is the function of the dimensionless parameters of the problem, r which are contained in the initial distribution To (r). Time tf is the sum of two quantities tf = tl + t 2, where tI is the time of formation of a profile close to the self-similar profile, Ta(r, ti), and t2 is the time left to this profile before focusing. Time t2 is determined on the basis of expression (10):

~~f-I 1 (0) t2 - T) (17) q0 TOt (0, tl)~~

4. Figures 4, 5 and 6 give the results of a numerical solution of the problem (5), (6) for Ko = 1, a = 2, qo = 1 and3 = (HS regime, Fig 4); 3= 3 (S regime, Fig. 5); =5 5 (LS regime, Fig. 6). In all three cases, the initial temperature perturbation given for t = 0 consists of linear

191 dependences of T on r, so that they form a triangle with amplitude Tom = 1 and base Aro = LT - AT /K o/qo0 5.44, which is a spatial period of the solution (14) for the S regime In the case of the HS and S regimes, the solutions are close to asymptotic in accordance with expression (7). In the case of the S regime, as calculations have shown, regardless of whether Ar0 < LT or Aro > LT the solution of the problem (5), (6) asymptotically approaches one penod of the solution (14) In this sense, LT is the "fundamental thermal length" of the S regime' burning occurs always over the "fundamental length", in a zone (of diameter LT) around the point of maximum temperature Tom. It takes the form of a burst Thus, if Aro < LT, there is first of all heat propagation. The region of burning grows until its diameter is LT. This instant corresponds to t = tI, i.e. to the establishment of a profile close to the self-similar solution (14). For t > tI, the rate of burning increases by several orders and an unusual burst of heat occurs (similar to a chain reaction [21], but only in the case of a non-linear medium). Numencal calculations show that, because of burning, in the peaking regime the release of heat in a structure occurs so rapidly that any temperature background (even a homothermlc one, albeit with a high tf value) is infinitely small m comparison with the temperature in the structure Consequently, as a result of the development of homothermic burning instability in the S regime, a structure with the greatest possible wavelength X = LT develops by virtue of its having the fastest law of growth (transfer of energy from perturbations with a smaller wavelength to those with a greater wavelength at the developed non-linear stage of instability formation) The development of such a structure is illustrated by Fig. 7. Thus, in the S regime, regardless of initial data, burning of the medium always occurs over the "fundamental length" LT, which is a function only of the following properties of the medium: a, Ko and q0. In the case of the LS regime, the T(r, t) profile inside the region of burning is close to the solution (7), (8) and undergoes substantial distortion at the boundary of the region. The region of burning is finite for, as in the case of the boundary problems in Ref. [4], the S regime - in which burning is localized over length LT - majorizes the LS regime If the excitation of burning is of the resonance type (i.e. tf t 2 and the burst begins immediately when t = to), the size of the localization region for the LS regime can easily be estimated on the basis of the "fundamental length" LT of the majonzing S regime. The resonance length has the form: T-- i ------Ta+l-1 L* 2-T f K0qo /2- (3+o++p1)1) R qK 0 To 2(18) (18) T CT * V qs u Q3P-l)

The equivalent qs of the majonzing S regime is determined on the basis of the condition that the temperatures and focusing lines tf of the majorizing S regime and the LS regime be equal. The size of the localization region in the LS regime depends not only on the properties of the medium but also on the maximum temperature Tom in the initial perturbation If Tom e- 0, then L - , and for the LS regime we can construct a solution on the basis of expression (8) with the heat wave front at infinity. The formula (18) is valid for Aro >LT, since in deriving it no allowance was made for the difference between tf and t2. If Aro < LI, then tf > t2 and the formula (18) should be corrected. The localization region (like tf) depends, in this case, also on the dimensionless parameters of the problem (rimarily on p = ro°

~~L(LS) = L* R (P, ) (1 9)~~

~~The following function may be proposed (again on majorant grounds) as the dimensionless function R:~~

~~192 R=[ r(, u ) ]m (20)~~

~~The validity of the formulas (19) and (20) is confirmed by a number of numerical solutions of the problem (5), (6) in the case of the LS regime for different values of /.~~

~~CONCLUSIONS~~

~~The principal conclusions which can be drawn from the results presented in this paper and in Refs [2-6] are as follows:~~

~~1. The action of the peaking regimes in a non-linear medium is metastably localized. This causes the development of structures if the medium contains volume sources~~

2. The development of a structure in the peaking regime is characterized by an instant tf when some quantity in the structure becomes infinite. At the asymptotic stage, the structure with the smallest instant of focusing develops in the medium. For example, in the case of the heat conductivity equation, depending on which regime (the boundary regime or the regime generated by a source) has the faster law of growth, the problem becomes either a boundary problem or a problem of burning.

3. The interaction of the growth of a quantity (heat, field, matter, etc.) in the peaking regime with its quasilinear diffusion determines the size of the localization region. In a compressible medium, particular sections of mass are localization regions. Each process has its "fundamental" length ( or mass). Let us consider, for example, the solution of the problem of compression of a finite plasma mass by a piston (0 pinch with a liner) [3]. The volume source of heat in the plasma is Joule heating. Figure 8 gives profiles of the dimensionless temperature 0 ~ T (tf-t)2/5 , the density 6 - p (tf - t)8 ' 5 and the axial magnetic field strength h - H (tf - t) along the dimensionless co-ordinate X - r (tf-t)-4/5. The example illustrates the existence in the plasma of structures of different types localized in the corresponding "fundamental" masses.

4. Studies of the special features of the development of structures lead to the conclusion that it is possible to superpose the solutions of non-linear structures. These solutions are characterized by definite scales (the dimension of the structure) and by a focusing time tf which depends on the initial perturbation amplitude Tm. Different initial data result in the formation of structures with correspondingly different instants of focusing. They can co-exist as a single structure, with a new instant of focusing, if there is a self-similar solution combining the entire set of initial data; this is permitted only when there is a definite discrete set of structure lengths and amplitudes Tmi.

5. The'actual physical processes (with allowance for gas dynamics, volume radiation, bounded- ness of the source, etc.) generally lead to replacement of the localized regimes (S and LS regimes) by propagating regimes and even by regimes without peaking. In a numerical or experimental study of a medium m which peaking regimes can occur, the features of such regimes should be taken into account, otherwise the stage of development of a localized peaking regime may go unnoticed. In particular, a special algorithm for selecting the time step was used in the numerical derivation of the solutions considered in the paper.

