<<

Energy and Technology Review ~awrence Livermore National Laboratory January 1985 The Novette Facility: A Step in the Evolution of High-Power Laser Systems

Work with the recently dismantled , LLNL's flexible, high-energy-density experimental facility, has profoundly affected our understanding of laser- coupling phenomena. Novette was the first in the Laboratory's series of successively more powerful and complex laser systems to incorporate full-power harmonic conversion of laser light in nonlinear birefringent media, and with it we probed the details of the design for the laser.

For further information contact In the course of inertial-confinement system in the evolving series of Kenneth R. Manes (415) 422-0681. fusion (ICF) research at LLNL, we have Laboratory was designed to exploit designed and built a series of the knowledge gained through successively more powerful and complex experiments with its predecessor. laser systems: Janus, Cyclops, Argus, The Novette laser was the first to Shiva, Novette, and soon the next incorporate full-power , Nova. All of these laser upconversion in order to test the systems used or will use chains of hypothesis that short-wavelength laser - amplifiers (in the form pulses couple energy more efficiently to of scaled modules), and they all share target plasmas than longer ones do. the same fundamental design, called Thus, the Novette laser provided MOPA-a Master Oscillator driving a a flexible, high-energy-density single-pass Power Amplifier. Each experimental facility to bridge the gap

10 DEFENSE PROGRAMS

between the Shiva and Nova lasers, theoretical and experimental studies have while allowing us to probe the details of indicated that coupling and, therefore, the Nova design. In the late summer of target performance would improve 1984, its work completed, the Novette significantly when the wavelength of the laser was dismantled and its amplifier incident laser light was shortened below chains were moved into the Nova laser 1 pm. Accordingly, the Nova laser is bay. designed for the harmonic conversion of In inertial confinement fusion, tiny but the fundamental laser pulse to powerful thermonuclear explosions are green or to near- light before produced by focusing a laser beam on the beam is focused onto the target. microscopic (0.1 to 5.0 mm) ­ The Novette laser produced these (D-T) targets. Because the laser wavelengths and also, for its final ICF beam is very intense (102 to 104 TW/cm2), experimental series, the fourth-harmonic it generates a dense plasma where it ultraviolet light (0.263 pm). impinges on the target material. This plasma interacts strongly with the laser ovette Design beam. Each of the Novette laser's two As recently as 1980, opinion was still N beam lines was optically divided as to the relative effect veness of similar to a Nova laser arm, the different lasers (and laser wavelengths) emerging beams having diameters of as fusion drivers. Most evidence now 74 cm (see Fig. 2) . Harmonic conversion favors a green (0.526-pm) to near­ took place in two unique mosaic arrays ultraviolet (0.351-pm) laser. of type II potassium dihydrogen The Novette laser fulfilled the need phosphate (KDP) crystals. The two for a short-wavelength source for approximately 4.5-TW green (2OJ) pulses multikilojoule target experiments. We were concentrated onto targets by two combined components from the earlier 74-cm-aperture £/4 doublet lenses. , until Novette the world's Because the two pulses arrived at the most powerful, with borrowed parts target within 5 ps of each other, a typical delivered for the Nova laser (our target could be irradiated simultaneously Fig. 1 100-TW -class laser system, which from two sides by approximately 9 kJ in reached its construction completion 1 ns. Artist's rendering of the Novette target area showing the refurbished Shiva tar­ milestone on December 19, 1984, by We devoted about half of the Novette get chamber and the final frequency­ firing all ten beams, frequency converted laser's experimental time to plasma conversion crystals and focusing-lens to 0.351 pm, into the largest chamber). physics studies, whose aim was to assemblies. We assembled the Novette laser on an accelerated schedule in an existing building adjacent to the new Nova laboratory. Construction began in January 1982, and the system was activated in 13 months. The facility included a Nova-style control system, the refurbished Shiva target chamber, and a full suite of target diagnostics (Fig. 1). The Novette laser produced 18-kJ, I-ns, infrared (1 OJ) pulses that were frequency doubled to green light (2 OJ) and focused onto targets. During 1983 and 1984, we gathered a large body of data on the performance of the Novette laser as a system. In this article, we compare calculations with measurements taken at several locations within the laser chains. Our work with the Novette laser has profoundly affected our understanding of laser-plasma coupling phenomena. Both

