Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 120
4 Laser-Plasma and Beam-Plasma Interactions INTRODUCTION The previous chapter was concerned with laser and particle beams insofar as they are used to produce HED plasmas, whereas this chapter is concerned with the physics of the beam-plasma interaction itself. As can be seen from the tables in Chapter 3, the most powerful focused lasers and particle beams today correspond to remarkable peak intensities—of order 1020W/cm2 for each. It is perhaps not surprising, then, that the interaction of these powerful beams with plasmas yields a host of new, and often very similar, physical phenomena. For example, both types of drivers may ionize material or create new matter through pair production. They may cause plasma blowout, produce nonlinear plasma wakes, self-focus, filament, scatter, hose or kink, form braided beamlets, generate radiation, accelerate particles to ultrarelativistic energies, and even refract at a boundary in a similar way (see Figure 4.1). These physical phenomena make up the intellectual theme of this chapter. The questions they raise make up a rich subfield for basic physics research. Answering these questions is of importance for a variety of applications for science and society. The answers may, for example, lead to breakthrough progress toward fusion energy, compact high-energy particle accelerators, and novel imaging techniques. They may also help us to understand the mechanism of ultra-high-energy cosmic ray (UHECR) acceleration and the formation of cosmic jets. The next two sections outline the fundamental physics questions and phenomena associated with high energy density beams in plasmas. To exhaustively delineate the
OCR for page 121
FIGURE 4.1 Refraction and periodic focusing of a particle beam at a plasma boundary. Courtesy of T. Katsouleas, University of Southern California. full array of phenomena associated with HED beam-plasma interactions is not the intent of this report, but the committee believes that the information provided will give the reader some idea of the wealth of phenomena associated with beams in plasmas. Then, in the secion on applications of HED beam-plasma physics, three applications are described rather extensively, followed by coverage of seven other applications. The final section discusses the opportunities for furthering this field, both experimentally and theoretically/computationally. QUESTIONS The intellectual theme of this chapter can be gleaned from the following list of illustrative questions of high intellectual value: At what intensities does dense matter become transparent? At what intensities does vacuum become opaque?
OCR for page 122
Can focused lasers “boil the vacuum” to produce electron-positron pairs? Can macroscopic amounts of relativistic matter be created in the laboratory and made to exhibit fundamentally new collective behavior? Can we predict the nonlinear optics of unstable, multiple-interacting beamlets of intense light or matter as they filament, braid, and scatter? Can the ultrahigh electric fields produced by laser wakes be used to make a tabletop accelerator with the luminosity and beam quality needed for applications in high energy and nuclear physics? Can short-pulse lasers be used to “ignite” a fusion plasma in the laboratory? Can lasers and particle beams simulate relativistic shocks and gamma-ray bursts in astrophysics? Can high energy density beam-plasma interactions lead to novel “next-generation” light sources? Are the same mechanisms responsible for laboratory plasma accelerators and plasma lenses also operating in the acceleration of particles from supernovae and the collimation of cosmic jets? Can ion beams produced by relativistic laser-plasma interactions be used as a source for beam-plasma physics, a diagnostic probe, or as a front-end component for accelerators? Can such interactions produce novel or economic radioactive ion sources? HIGH ENERGY DENSITY BEAM-PLASMA PHYSICS: PHENOMENA In introductory physics, we learn about beam propagation in terms of simple laws—Snell’s law and the laws of reflection and diffraction. In reality, beam propagation is more complex—beams filament, Raman scatter, frequency shift, and so on—even in ordinary media. When the beams reach the high energy densities that are the subject of this report, the medium through which they propagate becomes necessarily a plasma. The nonlinear optics of extreme beams, with their associated gigabar pressures, teravolts per centimeter electric fields, and gigagauss magnetic fields, is no less rich than in regular media. The beams exhibit a wide range of propagation phenomena. These include the familiar, such as focusing and stimulated scattering instabilities, as well as the less familiar, such as braided light and relativistic shocks. In addition, the beam-plasma interactions lead to new radiation-generation mechanisms and to the high-gradient acceleration of particles in the plasma. Recent advances in experimental and computational capabilities are creating exciting new opportunities on two fronts: first is the use of new tools to address questions that have defied solution for a number of years, and second is the exploration of new regimes opened through advances in beam infrastructure at very high
OCR for page 123