~~6. The phenomena considered here show that deep internal connections exist between the non-linear processes in a medium, decay of that medium into individual structures and the peculiar~~

193 thermodynamics of peaking regimes, which are accompanied by a complication of the organization of the medium and the appearance of a special "physics of a plasma with structures". The processes of transfer in such a medium, the conditions of initiation of the fusion reaction, stability and a number of other properties are fundamentally different. It may prove possible to use fine non- linear effects in the search for new approaches in controlled nuclear fusion. However, these phenomena are of great importance quite apart from that.

~~REFERENCES~~

[1] NUCKOLLS, J., WOOD, L., THIESSEN, A., ZIMMERMAN, G., Nature 239 (1972) 139. [2] ZMITRENKO, N.V., KUDRYUMOV, S.P., Dokl. Akad. Nauk SSSR 218 (1974) 1306,219 (1974) 578 [3] ZMITRENKO, N.V., KUDRYUMOV, S.P., "Avtomodel'nyj rezhim szhatiya konechnoj massy plasmy v zadachakh z- i 0-pincha" (The self-similar regime of compression of a finite plasma mass in z- and 0-pinch problems), preprint IPM (Institute of Apphed Mathematics) No.19, Moscow (1974), deposited with All- Union Inst. for Scientif. and Tech. Info. (VINITI), No. 3398-75 DEP. [4] SAMARSKIJ, A.A., ZMITRENKO, N.V., KUDRYUMOV, S.P., MIKHAJLOV, A.P., Dokl. Akad Nauk SSSR 223 (1975) 1344. [5] SAMARSKIJ, A.A., ZMITRENKO, N.V,, KUDRYUMOV, S.P., MIKHAJLOV, A.P, Dokl. Akad. Nauk SSR 227 (1976) 321. [6] KUDRYUMOV, S.P., "Lokalzatsiya tepla v nelinejnykh sredakh" (Heat localization in non-linear media), preprint IPM No. 39, Moscow (1976). [7] KERKIS, Yu.A., SOKOLOV, V.S., TRYNKINA, N A, FOMICHEV, V L., Dokl. Akad. Nauk SSSR 211 (1973) 69. [8] ZAKHAROV, A.K., KLAVDIEV, V.V., PIS'MENNYJ, V.D., ROTHART, L., SAENKO, V.B., STAROSTIN, A.N., YAN, G., Dokl. Akad. Nauk SSSR 212 (1973) 1092. [9] KVARTSKHAVA, I.F., MATVEEV, Yu.V., BUTOV, I.Ya., SAMARSKIJ, A.A., KUDRYUMOV, S.P., POPOV, Yu.P., "Rol'samoorganizatsij pinchevykh razryadov v nagreve i uderzhanu plazmy" (Importance of the self-organization of pinch discharges for plasma heating and confinement), in Plasma Physics and Controlled Nuclear Fusion Research (Proc. Conf. Tokyo, 1974) 3, IAEA, Vienna (1975) 149, Nucl. Fusion, Suppl.(1975) 175 [10] TIKHONOV, A.N., SAMARSKIJ, A.A., ZAKLYAZ'MINSKIJ, L.A., VOLOSEVICH, P.P., DEGTYAREV, L.M., KUDRYUMOV, S.P., POPOV, Yu.P., SOKOLOV, V.S., FAVORSKIJ, A.P., Dokl Akad. Nauk SSSR 173 (1967) 808. [11] VILENSKAYA, G.G., NEMCHINOV, I.V., Dokl. Akad. Nauk SSSR 186 (1969) 1048. [12] D'YACHENKO, V.F., IMSHENNIK, V.S., "K magmtogidrodinamicheskoj teori pmch-ehffekta v vysoko- temperaturnoj plotnoj plazme" (Magnetohydrodynamic theory of the pinch effect in a dense high-temperature plasma), m Voprosy teorii plazmy (Problems of Plasma Theory) 5,Atomizdat, Moscow (1967) 394. [13] SAMARSKIJ, A.A., DORODNITSYN, V.A., KUDRYUMOV, S.P., POPOV, Yu.P., Dokl. Akad. Nauk SSSR 216 (1974) 1254. [14] KOMAROV, N.N., KVARTSKHAVA, I.F., FADEEV, V.M., Nucl. Fusion 5 (1965) 192. [15] SOKOLOV, V.S., Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Tekh. Nauk No. 13 (1973) 86. [16] ZEL'DOVICH, Ya.B., KOMPANEETS, A.S., "K teorii rasprostranemya tepla pn teploprovodnosti, zavisyashchej ot temperatury" (Theory of heat propagation at a heat conductivity dependent on temperature), in Sbornik posvyashchennyj 70-letiyu akad. A.F. Ioffe (Volume commemorating the 70th birthday of Academician A.F. Ioffe), Izd. Akad. Nauk SSSR, Moscow (1950) 61. [17] BARENBLATT, G.I., Prikl. Mat. Mekh. 16 (1952) 67. [18] OLEJNIK, O.A., KALASHNIKOV, A.S., CHZHOU Yuj-lin', Izv. Akad. Nauk SSSR, Ser. Mat. 22 (1958) 667. [19] SAMARSKIJ, A.A., SOBOL', I.M., Zh. Vychisl. Mat. Mat. Fiz. 3 (1963) 702. [20] BASOV, N.G., KROKHIN, O.N., Zh. Ehksp. Teor. Fiz. 46 (1964) 171. [21] SEMENOV, N.N., "Tsepnye reaktsii" (Chain Reactions), ONTI publishing house, Leningrad (1934).