11 develop a better understanding of short­ energies of up to 1 J each. At the master wavelength, laser-plasma interaction oscillator, the laser beam was spatially phenomena relevant to inertial Gaussian with a diameter of 1.5 mm confinement fusion. The balance of the at 10% of its maximum intensity. As experimental time was divided between the beam proceeded through the high-density implosion research and the preamplifier, it was expanded and passed very successful laboratory x-ray laser through an aperture so that only the experiments. Finally, laser data were smoothest central region exited the obtained to help define the Nova laser preamplifier and was presented to each configuration. of the arms. Figure 2 is an optical schematic that In each arm, seven beam-expanding shows the Novette laser's relation to spatial filters separated stages consisting other LLNL MOPA-system architectures. of rod and disk amplifiers with We have gradually improved the increasingly larger apertures. The peak fundamental design by using output flux from each stage was components that were increasingly chosen to be near its operating limit. reliable, efficient, damage resistant, Propagation was controlled by these and cost-effective. Along the way, we spatial filters to minimize local intensity have learned a great deal about the fluctuations and to keep the maximum propagation of high-power laser pulses laser flux below optical-damage limits. in the dielectric media typically found By maximizing the fill factor, or in neodymium-glass amplifier chains. efficiency, with which each amplifier's We based the design of the Novette aperture was used, we minimized test bed on the requirements of the Nova the peak laser flux that each optical laser. Because the output of a repeatable, component had to tolerate to produce high-quality master oscillator is typically a given output. about 10 - 4 J, the Nova power amplifier The principal design constraint on 9 must provide a of 10 . In the the maximum energy delivered by each Novette laser, therefore, the weak pulse stage is the damage threshold of its produced by the master oscillator first optical components, which depends, passed through a preamplifier consisting in tum, on the purity of the optical of crystalline and glass-rod amplifiers, medium, the optical surface, the pulse beam-expanding spatial filters, and duration, and the wavelength of the laser beam-splitting elements, to produce two light. To raise the damage threshold in parallel, spatially uniform beams with the Novette laser, components such as