energy densities. In the first category are questions such as: Can we predict the evolution of one or several beamlets propagating through a dense plasma? The importance of this question for laser fusion has been appreciated for more than 30 years. Many advances have been made, yet the answer remains an open challenge. It requires detailed understanding of parametric processes such as Raman and Brillouin scattering in the presence of trapped particles and complex thermal transport. Fortunately, new diagnostics such as x-ray “backlighters” and new computational tools such as advanced algorithms and massively parallel computing offer the possibility of finally answering this question. At the other end of the spectrum are new questions and opportunities that arise as a result of the tremendous advances in short-pulse laser and particle beam technology. One of these is the fast ignitor: Can a short laser pulse be used to accelerate and deliver hundreds of mega-amperes of plasma electrons in a short burst to the core of a fusion fuel pellet and ignite it? The answer to this question may contribute to realizing the international goal of fusion ignition. Other questions emerge at still higher laser and beam energy densities. Chirped pulse amplification (CPA) laser technology has enabled a proliferation of multiterawatt laser systems. When focused, their peak fields exceed several gigavolts per centimeter, and the quiver energy of electrons in these fields exceeds several MeV. These HED beams are creating macroscopic amounts of relativistic matter in the laboratory for the first time. Not surprisingly, they are producing a bounty of new relativistic phenomena such as relativistic transparency. In relativistic transparency, the electron mass increases sufficiently in the laser field to reduce the plasma frequency (given by where np is the plasma density and γ is the relativistic Lorentz factor) below the laser frequency. At this point, the plasma becomes transparent to the laser pulse it would normally reflect. Other examples of relativistic phenomena accessible with current laser technology include highly nonlinear plasma wakes in which the plasma is driven to complete blowout, ultrastrong plasma lensing of both photons and particles, and intense radiation generation from the terahertz to x-ray frequency range by various mechanisms. Electron beams with energies up to 100 MeV with small normalized emittance (of order millimeters to milliradians) and nanocoulombs of charge have been generated by plasma wakes in millimeter gas jets. Although the electron beams in these experiments had large energy spreads, the acceleration gradient they achieved was more than a thousand times the gradient of a conventional linear accelerator. This leads to the question: Can wakefield acceleration yield sufficient energies and beam quality so as to enable high-energy physics on a tabletop? Similarly, focusing by plasma lenses has been recently demonstrated at the Stanford Linear Accelerator Center (SLAC) at field strengths of 100 MG/cm. Comparing this
OCR for page 124
to the 5 kG/cm strength typical for quadrupole magnets leads naturally to the question: Can short plasma lenses enhance the final focus of a linear collider? On the horizon are yet higher density beams and lasers. Chirped pulse amplification technology applied to high-energy lasers is making it possible to consider multipetawatt- to exawatt-class lasers in the near future. The focused field gradients of such lasers will exceed teraelectronvolts per centimeter, and the quiver energies will exceed gigaelectronvolts. As such extreme beams propagate in plasma, they can be expected to create copious electron-positron pairs and possibly heavier pairs. It may be interesting to consider questions such as: Can beams undergo a stimulated pair scattering instability by coupling parametrically to the pairs they create? Can backscatter amplification or other techniques be used to make even higher energy density pulses, exceeding even chirped pulse amplification limits? Using ultrahigh-intensity lasers, it may become possible to simulate some of the properties of black holes. At an intensity of 1026 W/cm2, an electron will undergo an acceleration of 1027 g, comparable to the gravitational acceleration at the event horizon of a black hole. This high acceleration could be used to study Unruh radiation, which is similar in many respects to Hawking radiation, induced by gravitational fields. At modest accelerations, the radiation is described by Maxwell’s equations. But it is interesting to study at very large accelerations whether, as Unruh has suggested, there is radiation beyond that predicted by Maxwell. The source of the extra radiation is an effective temperature kBT=ℏa/c associated with a particle undergoing acceleration a, causing it to emit blackbody radiation (given by Stefan’s law, σT4). At sufficiently high intensities, even vacuum can be broken down. The field necessary to achieve pair creation (“boil the vacuum”) is the Schwinger field, Es, which is also equal to the field required to accelerate an electron to its rest-mass energy in a Compton wavelength, λc=2πℏ/mc. The Schwinger field is Es=m2c3/qℏ≈1016V/cm and corresponds to an intensity of Is≈0.5×1030 W/cm2. To reach the Schwinger field would require, for example, a laser pulse having an energy of 100 MJ, pulse duration of 10 fs, focused down to an area of µm2. Although such fields are beyond the horizon, other nonlinear quantum electrodynamics effects could be accessed at more modest fields. The optical mixing of two pulses in vacuum can produce a standing wave “density grating” composed of a virtual electron-positron plasma. For petawatt, kilojoule-class lasers, a nontrivial pair probability density can be created. It may be possible to scatter off of this grating with a third laser, thereby demonstrating the nonlinear optics of vacuum. Finally, it is noted that an alternate path to the Schwinger field could be an x-ray free-electron laser. If 1.5-Å x rays could be focused to a diffraction-limited spot size of roughly 2 Å, the Schwinger field could be reached with an energy of 2 J at a pulse length of 10 fs.