~~194 0<6) I 1I6 BaTT .- 4 CMo -1 I 1 I 1 1 I 1 1 iI 1 1 1 1~~

~~FIG 1 Law of the growth of the power of a laser pulse in a peaking regime.~~

~~195 T~~

~~I: t =-I,02 I0- I 2: t =-3,I I0- 2 3: t ,=-I,05 I0- 2 4: t -3 IO- 4 5: t =-2,4 IO- 5 6: t =-i - 6 20'~~

~~0 00~~

~~0.0 o0.2 0.4 0.6~~

~~To = 1.06, Ko = 0.5, tf = 0, to = - 0.25.~~

~~196 T~~

~~2 \~~

~~~~~I~~~~~~~~~~

~~O r 2 b A~~

~~FIG.3. Numerical estimate of two periods of the solution (14) for the problem (5), (9) Temperature profiles T(r, t) at different instants.~~

~~197 3- -PII +^ 11. 10SA1 II II ** ** ** ** C\**l lC?o * 1 I- -4 C CV) ed !u~~

*S As t:eI E0, 0) RF-

~~198 T~~

~~aJk- s 6 8 D1 12 V.~~

~~FIG 5. Profiles of temperature Tat different instants in the numerical estimate of localized burning (S regime).~~

~~199 T~~

~~t= 0,679 = 0.000~~

~~4 8 3D 12~~

~~FIG. 6. Numerically estimated example of the burning of a structure with decreasing effective width (LS regime).~~

~~200 T 6 " 0,4840338 7~~

~~IO5~~

~~IO'~~

~~I -8 -4 0 4 8~~

FIG. 7. Temperature profiles T(r, t) in a numerical solution of the problem of burningof a homothermic background; the development of instability led to the formation of a structure of thickness - LT.

~~201 I0.o~~

~~& A ti~~

~~20f 1 It i / hA k' t , \ \I I ,, 'v IO0 1 \ i ,i \~~

~~.1"\ ,1"\ %e I \A% I s _ _ iA 0.0 O.I \ 0 .2 0.31 .I~~

~~\\ i \ \~~

~~FIG.8. Thermal (temperature profile) and magnetic (field profile) structures in the self-similar problem of compression of a 0-pinch by a liner.~~

~~202 Laser Fusion Research in Osaka~~

~~C. Yamanaka, M. Yokoyama, S. Nakai T. Yamanaka, Y. Izawa, T. Sasaki, J. Mizui* M. Matoba, T. Mochizuki, Y. Kitagawa, K. Tanaka, N. Yamaguchi* H. Azechi and T. Norimatsu~~

~~Institute of Laser Engineering Osaka University~~

~~Institute of Plasma Physics* Nagoya University~~

~~Abstract~~

In the investigation of inertial confinement fusion the "qekko Project" uses a high power glass laser and the "Lekko ProJect" is an E-beam C02 laser system. The two different types of laser give a direct comparison of experiments at different wave lengths, pulse duration and power. The research program is directed towards the understanding of the coupling of laser light and plasma, the compression of target plasma by the implosion process and the optimum design and fabrication of target pellets.

~~1. Introduction~~

In Osaka we have investigated the inertial confinement fusion. There are two main laser projects, one of which is "Gekko project" using high power glass laser systems l ) and another is "Lekko pro- 2 ject" by E-beam CO 2 laser systems ). Beside these large projects for fusion research, we have several basic research schemes: experiment on interaction between laser light and plasma and REB experiment concerning the auxiliary heating of plasmas. The use of different types of laser gives a direct comparison of experiments done at different wave-lengths, pulse duration and power. The current problems of laser fusion research are to get fundamental understanding of (1) the coupling of laser light and plasma concerning various nonlinear phenomena and instabilities

203 (2) the compression process of target plasma by implosion (3) preferable pellet structure design and fabrication. To perform these experiments, the method of computer simulation and fine diagnostics using computer aids are very important.

~~2. Interaction Experiment~~

Laser plasma interaction has been extensively investigated using laser intensity range up to 10 1 6 W/cm2 for glass laser and 2 1014W/cm for CO 2 laser. Increasing the laser intensity, we can successively observe the appearence of the different absorption processes which have their own characteristics to be measured. (1) Gekko experiment ( glass laser ) Above the laser intensity 1013W/cm 2 , the electron of hydrogen plasma begins to have two components of velocity distrubution 3) due to the parametric decay instability. The act of this instability is endorsed by the observation of the satellite 4) spectral structure in Brillouin back scattered light in the red wing of SHG light. The isotopic shift of the satellite between D plasma and H plasma can be explained by the relation of parametric resonance. At this power level the specular reflectivity begins to decrease. And also the fast ions begin to appear. If we use longer laser pulse or front focusing condition to the target surface, the threshold laser intensity for induced backscattering is reduced. This is due to the formation of large density scale length in which the induced emission is rised. Above the laser intensity of 1014W/cm2, the backscattered laser light begins to have spectrum broadening 5) which is attributed to the temproal density modulation induced by the ponde- romotive force of swelling light field at the turning region of resonance absorption. The density modulation, swelling light field and the spectrum broadening have been computed by the ID code6 ) . The computed data indicate a similar behavior to the physical experiment. The results are shown in Fig. 1. The corresponding experimental results in hydrogen plasma are given in Fig. 2. As well known, linear conversion of E field is observed in the oblique incidence case. The back focusing condition of the beam to the target is necessary to indroduce the oblique incidence in lens collimation. As the density modulation is introduced, high energy jet stream of 50-lOOKeV is observed with

204 Be target in the angle of 10°-20 ° to the target normal, accompany- ing a strong magnetic field. At the laser intensity of above 1014 W/cm 2, the fraction of fast ions was 30 to 50%. The energy carried by these fast ions was about 70% of total kinetic energy. The experimental data is given in Fig. 3. The fast ion can be suppressed by using a laser beam with weak prepulse or also front focusing condition. When ananasecond pulse is used, the fast ion is also reduced to the amount of energy carried out as small as 20% of the total energy. Since dB/dt is proportional to VTxVn, the magnitude of magnetic field depends upon the spatial distribution of T and n. If the configuration of VT and Vn is given by (a) in Fig. 4, the B field and the ambipolar E field drive a plasma to the center line by E/B drift. This case induces the high energy ion jet stream in a narrow angle of cone. In the case of (b) of Fig. 4, ions will be emitted in a broad angle from the target. The latter case is introduced by the laser irradiation with prepulse or by the front focusing which produce the corona region in front of the target. In table 1, these irradiation conditions are summarized (2) Lekko experiment An E-beam controlled C02 laser is used. In the case of wavelength 10.6p, there are also three distinct stages according to the laser