Master osc:IIator 3.75 9.4 9.4 Janus size

Argus, Cyclops, Shiva size

Nova, 74 cm Novette size 31.5-cm disk amplifiers

!2;] n-cm aperture amplifier

Fig. 2 t>

12 DEFENSE PROGRAMS spatial filter lenses, which in previous array, while the lower band represents laser systems were antireflection-coated an estimate of the highest (damage with dielectric films, were replaced with limited) 0.53-.um on-target energy versus optics treated by a neutral solution laser pulse duration. process.] Such optics not only display Our designs for each amplifier stage excellent transmission (typically greater of the laser were based on damage risk than 99.5%) but also have surface­ assessments made using MALAPROP damage thresholds of 12 to 16 J/cm2 for computations. The components most I-ns pulses, approaching those of the vulnerable to laser-induced damage were polished glass substrate. We tested this the input lenses to the last two spatial new technology on the large-diameter filters . Neutral-solution, antireflection spatial-filter lenses, windows, and target­ coatings on these optics have survived focusing lenses of the Novette laser in unharmed after being subjected to more preparation for the Nova laser. than 100 shots of I-ns duration with A major design goal for both the average fluxes of 5 to 7 J/cm 2 and peak Novette and Nova lasers was to fluxes twice as large. Indeed, bulk maximize the output energy without damage as a result of microscopic damaging optics. We would like to metallic inclusions has set the practical propagate an ideal beam that exactly fills flux limit for these lenses. the power amplifier with an output flux just below the damage limit. However, armonie Conversion such a beam does not propagate for any The conversion of laser light useful distance without diffraction of the H into its harmonics in nonlinear beam edges, which leads to constructive birefringent media is a well-known interference and regions of excessive process that was first demonstrated more optical flux. The first spatial filter in each than 20 years ago. of the Novette arms had a 2-mm-diam Because shorter wavelengths had been pinhole in its focal plane and expanded shown to couple more effectively to the beam from its input diameter of plasmas, considerable interest arose in 2.6 to 3.75 cm. The beam profile that using this up conversion (harmonic) emerged was nearly uniform, except for process to generate short-wavelength six faint rings. Succeeding spatial filters beams to irradiate fusion targets. During re-imaged, or relayed, this beam onto the the last three years, efficient conversion entrance lenses of the following spatial of l-.um light to its second, third, and filters and eventually onto the target­ fourth harmonics in KDP for beams chamber focusing lens. The net effect of up to lO-cm in diameter has been this strategy was to achieve a measured fill-factor of 85%. This is substantially greater than the 70% or less of earlier . lasers. The MALAPROP computer code (see box on p. 18) has proven invaluable for estimating peak fluxes impinging on optical components.2 We used MALAPROP to calculate the electric field distribution experienced by each component during a hypothetical shot. We introduced realistic dust and damage -:::l .e-:::l effects into each calculation to model o 5 typical operating conditions. Fig. 3 MALAPROP was used to predict the Performance of the Novette chain at its ultimate performance of the Novette fundamental 1 w frequency (1.05 pm) and second harmonic 2w (0.52 pm). The laser. Figure 3 summarizes the calculated upper curve is the maximum 1.053-pm capabilities of a single Novette beam. o ~--~~--~----~----~----~ energy deliverable to the KDP array; The upper curve illustrates the maximum o 2 3 4 5 the lower band is an estimate of the 1.053-.um energy deliverable to the KDP Pulse length, ns highest 0.53-pm on-target energy.

13 accomplished in several laboratories. its last target with a pulse of green Currently, material and optical coating light. The next morning, we began to limitations restrict the harmonic fluxes disassemble it, and what had been up to tolerable in larger, high-power laser that time the two most powerful single­ systems. The handling and focusing of arm laser-amplifier chains ever built the second harmonic, 0.526-,um light, is faded into history. By the end of possible with the same materials and September, the electrical system for the coating technologies used for 1.0S3-,um Novette laser was being installed. light and is, therefore, relatively Before the new year, the Shiva laser straightforward. At the third harmonic had irradiated its last target. Gradually, (0.3S1,um), we are concerned with we removed from it and refurbished the solarization of most optical , low components required for the Novette damage thresholds, and high nonlinear laser system. The remaining components index of refraction. remained in their cradles on the Shiva Although solutions to these problems spaceframe until they were needed for have been identified for the Nova laser, the Nova laser. we were unable to apply them rapidly During the summer of 1982, we began enough to build the Novette laser and to integrate the central control system meet an early experimental schedule. For software and operation of the smaller most studies, therefore, the Novette laser stages of the laser. The large laser was restricted to the second harmonic. components, with optical apertures of We used the Novette laser to test the 31.5 and 46 cm, began to arrive in first Nova second- and third-harmonic August. By early September, the Novette converter assembly, and we designed 31.S-cm amplifier sections delivered and used a unique fourth-harmonic approximately 4 TW each in separate converter for experiments in the spring 100-ps shots. Next, we installed the final and summer of 1984, the last rCF 46-cm amplifiers and the two frequency­ experiments conducted with the Novette conversion arrays. These two I20-m laser. optical trains had to be folded twice to Harmonic conversion of the Novette fit the 70- by IS-m space previously laser's output beams required doubler occupied by the . optics with diameters greater than 74 cm. In January 1983, we began a series To meet this requirement, the two of full-power, two-beam shots. On frequency conversion arrays were January 24, 1983, the highest powers mosaics of smaller crystals. Array yet recorded for a two-beam laser were architectures for initial operation were achieved: 2S TW at Iw (infrared) dictated by the size of available KDP converted to 13 TW at 2w (green) for crystals. We built one S- by S-element 100 ps. By this time, the Novette laser array of IS- by IS-cm crystals and was already engaged in 100-ps x-ray another 3- by 3-element array of 27- by laser studies. 27-cm crystals. The crystals and their Figure 4 is a histogram of high-energy "eggcrate" support structures were laser shots on the Novette system. Those sandwiched between neutral-solution labeled "target shots" contributed to processed windows to ensure that the several plasma physics investigations; crystals were all oriented within a those labeled "other" include all laser­ 100-,urad error cone. The fourth­ system diagnostic calibrations and harmonic converter used the same IS-cm laser-related experiments. crystals to produce green light, mounted The time required for cooling of this time in a rigid frame that also the Novette output amplifiers and supported a second mosaic of IS-cm subsequent laser-system realignment crystals to double the frequency once would limit this system to about eight again. target shots during a 16-hour working day. In practice, however, target and ovette Operation target diagnostic preparation and Near midnight on August 31, alignment generally cut potential system N 1981, the Argus laser irradiated usage by more than a factor of 2.