intensity. The first is the decrease of reflectivity from 50% to 18% in the case of polyethylene target at the laser power 101 0W/cm 2 . The ion velocity begins to have spread at the maximum of reflection. The experimental data are shown in Fig. 5. This threshold 1010 W/cm2 corresponds to the parametric decay instability. At the rear side of the target only the slow component of ion is observed. The second is the onset of the backscattered light ( eliminated the specular reflection ). The threshold laser power is 2xl010 W/cm2 for longer pulse (70nsec) and 3xl011 W/cm2 for shorter pulse (4nsec). This is due to convective nature of stimulated Brillouin scattering. The data is given in Fig. 6. The spectrum of backscattered light is shown in Fig. 7 which is measured by IR vidicon using pyroelectric image tube. At the laser power less than 1012W/cm2 , the blue shift is observed with two humps indicated by arrow A and B. The shift of B corresponds quite well to the doppler shift of plasma expansion (2x107cm/sec). The hump A fits to the Brillouin backscattering corresponding to

205 an acoustic frequency 1.8x 1 0 0Hz, ion sound speed 9xl0 6cm/sec which shifts to red from the blue shifted hump B. Above the laser intensity of 10 1 2 W/cm2 , other components indicated by C appear on both sides of the hump B symmetrically. For the long laser pulse, the general features are similar to the short pulse case, but one component C in red side is only observed. The third stage is marked by this spectrum spread shown by C which is due to the self- phase modulation of laser light caused by the nonlinear refractive index of plasma. The resonance absorption induces the strong swelling of the laser field near the turning point where the pon- deromotive force acts to repell the plasma. The frequency shift5) due to this process is given by

Aw-ko aF(t) 1 I In(x) dx at i^ne no(x ) dx where Ine2(col1/2 kTe+ Ti )e Ine is the laser intensity -p -1 e_ _ k ZTe n where the electron quivering energy is comprable to the mean thermal energy of the plasma. The frequency shift to red is 2.5xl01 0 Hz (80A) at X=10.6 pr n0=10-2, x=100lm, kTe= 3 00eV and I=1012W/cm2. This estimation agrees well with the experimental data. For the long pulse case there is no blue side C hump. This can be explained by the pulse wave form which has fast rise lnsec and slow fall of 70nsec. This result strictly corresponds the former report of glass laser case. In this third stage, self magnetic field is detected up to a few M Gauss. The ion beam is ejected in a cone of vertical angle 30 ° . The B field has a positive correlation to the ion beam intensity. (3) Interaction to plasma focus As a target, we use the plasma focus which is produced by Mather type 50KJ, 50KV, and 1.25MA machine. The focus is lcm in length,2mm in diameter. The neutron yield from deuterium plasma is 109 per shot in 80nsec detected by dysprocium foils. The laser is a double discharge TEA CO 2 laser of 6 modules in series, which can deliver 80J in 70nsec. The interaction experiment is designed to induce the turbulent state in plasma focus by the laser irradiation. As well known, the characteristics of plasma focus is shown in Fig. 8. The instability threshold is given by Pc-V 1 - 5, where Pc is critical pressure and V is applied voltage. The irradiation of laser light

~~206 is set to the marginal state or to weak focus state. Fig. 9 shows the introduction of instability and increase of neutron yield by laser. Apparent neutron spectrum are shown in Fig. 10~~

and Fig. 11. In summary when the CO 2 laser pulse of 70nsec-70J irradated to the focus pinch plasma of 120ns-300J the instability is induced to enhance the neutron yield up to twice of the yield of plasma focus.

~~3. Compression ExDeriment~~

Attainina the efficient compression to initiate a thermonuclear burn in a laser driven implosion requires that the ablation and compression processes maintain the spherical symmetry. To get the information of the ablated plasma and the compressed matter, two types of the experiments are performed. One is a physical simulation using a transparent gas filled cone. Another is a spherical compression experiment of various micro-balloon targets. The structure of pellet is yery important to achieve the spherically strong compression as well as the pulse shape of the laser beam. Glass micro-balloon tamped by nickel and gold, which is pressurized by the deuterium gas is testing. A solid deuterium micro-pellet coated by heavy metal is developing. Fig. 12 shows nickel coated glass micro-balloon whose diameter and thickness of nickel layer are 100 and lpm respectively. For these experiments, an energy analyzer of a particle, ion energy analyzer, multichannel x-ray spectrometer using Bragg crystal, X-ray microscope, X-ray pinhole camera, and X-ray streak camera have been prepared. Special phosphor-SIT camera system is developed to improve the detectable level of X-ray camera. The sensitivity of this system is 500 to 2000 times larger than that of X-ray film. Simultaneous measurement method of interferometer, Schrielen and shadowgraph has been also developed. The whole diagnostic apparatus, the target chamber and the laser system are commanded by the OKITAC 50/40 computer system. (1) Shock wave in solid target To study the behaviors of the shock wave in solid target and the energy delivered to the compressed matter and ablated plasma, a transparent solid slab target, lucite is illuminated by a 30 picosecond second harmonics of glass laser light to catch the laser produced shock wave. The shock wave and blast wave into solid and environmental gas are driven by the main laser beam.

207 Series of the schrielen photographs of the blast wave are shown in Fig. 13. Above the laser intensity of 5x10 1 3 W/cm2 , the blast wave in gas is accompanied with the second shock. This baw shock is driven by fast ion stream, as the threshold of the baw shock coin- cides to that of the appearence of the fast ion. The time history of the blast waves is shown in Fig. 14. The solid and broken lines represent the shock and sound waves in gas and solid matter respectively. We can estimate the energy delivered to the matter. They were 10 and 20% of the incident laser energy respectively. (2) Effect of the multi-layer target Films of polyethylene and nickel coated polyethylene are used as the laser irradiated target. The thickness of the polyethylene and coated nickel is 4pm and lpm, respectively. The laser beam ( 2 0 0 ps, 8J, focal diameter 20pm ) is irradiated to either side of the film. The specular reflection and the transmitted laser light from the plasma are measured. When the laser beam is irradiated from the polyethylene side the transmitted laser energy is almost constant at the laser intensity of 1014 1015W/cm2 . But when the laser beam is irradiated from nickel side the transmitted laser energy becomes maximum at 8x10 W/cm and the specular reflection becomes minimum as shown in Fig. 15. This effect may be related to transportation of light by photon trap- ping in the plasma. But at present time we cannot well understand the mechanism of this enhanced transmission of the laser light. The expansion velocity of plasma is measured by a shadowgraphic method using the second harmonic light of the mode locked laser light and streak camera. The power dependence of the expanding velocity of the plasma at the density of 1019cm 3 is shown in Fig. 16, where VA and VB are respectively the velocity to the laser side and to the rear side of the target. VA is always larger than VB. Both expansion velocities by the nickel coated target are larger than these of the no coated polyethylene target. And both expansion velocities by the double layer target irradiated from the nickel side are larger than these irradiated from polyethylene side at low laser intensity. Increasing the laser intensity, they become almost same values. As shown in Fig. 17 the expanding velosity of the plasma along the target surface irradiated from the polyethylene side

208 shows the largest value. This is caused by the small thermal conductivity in the target normal direction by the high Z plasma. The time of flight of ions measured at the rear side of the target shows the steep rise when the laser beam is irradiated from the polyethylene side of the double layer target as shwon in Fig. 18 (b). By these experimental results we could say that the double layer target as low Z material coated on the high Z material is effective to equalize the surface condition and to compress the target. (3) Micro-spherical pellet illumination Spherical compression experiments are investigating by two beam glass laser system "Gekko II". The maximum output energy of each beam is 200J~40J in 3ns-0.3ns. The laser beam is focused onto the glass and nickel coated glass micro-balloon through an aspheric lens whose focusing length is 10cm, F/l.l. The diameter of the micro-balloon is 100pm. The thickness of the glass wall and coated nickel are 2 and 1 to 0.5im, respectively. For the one beam irradiation experiments a polyethylene backuped glass micro-balloon is used. The slow rise laser pulse is rather effective to spherical compression as shown in Fig. 19. The compression ratio of the one beam method is 70 at 80J in 3ns. When the focused laser beam concentrates to the small portion of the pellet, micro hot spots are observed in the X-ray pinhole picture as shown in Fig. 20. These micro spots are caused by the strong nonlinear coupling of laser beam and plasma. In high intensity region modulational instability produces cavitons and accelerated electrons in the cavity wall can emitt the strong high energy X-ray.

~~4. Conclusion~~

The high power laser technology has recently developed to contemplate the implosin experiment. However the fabrication of target and the characteristics of multi-layer pellet have just come to be investigated. Fundamental research on laser-plasma interaction still have many interesting problems to be solved. To solve the fusion problems, international cooperation to attack the goal is very important.

~~209 REFERENCES~~

~~1) K. Yoshida et al: Technol. Rep. Osaka Univ. 26, 127 (1976)~~

~~2) M. Matoba et al: ibid, 139 (1976)~~

~~3) C. Yamanaka et al: Phys. Rev. A6, 2335 (1972)~~

~~4) C. Yamanaka et al: Phys. Rev. Lett. 32, 1038 (1974)~~

~~5) C. Yamanaka et al: Phys. Rev. All, 2138 (1975)~~

~~6) C. Yamanaka et al: IQEC K-10 (1976), Opt.Comm. 18, 104 (1976)~~

~~7) Y. Kitagawa et al: to be published in J. Phys. Soc. Japan~~

~~210 O~~

~~LL Ii~~

~~0 . u0 ils~~

~~0 i -I U -1 H~~

~~-oJ1~~

~~o o r i 11^ i V^~~

~~O'L o 01. 001. 0o P4~~

~~-I4 CC I I~~

~~' d~1~~~~

~~Q)a,~~ < ~^,d~~~~~~ f ._4._40-Ioi3t 04-i o (U / 6 Nt.z) N 1)L \ q *) g d t~~

~~u 0d~~

~~v . U)~~~U 4r4 0 I ! 1 - *H tn~~

~~^ /~~

~~(III ' 9 L/ x e I x) N s O A Nk(^)NtN £ SDI/fer I~tL/l 211 - - ' ~ - .Zi -I Scattered light lo = 3 1x 1d 1 180O /ghm2~~

~~135 ° 1x1014~~

~~Incident li ght~~

~~Io I I t I t I I lOA/div.ldiv 1.06421642 pmm~~

~~Fig. 2 Spectral broadening of back scattered light by self phase~~

~~modulation.~~

~~212 Lu 0~~

~~a e 0 r-~~

~~: I ---1 a) oZ -O 0 ,4~~

-1 >j: 0 rd 2 0 >t_ .r- o 0 0 0 cU 0 - 1i CO1f1 n1 i 1 ow> E,,, I,,, , " .1 a, ill 1S~ tA IJ HI / / / / :>, - / 1 17 r, 1 1/ i U) 0o ·- 0 ...... 17_) < F-1 z -- i 1 1 j H 0 I Ii Ii . LL_I1_ aJ 1-1 4h ; t E t 0

~~E e o ,,- z ) 0 o H U) CO *J~~

~~I '~~

;

~~U) I~~

~~I (D~~

~~1 11 eI z~~

~~0 0 T v--: CI~~

~~213 COLLIMATION OF SUPER THERMAL ION~~

~~T E IE GTE - .. - 1 \a - r- LLaser -K Laser~~

~~IB (a) (b)~~

~~HIGH FREQUENCY OSCILLATION (ELECTROMAGNETIC MODE) DEPEND UPON HIGH ENERGY ELECTRON CURRENT~~

~~1.CURRENT PINCH 2.L-WAVE T-WAVE MODE CONVERTION 3.SPACE AVERAGE EFFECT OF MICRO SCALE MAGNETIC FIELD PRODUCED AT FOCAL POINT~~

~~Fig. 4 Schematic explanation of production of superthermal ion~~

~~stream.~~

~~214 U a10 0 50 3, 0 E " C-) 40~~

~~-0 0 x 30 (A -b-, a 5 U 0 20 / O' 0 -1 10 u 0 0 I) 10~~

~~.,I, Oxo t . . . I ... 100 1010 101 Incident laser intensity ( w/cm )~~

~~Fig. 5 Reflectivity and ion velocity spread due to laser pulse.~~

~~215 aI 70 ns o 4 ns~~

~~i =500 I 2 0_1 in In~~

~~/ I/ £t- i 10 0 0 U '~~

~~in\ 0 M3 i 0~~

~~. I . . . i I I , I .... 102 (MW) I! e . I . . l . . . | . a . . . ,.. . . . · . 1 . . 101 1011 101m Incident (W/cm2)~~

~~Fig. 6 Power dependence of Brillouin back scattering for long and~~

~~short laser pulse.~~

~~216 Ir i~~

~~1 3x10'T :23£0t2.3 x` 11~~

~~9.5x101~~

~~9,', M ~ t^ 2 ' '' '' * 4 'a W/cm t e Ci -;de I d e .d~~

~~1 *^.(1 ,; i· '**-~'X'i '>E, "i" yE 1 ,$ ta,0 ,ii ?W j Ii~~

~~(4 ns .,: ti-thty, ( 7O nos)~~

~~Fig. 7 Spectrum of back scattered light.~~

~~217 8~~

~~6 /H 6 /OH 2~~

~~oo ~~~

~~a, ofl * 1 L. -1::: 2 EL0~~~~~~o/ ._-"He~~

~~O Hf - - I- ,-,i--- i - 16 18 20 22 24 26 Applied Voltage ( kV)~~

~~Fig. 8 Typical characteristics of plasma focus.~~

~~218 CO2 LASER IRRADIATION AT AND AFTER Focus PINCH 20 KV, 5,5 TORR, C02 LASER 60J~~

~~AT Focus PINCH AFTER FOCUS PINCH~~

~~CURRENT CURRENT~~

~~LASER LASER~~

~~Y-RAY NEUTRON Y-RAY NEUTRON~~

~~13 M FROM FOCUS,200 NSEC/DIV. 13 M FROM FOCUS, 200 NSEC/DIV.~~

~~Fig. 9 Neutron yield and current behaviour of focus plasma irradiated~~

~~by CO 2 laser.~~

~~219 1 I~~

~~with Laser /I EL=0.4MeV - '( ;L('T L=4keV)~~

~~c5 / o/!A0 A AEo=0.28M( (T IL-"' :01- / o=1.9key -o - / \ \0 z - / / |\ \ without Itr /t~-~ / ;~\- Laser~~

~~0.01. i / 1 2 3 E (MeV)~~

~~Fig. 10 Energy spectrum of neutrons at 20 kV-8 kJ plasma focus~~

~~irradiated by 80 J-70 ns CO 2 laser.~~

~~220 1~~

~~-- 0~~

~~~-QJ H~~

~~U Co N1 000~~

~~t~ ! a!1nI~ 1Ist ~ rCI~~~~~~~~.4, Y.- ,g-.L~~

~~(U3)1 1IH±D3dS NOWINO3N~~

~~221 cC . 0 o 0 t~o m0 A~~

~~^0) U~~

~~a) O oX~~

~~Cr CA0 -)~~

~~(t .) -r-I~~

~~tO ' 0(n '~~

~~222 Target Air 0 200m 0 200p m~~

~~Fig. 13 Schrielen photograph of the blast wave in solid and air.~~

~~223 -- Laser Target Air~~

~~100 0 100 200 rs (Pm)~~

~~Fig. 14 Time history of the blast wave.~~

~~224 -b-i~~

~~n- I.-~~

~~c50 0~~

~~, - 1 Laser Ni(CH '== Ni+(CH2)n U)-0~~

~~I I I I I Co 2 4 6 8 10 Laser Intensity x1015 (W/cm 2 )~~

~~Fig. 15 Power dependence of the transmitted laser light when the laser beam is irradiated from the nickel side of the double~~

~~layer film. Thickness of polyethylene and nickel is 4 pm~~

~~and 1 pm, respectively.~~

~~225 8 0 () E VA -. p- VA o x'5 'Xx66~~

~~O Ni+ (CH2)n c6 (CH2)n~~

~~IS (CH2)n{ (0 +Ni LLU~~

~~2 L) Laser VB~~

~~I I I I I _1 0 2 4 6 8 10 12 Laser Intensity x1015 (W/cm 2) Fig. 16 The power dependence of the expanding velocity of the~~

~~plasma.VA and VB are the velocity to laser side and to~~

~~rear side of the target. CCH 2 )n +Ni and Ni +(CH 2 )n~~

~~mean that laser beam is irradiated from (CH2 )n and Ni~~

~~226226 side, respectively. (CH2)n+ Ni~~

~~8 (CH2)n~~

~~E U +(CH2)n u 06~~

~~Vr U~~

~~0o Laser >4 0) C -n~~

~~0 2 4 6 8 10 12 Laser Intensity x 1015 (W/cm 2 )~~

~~Fig. 17 Expansion velocity of plasma along target surface.~~

~~227 CN E n O u : Cf) C C O O I ° 0 0 (N X CN LO O0~~

~~0 u-I 0 0 0 .o4~~

~~4-4 C) r-1 r4~~

~~IH0 U 0 a)~~

~~U) 4J (N n u-M o to0 CM X (0~~

~~LL_ -. u~~

~~U cm '4~~

~~m .l icr *>- 0Hr *,1 o fU~~

~~U c:~~

~~( O Co . X Z tI~~

~~228 One Beam Two Beam~~

~~Fig. 19 X-ray pinhole photographs of pellet irradiation.~~

~~(a): One beam irradiation to glass micro-balloon back-~~

~~uped by thin films.~~

~~(b): Two beam irradiation With nilkfle coa.~~

~~229 4-~~

~~Fig. 20 X-ray pinhole photograph where micro hot spots appears.~~

~~230 Table 1 Relation between irradiation condition and plasma heating~~

~~OUT FOCUS IN FOCUS~~

~~Ps ¼ 11__s-11 <~~

~~DECAY RESONANCE INSTABILITY ABSORPTION~~

~~ION ENERGY El E 2 SPREAD AE 1 AE2~~

~~HEATING EFFICIENCY OF MAIN BODY '1 '2 v~~

~~PRECURSOR PLASMA SUPER THERMAL ION SLOW RISE LASER PULSE SELF-PHASE MODULATION PREPULSE~~

~~231 A HIGH-POWER LASER SYSTEM FOR THERMONUCLEAR FUSION EXPERIMENTS~~

Eh.A. AZIZOV, L.P. IGNAT'EV, N.G. KOVAL'SKIJ, Yu.A. KOLESNIKOV, A.F. MAMZER, M.I. PERGAMENT, Yu.P. RUDNITSKIJ, G.V. SMIRNOV, V.A. YAGNOV, V.G. NIKOLAEVSKIJ I.V. Kurchatov Institute of Atomic Energy, Moscow, USSR

~~Abstract~~

A HIGH-POWER LASER SYSTEM FOR THERMONUCLEAR FUSION EXPERIMENTS. A high-power laser system has been designed for an energy output of - 3 X 104 J. Neodymium glass was selected based on the level of technical progress, operating experience and the availability of components. The operating performance that has been achieved to date is described.