14 DEFENSE PROGRAMS

Dramatic reductions in productivity, Triggered by the Novette computer measured by number of target shots per control system, the master oscillator and month, came at the beginning of an preamplifier system provided all of the experimental sequence when an entirely subnanosecond timing and trigger pulses new target type was first introduced. required by the laser and target systems. Figure 4 suggests that rates in excess In addition, the computer control system of 60 target shots per month are automatically measured and aligned the achieved if a standard target design beam and shaped the pulses in time and is used. space. The devices installed in the master From March 1983 through August oscillator system have provided pulses 1984, we successfully completed several that are Gaussian shaped in time and important studies and experiments. We: range in duration from 100 ps to 1 ns for • Conducted two series of x-ray laser most target studies. Some experiments experiments with 100-500 ps pulses. used 3-ns-Iong pulses, shaped to • Completed wavelength scaling compensate for amplifier saturation, in studies using 1-ns pulses. order to produce temporally rectangular Fig. 4 • Achieved densities of about 100 output pulses of 6 to 8 kJ . Histogram of high-energy laser shots times liquid D-T by using targets of At the preamplifier output, the single on the Novette laser. The target shots contributed to several plasma physics similar geometry. optical pulse was split into two beams investigations, and the others were • Completed green-target scaling whose relative timing was adjusted by laser-system diagnostic calibrations studies and carried out fourth-harmonic means of variable-path-length sections and laser-related experiments. target experiments. For these ultraviolet experiments, we constructed a new tandem 5- by 5-element mosaic array of KDP crystals. Its success influenced August the Nova array design, which we July subsequently tested with the Novette June laser. Essentially a user facility, the Novette May system delivered energies as specified by April experimenters. Delivered energy was generally within ± 10% of the requested March value. Typically, 53 to 55% of the February _ Target infrared energy produced by each chain January Other was converted to green light and December delivered to targets. The energy delivered was limited principally by losses in November 'It lenses and debris shields that become ~ October coated with target material deposited by ~ c!.. September co previous shots. Ol ~ August ovette Performance July The laser pulse that eventually June reached the target emerged N May from Novette's actively mode-locked, Q-switched neodymium:yttrium-lithium­ April fluoride (Nd: YLF) oscillator as one of a March train of about 20 pulses at 8-ns intervals. Each pulse exited the master oscillator February with subnanosecond accuracy (± 2% in January energy and ± 1% in pulse duration) December relative to the Q-switch time and was November consistently selected by means of a Pockels cell gate. Future Nova experiments will require longer shaped o 10 20 30 40 50 60 pulses. No. of shots