As shown by calculations intended to demonstrate the physical feasibility of controlled thermonuclear fusion initiated by a laser beam, a laser pulse with an energy of 104- 105 J lasting between 1 and 10 nsec is required. Furthermore, a number of ideas regarding the physical processes and effects used in numerical calculations of the compression and heating of deuterium- tritium targets by a laser beam need serious experimental verification that is possible only over this energy range. Hence the neodymium glass laser system designed by us is calculated for an energy of ~ 3 X 104 J. In the design fairly rigid demands are made upon the shape of the laser beam, construction of the target and irradiation symmetry. The design provides for the possibility of doubling the output energy by increasing the number of high-power amplification stages or by a transition to active elements with better characteristics. Our selection of a neodymium glass laser system was based on the level of technical progress attained, extensive experience gained in operating neodymium lasers and the industrial availability of virtually all the component parts. Given these circumstances, it is possible to build a laser system with the required parameters in a very short time. As we know, the total surface area of the active elements in the output stages of a laser system is determined by the ratio between the total energy and'the maximum permissible coherent emission energy density for glass. Experience gained in operating the "Mishen'-2" device shows that non-linear effects limit the maximum energy density to 4-5 J/cm2. Consequently, to obtain 3 X 104 J the output aperture of the laser system should have an area of not less than 6 X 103 cm2. The high-power amplifier stages of the designed laser system use active elements of rectangular section 40 X 240 X 750 mm in size, made of LGS-247-2 and LGS-I glass. Active elements of this type of design appear to be optimal for high-power lasers in the nanosecond range. The active element is pumped by two plane pumping devices with 18 IFP-8000/I lamps in each. The distribution of the inverse population along the smaller side of the active element perpendicular to the plane of the pumping devices can be satisfactorily approximated by the equation N(x) = No I - 5 where x varies between 0 and 4. For a pumping energy of equation N(x) = N0 [(x- 2)2], i41]~,,,,,,1 300 J the inversion at the centre of the active element attains 0.3 J/cm2 , while the weak signal amplification factor is ~ 5. The temperature differential between the lateral surface and the central section of the active element during maximum inversion does not theoretically exceed 0.5°C, while the thermo-optical distortions of the wave front are approximately 0.1 X. Interfero- metric measurements fully confirm the theoretical results.

233 Figures 1 and 2 show in diagramatic form the optical system of the laser and the layout of the principal units in the experimental set-up. As can be seen from the figures, the laser system consists of a dnver oscillator coupled to an alignment oscillator, a set of preamplifying stages and high-power amplifymg stages with active elements of reactangular section joined in series- parallel. The magneto-optical Faraday rotator and passive dye cells placed between the pre- amplifications stages step up the energy contrast and protect the optical cells from radiation reflected by the target. To ensure the required coherent emission energy it was decided that the number of high-power output stages would be 64 (the total number of high-power amplification stages being 102). The pumping lamps are fed from an inductive storage unit with a total energy capacity of 100 MJ. A diagrammatic representation showing how the electrical system for pumping the final amplification stages functions is given in Fig. 3. As can be seen, the feed system consists of a rotating wheel generator, sectioned magnetic energy storage units and commutatmg devices intended to eject the energy and shape the required current pulse in the pumping system. In the layout selected a separate symmetrical load with a mid-point connected to the common zero busbar is fed to each of the N storage unit sections. When N + 1 circuit breakers have been triggered, N independent circuits with a current Io feeding the independent loads are formed. This feed system makes it possible, first, to halve the voltage at the pumping system elements, second, to make it simpler to attach a large number of load channels separated in space, and, third, to make autonomous the work of each load, which then, for practical purposes, does not affect the work of the remainder. Irrespective of the way in which the pumping lamps in the load sections and initial current 10 are connected, the energy system ensures a supply of - 300 J to each of the pumping devices over a period from 5 X 10-4 to 1 X 10-3 sec. The driver YAG oscillator and the light pulse shaping and programming system, which consists of two Pockels cells, are such that up to 0.1 J can be obtained in one transverse mode for a pulse duration of 1-3 nsec and energy contrast of not less than 106. To align the system and focus the laser beams onto the target use is made of a continuous YAG oscillator, which emits 10 W in one transverse mode. The optical axes of both generators coincide with an accuracy of 10". The set of preamplifying stages with cylindrical active elements, in which the diameter of the laser beam is gradually increased by means of Galileo telescopes combined with spatial filters from 3-4 mm to 20 mm, 45 mm and 260 mm, terminates with an amplifier with an active element 40 X 240 X 750 mm in size. At the present time the following laser pulse parameters have been obtained at the output from this system:

~~Energy 300-400 J Beam size 40 X 240 mm Divergence 10 - 4 rad Energy contrast > 10s Pulse length 3 nsec~~

The passage of the laser beam through the high-power amplification stages for the system of spatial separation of the light beams shown in Fig. 1 was calculated with a computer. Use was made of an equation for a travelling wave amplifier [ 1 ] m which the dependence of the induced emission section on the energy density and the non-linear losses were taken into account. The former effect is due to non uniform broadening of the spectral bands for neodymium, and the latter is caused by losses due to Raman and Brillouin scattenng and two-photon absorption. Numerical calculations were also made for the large-scale distortions caused by non-uniform distribution of the intensity across the beam section, and by non-linearity of the refractive = - 13 index of glass (the non-linearity factor for silica glasses is n2 2 X 10 CGSE [2].

234 Fig. 4 shows the energy density distribution over the laser beam section at the output from the final amplifying stage for uniform distribution of the intensity after the preamplification system. The same diagram shows the corresponding wave front distortions due to non-lineanty of the refractive index. A noted improvement in the quality of output beams can be attained by feeding into the high-power amplification system a beam with a specially profiled energy density distribution. Fig. 5 shows the energy density profile for which a plane wave front and fairly uniform intensity distribution are obtained at the output It is intended to make the corresponding profiling filters by tinting the glass with the aid of an electronic beam or gamma rays. To suppress minor wave front distortions we shall use spatial filters. The divergence of the laser emission, which is a function of the considered effects and aberrational distortions, should not theoretically exceed 5 X l0 -4 rad.

~~REFERENCES~~

[ 1] MIKAEHLYAN, A.L., TER-MIKAEHLYAN, M.L., TURKOV, Yu.G., Solid-stage optical generators (in Russian), Sov. Radio, Moscow (1967) 337. [2] AKHMANOV, S.A., SUKHORUKOV, A.P., KHOKHLOV, R.V., Usp. Fiz. Nauk 93 (1967) 19.

~~235 0~~

~~4.4~~

~~4- '4 0~~

~~'4 4.4 '-4 '4~~

~~0. 0 '4~~

~~0 0 '4~~

~~'4 4.4 '4 4.- '4 4.-~~

~~236 &.-L~~

~~FIG.2. Experimental set-up~~

~~237 'I;~~

~~II I t~~

~~41 ~~~~~~~~~~a UBj PSii"1~~

~~~~~h j BS~ ts E se<' |l^1I j^^ S~~~~

~~~~"^tJ1 t'^ 5 ^~~~~

~~~~1.f~~SS ^18 ^^~~~~

~~~~238 i7 I I I * S /~~~~

~~~~r I3 82 I /y 11~~~~

~~~~o Q5 I 4r, 2 PaccmoRnue (clJ~~~~

FIG.4. (a) Energy density distribution over section of light beam at laser output, (b) Difference in optical path lengths in the high-power amplifying stages due to non-linearity of refractive index. A light beam of uniform intensity is fed to the input of the high-power amplifiers

~~~~a I~~~~

~~~~%14~~~~

~~~~0 4 2 CM~~~~

~~~~FIG.5. Optimum intensity profile of laser beam fed to input of high-power amplifiers.~~~~

~~~~239 STUDY OF THE ACCELERATION OF THIN METAL FOILS ACTED ON BY HIGH-POWER LASER EMISSION~~~~

V.N. BELOUSOV, V.L. BORZENKO, I.N. BURDONSKIJ, E.V. ZHUZHAKALO, N.G. KOVAL'SKIJ, A.N. KOLOMIJSKIJ, A.A. MALYUTIN, Yu.K. NIZIENKO, P.P. PASHININ, M.I. PERGAMENT, A.I. YAROSLAVSKIJ, V.N. KONDRASHOV, V.V. GAVRILOV I.V. Kurchatov Institute of Atomic Energy and Lebedev Institute of Physics, Moscow, USSR

~~~~Abstract~~~~

STUDY OF THE ACCELERATION OF THIN METAL FOILS ACTED ON BY HIGH-POWER LASER EMISSION. The experimental study of the acceleration of metal foils when irradiated by laser pulses provides information on the mechanisms of absorption and thermal conduction of plasmas and targets in the solid phase. The simplest and most convenient geometry for diagnostics is plane geometry which has been used in experiments with the Mishen'-2 device. The experimental results are compared with theory.

The results of numerous calculations made recently in the USSR and the United States [1 ] show that the highest energy breeding ratios due to the occurrence of nuclear fusion reactions are attained when spherical shell targets are irradiated by laser pulses. To optimize the compression process and to attain ignition and highly efficient combustion of the thermonuclear fuel, we vary, in the computer calculations, the geometrical dimensions of the shells, which play the part of a compressing piston, together with their composition and the ratio between the shell (piston) mass and the deuterium-tritium mixture filling the shell. Within the limits of the model adopted for the calculations, the laser emission is absorbed in a thin layer of plasma corona surrounding the target, close to the critical density points. Energy transfer from the absorption zone to the surface of the solid target is effected by electron thermal conductivity, while the acceleration and heat-up of the shell material is due to the propagation through it of thermal and shock waves excited during ablation of material from the surface. Present ideas as to the mechanisms governing absorption and thermal conductivity in the plasma and the target solid phase, which are based on a variety of models, can hardly be considered definitive. It is just for this reason that there must be experimental study of the acceleration of metal foils of differing thickness when acted on by laser pulses with a view to simulating actual processes and comparing data obtained with the theoretical computations. The simplest and most convenient geometry for diagnostics is plane geometry. Our studies of the acceleration of thin metal foils when irradiated by high-power laser pulses were carried out in the Mishen'-2 device [2]. As can be seen from Fig. 1, the laser system in the Mishen'-2 with active elements made of neodymium glass consists of an oscillator, a system for shaping a brief pulse in Pockels cells, preamplification stages and final high-power amplification stages with active elements of rectangular cross-section (4 X 24 cm 2 ). To suppress super- luminescence and to protect against light reflected by the target, use is made of passive dye cells and Faraday (magneto-optic) rotators. The laser emission energy at the output of each of the three parallel final stages attains 300-350 J in a pulse lasting 3.0 nsec. The divergence of the light beams is not more than 2 X 10- 4 rad; the energy contrast in the system is not less than 10 s .

241 The light emission was focused on the surface of a thin foil placed in the middle of a vacuum chamber, using a lens with a relative aperture of 1:5 (f - 170 cm). Monitoring of the engineenng parameters of the device, control of the charging of the capacitor bank in the optical pumping lamp circuits (total capacity of bank ~ 1.5 MJ) and testing to ensure that the diagnostic equipment is ready for action are all carried out with a M-6000 computer. To study the properties of the plasma formed, to record the laser emission parameters and to measure the speed and pulse of the accelerated part of the foil we used a broad range of diagnostic techniques. The distribution of the emission sources in the visible and X-ray spectral regions was determined by means of pinhole cameras. Fast-action cameras employing image converter tubes enabled us to observe the dispersion of the plasma both by recording its natural luminescence and by the shadow method, in which for brightening purposes we used part of the main laser beam converted into the second harmonic (5300 A). The electron temperature of the plasma was determined by the standard absorption method. We also used electrostatic analysers for the charge particle energy and studied the spectrum of the plasma-scattered light. In preliminary tests we irradiated aluminium foils 20-100 pum thick with pulses from one of the Mishen'-2 laser channels. A typical shadow photoscan showing the dispersion of the plasma jet and acceleration of the foil material is given in Fig. 2. The same figure shows the - 3 measurement system. Movement of the plasma jet with a density of ' 1021 cm and boundary of the accelerated foil are shown in Fig. 3 (the experimental data are represented by solid lines and the calculation results by broken lines).

~~~~REFERENCES~~~~

[1] CLARKE, J.S., FISHER, H.N., MASON, R.J., Phys. Rev. Lett. 70, 89, 1973. k2 ] ALEXANDROV, V.V., et al., Plasma Phys. and Controlled Nucl. Fusion, 1974, Proc. Int. Conf. Tokyo, 1974, Vol. 2, Session VIII.

~~~~CXEMA YCTAlObKU Muleub ]~~~~

~~~~242 FTi<. 2-~~~~

~~~~\ / I v J f c / T ) f ^//fMa~~~~

~~~~oS10 P y'i' f0 [cm]~~~~

~~~~243~~~~