15 and whose amplitudes were set using nonresonant parasitic oscillations in waveplate-polarizer pairs. The wings of each disk imposes a size limitation on the spatially were next glass disks. With well index-matched truncated so that only the central absorbing cladding on the disk edges, 26-mm-diam portion passed into the ASE depumping limits the gain-length amplifier chains. Thus, the first 28-cm­ product along the major axis of the disk long by 5-cm free-aperture rod amplifier to between 3 and 4. To preserve the high was presented with a circular beam stored energy, the Novette 46-cm disks having an average intensity of about were split by an absorbing stripe along 2 60 MW/cm . their minor axis. As we installed each After passing through the rod amplifier, we measured its small signal amplifiers, the beam passed through gain (see Table 1). relaying spatial filter and glass-disk The glass media were LHG-8 and amplifier stages of increasing size. The LG-750 phosphate, and the saturation first two amplifier stages, with 9.4- and flux encountered over the Novette 2 l4.5-cm clear apertures, were removed operating range was 4.02 to 4.05 J/ cm . from the Shiva laser and refitted with The Nova box amplifiers had flashlamps phosphate glass disks. The 20.8-cm-diam arranged only on the sides facing the stage used the box amplifier design, disks. Each of the 46-cm amplifiers, for which is more efficient than the example, was pumped with a nominal cylindrical design of the Shiva amplifier. -bank energy of 500 kJ at 20 kV, Following the 20.8-cm stage, the beam giving a stored energy of more than 8 kJ. was expanded to 31.5 cm, passed through Figure 6 shows measured Novette the final Faraday rotator isolator in the infrared output energies at the inputs to chain, and was turned by two mirrors the second harmonic converters (one in down the final leg of the space frame, each arm) for two pulse durations, 0.1 which held four 3l.5-cm amplifiers and and 1.0 ns. The O.l-ns pulses were never three 46-cm output amplifiers. Each of energetic enough to significantly saturate these amplifiers contained two disks the two output amplifiers used at this (Fig. 5). The large Nova box amplifiers pulse duration. We took the l-ns data stored approximately 2% of the electrical using three output amplifiers and energy in the capacitor bank as inversion somewhat larger spatial filter pinholes. in the disks at pulse-propagation time. Measurements show clear saturation Gain depletion by amplified of the 1.0-ns gain in keeping with a spontaneous emission (ASE) or by theoretical calculation based on Ref. 3.

Fig. 5 One of the six 46-cm output amplifiers used in Novette, each containing two split neodymium-glass disks.

16 DEFENSE PROGRAMS

Since November 1982, we have After leaving the 46-cm amplifier accumulated photographs of the output stage, the two beams were expanded to infrared beams presented to the second­ 74 cm in diameter and refl ected by harmonic crystal arrays at the rate of mirrors over 1 m in diameter up to the about two per day. Virtually all of these level of the target chamber, which they images, like the 10-TW sample shown in Fig. 7, are quite uniform in appearance, with slight evidence of nonlinear breakup. At 12 TW, small-scale, self­ Table 1 Amplifier gains measured as each amplifier was installed in the focusing effects are more apparent. The Novette laser. prominent dark shadow bisecting the image is caused by the absorbing region in the largest (46-cm) amplifier disks. Amplifier type Diam, em Nominal gain Intensity-dependent modulation grows as the output power increases. Our Rod 5.0 25.9 simulations from MALAPROP indicate Disk 9.2 6.6 Disk 15.0 that 1-ns operation becomes 4.2 Disk 20.8 2.1 unacceptably risky above 10 TW per Disk 31.5 1.8 beam. Disk 46.0 1.8

1500 .------, (a) 0.1 ns Fig. 6 Measured Novette infrared output ener­ gies at the inputs to the Novette laser's second-harmonic converters for two ...., ...., 8000 pulse durations (a) 0.1 ns and (b) 1 ns. >; 1000 Data points are shown for experiments ~ >; Q) ~ with the north and south beams of the c: Q) Q) a; 6000 laser. The solid curves indicate the orn calculated saturation of the amplifier. c: orn Q) c: rn Q) rn4000 ~ 500 'S 'S S- o ::l o 2000

20 40 60 80 100 120 140 20 40 60 80 100 120 140 Mid-disk sensor energy, J

4 .------~ (b) 70

60 3 "'E 50 t) III ~ Fig. 7 ~ Cl 40 2 a; ~ Output beam profile of a 10-kJ, 10-TW .§ 'iii pulse presented to the second­ C 30 c:

17 MALAPROP Generally, it models one time slice of a light beam of a single wavelength. A The MALAPROP computer code cross section of the beam is spread models the propagation of laser light. across a fixed mesh as complex values Using it primarily as a design analysis representing intensity and phase. As the tool for high-energy, inertial-confinement beam propagates through components in fusion (lCF) laser systems, we have been the modeled system, the intensity-phase able to account in a quantitative manner values change to include the effects of for effects, diffraction, the components. Specific components and noise sources. The code has been modeled include free-space propagation, instrumental in optimizing performance apertures, noise sources, amplifiers, versus cost in the design process of the lenses, and spatial filters. MALAPROP Novette and Nova laser systems. uses well-tested approximations, and we MALAPROP contains 13 000 lines of have validated its simulations in many FORTRAN and is our major program for experimental situations. analyzing laser systems. We also have The major specialty of MALAPROP is auxiliary programs to calculate some the coupling of nonlinear optics effects input parameters and summarize results. and diffraction, together with appropriate Analysis of a laser system such as noise sources. At the high intensities Nova uses about 15 minutes of Cray 1 of ICF systems, self-focusing of small computer time and 650 000 (64-bit) perturbations frequently limits the words of memory. maximum performance of the system. MALAPROP simulates coherent light (Self-focusing occurs when the light propagation. Its routines, or sets of beam is focused toward a narrow spot routines, model all of the components and thus becomes more intense; this can contained in an ICF laser system. damage some optics.) By modeling these effects, system designers often avoid 20 reducing the performance of the system ! and therefore reduce costs without I degradation of beam quality. 18 I We have modeled several laser I systems with MALAPROP. A. J. Glass 16 i first conceived the code in the mid-1970s ! as a tool to help understand results of I complex nonlinear physics. However, as 14 Propagation its capability expanded, particularly to N in air E include two-dimensional effects and

18 DEFENSE PROGRAMS

approach from opposite directions. cylindrical lenses and installed them Before entering the lenses that focus the beyond the II4 lenses within the pulses on the target, each beam passed evacuated target chamber. The final through an apodizer plate, the mosaic of optical component traversed by the laser KDP crystals, and an infrared beam beam on its way to the target was a glass dump. Bead-blasted bands on the or fused silica shield that protected the apodizer plate softly shadowed the low­ lens assembly from most of the target damage threshold interstices between debris. Although these shields were crystals in the second-harmonic replaced several times during the laser's generation array, which was located 1 m operating life, they were coated rapidly downstream. By preventing diffraction by debris from the massive targets from uncontrolled phase discontinuities irradiated and typically had between KDP crystals, this technique transmissions of only about SO%. also shields the final focusing lenses Routine monitoring of these optics from potentially damaging intensity allowed system operators to adjust ripples, but at a cost of 10 to 15% of the output and meet the needs of the beam area. experiments. Almost every shot yielded a measurement of the second-harmonic conversion efficiency, and some of these data are plotted in Fig. S. We have seen KDP frequency-conversion array 0.8 .-----:-=-:------, Fig. 8 efficiencies, corrected for beam area, of Measurement of the second-harmonic more than 75% for incident infrared conversion efficiency compared to the 2 intensities of 2 GW/cm . The theoretical theoretical curve calculated for a single 0.6 curve shown in Fig. S was calculated for >- 1.S-cm·thick, type II KDP crystal. The c:: bar indicates the spread of intensities a single 1.S-cm-thick, type II KDP crystal III'" in the laser beams, that is, the percent­ detuned in angle less than 100 ,urad and ~ ~ age of the total laser power that was presented with an infrared beam having gO.4 delivered to the second·harmonic array the same intensity spread as the Novette '0; at an intenSity below the indicated iii > value. For example, 80% of the total laser's. By this measure, the mosaic is c:: 80% o power is below the 80% pOint on the behaving as if it were a single crystal. If U 50% bar. second-harmonic irradiation was desired, 0.2 I 20% the beam was next passed through a - Theory filter or "beam dump," which absorbed 99% of the remaining infrared light and OO---~--~----~-~----~--~ passed 96% of the (green) energy. o 2 3 4 5 6 If a fourth-harmonic pulse was Input infrared intensity, GW/cm2 needed, the green beam was frequency doubled again in an array of O.S-cm-thick type I KDP crystals. Figure 9 shows that the Novette laser was able to deliver about 1.5 kJ of 0.26-,um radiation to 2 targets in pulses slightly shorter than ~ 1 ns. ~ '3 The converted light from each beam :!. >- proceeded through II4 focusing lenses, Cl iii which concentrated it on the targets. c:: III Most target designs called for a focal 'c'" region between 200 ,urn and 1 mm in 0 E ~ Fig. 9 diameter. .s:: X-ray laser experiments conducted on i= The Novette laser delivered about :; 1.5 kJ of fourth-harmonic (0.26-llm) en­ the Novette system required line foci 0 LL ergy to the targets in pulses slightly 100,um by 1 cm precisely oriented with 0 shorter than 1 ns. The error bars repre- respect to the x-ray laser diagnostics. To 0 2 4 6 8 10 sent uncertainties in measurements meet this request, we devised weak Input energy (lw), kJ based on analysis of instruments.

19 ummary operational experience with many new The Novette laser system was the technologies: S latest embodiment of the rapid • Segmented laser-disk amplifiers, evolution of powerful rCF laser systems. which can be scaled to almost arbitrarily (The relative energies of LLNL systems large apertures. from Janus to Nova are shown in • Neutral-solution antireflection Fig. 10.) Each of its two relatively coatings on laser optics, which have compact arms exceeded the total output moved the nominal damage threshold of all of the Shiva laser's 20 arms. The from 3-5 J/cm 2 to 15 J/cm 2 in 1 ns. Novette system was assembled in 30% of • Nova-style digital control and the time required to build the Shiva laser diagnostic techniques (fiber-optically and achieved a shot rate on target coupled together) that can be made to exceeding that of the Shiva system. An work routinely by achieving up to six evolutionary step of this magnitude in full-power shots in a single day. so short a time approaches a revolution • Large-aperture, multi-element KDP in high-power laser technology. The crystal arrays used to demonstrate Novette laser completed the rCF conversion to the second, third, and experiments planned for it, delivering fourth harmonics and to conduct target up to 9 kJ of 0.53-,um light to a fusion experiments with second- and fourth­ target in 1 ns. harmonic energy. ~ As a test bed for the Nova laser, the Novette laser provided the first Key Words: computer code-MALAPROP; inertial confinement fusion (lCF); laser-Argus, Cyclops, 7 10 janus, Shiva, Novette, Nova; neodymium-glass amplifier. 106 - Nova, 105 - Novette. ~ 104 - Shiva •• C) Notes and References iii c:: 103 t-- • Argus 1. G. R. Wirtenson, N. j. Brown, and L. M. Cook, W "Scaling Up the Neutral Solution Anti­ 2 10 f- • Cyclops reflection Process," Optical Eng. 22 (4), 450 • Infrared (1983) . • Janus 10 f- • Green light 2. W. W. Simons, j. T. Hunt, and W. E. Warren, A Blue light " Light Propagation Through Large Laser 1.0 f- Fig. 10 Systems," j. Quantum Elect. QE-17 (9), 1724 The energy levels of LLNL's succes­ I I I I (1981). sively more powerful and complex 1960 1970 1980 1990 2000 3. W. E. Martin and D. Milam, j. Quantum Elect. MOPA-design laser systems. Year QE-18, 1155 (1981).

20