OCR for page 125
HIGH ENERGY DENSITY BEAM-PLASMA PHYSICS: APPLICATIONS Three applications are described below in some detail, followed by discussion of seven more applications in briefer presentations. Three Important Applications Multi-GeV Electron Acceleration in Plasma Wakefields The high cost and size associated with conventional rf accelerator technologies has been a prime motivation in advanced accelerator research for more than two decades. Wakefield accelerators driven by high energy density laser or particle beams promise an entirely new type of technology for building compact high-energy accelerators. Laser pulses propagating in plasmas can generate large-amplitude plasma waves, that is, wakefields, which can be used to trap and accelerate electrons to high energies. The amplitude of the plasma wave is largest when the laser pulse duration (or its modulation) is on the order of the plasma period. This laser-plasma interaction forms the basis for the laser wakefield accelerator (LWFA). A wealth of new and interesting experimental results on LWFAs has been obtained in recent experiments around the world (see Figure 4.2). On the theoretical and computational front, detailed analyses of the propagation and stability properties of intense laser pulses in plasma channels have been conducted. Recent advances in algorithms and high-performance computing are enabling fully self-consistent modeling of full-scale wakefield experiments in three dimensions for the first time. This work provides a strong foundation for next-generation wakefield accelerator research aimed at producing electron beams with gigaelectronvolt energies and high beam quality. To reach multigigaelectronvolt electron energies in an LWFA, it is necessary to propagate an intense laser pulse long distances (many Rayleigh ranges) in a plasma without disruption. However, a number of issues associated with long-distance propagation in plasma must be resolved before a viable, practical high-energy accelerator can be developed. These issues include optical guiding, instabilities, electron dephasing, and group velocity dispersion, all of which can limit the acceleration process. The scale length for laser diffraction is given by the Rayleigh range; therefore, the acceleration distance is limited to a few Rayleigh ranges. Since this is far below that necessary to reach gigaelectronvolt electron energies, optical guiding mechanisms such as relativistic focusing, ponderomotive channeling, and preformed plasma channels are necessary to increase the acceleration distance. There is, in fact, ample experimental confirmation showing extended guided propagation in plasmas and plasma channels. The combining of such guiding techniques with a
OCR for page 126
FIGURE 4.2 Progress in plasma accelerator energy: maximum energy gains and laboratories where they were obtained. suitable injector geometry to achieve controlled acceleration of a monoenergetic beam is the next key step in LWFA development. Another approach to achieving longer acceleration distances is being pursued at several particle beam facilities. These take advantage of the natural tendency of particle beams to propagate longer distances without spreading, compared to lasers. For example, at the SLAC, electron beams have been used to generate wakefields over a meter and to accelerate electrons by as much as 350 MeV. Lasers incident on solid targets can also be used to accelerate heavier particles— protons and ions. The petawatt laser at Lawrence Livermore National Laboratory (LLNL), incident on a thin target with intensity approaching 3×1020 W/cm2 has generated ~2×1013 protons with energy in excess of 10 MeV, with the highest energy approaching 58 MeV. The generation mechanism has been attributed to the electrostatic field set up by the escaping jet of hot electrons from the back of the target. Experiments with the 50-TW Vulcan laser at Rutherford Appleton Laboratory (RAL), with on-target intensity approaching 5×1019 W/cm2, have demonstrated
OCR for page 127
generation of protons with energies approaching 30 MeV and Pb46+ ions of up to 430 MeV energy. The energetic ion production in this experiment has been attributed to a combination of ponderomotive acceleration at the front surface and escaping hot electrons creating “coulomb explosion” at the rear surface. Intense Laser-Plasma Interactions with Long-Pulse Lasers High-power lasers with pulse lengths >1 ns have proved to be a remarkably versatile tool for the study of high energy density physics, with important applications ranging from inertial confinement fusion to stockpile stewardship and astrophysics. Over the past four decades, the energy of these lasers has increased at a rate comparable to the growth in computer power, culminating in the National Ignition Facility (NIF) now under construction. This megajoule-class 0.35-µm laser will provide a crucial test of inertial fusion as a future energy source and will also enable a broad study of high energy density science for stockpile stewardship and other applications. The field of laser-plasma interactions is a vital enabling technology for these many applications as well as a remarkable testbed for understanding broadly applicable nonlinear plasma science. The challenges associated with the interactions of long-pulse high-energy lasers with plasmas is well illustrated by considering the nominal hohlraum target for achieving ignition on the NIF. As shown in Figure 4.3, this hohlraum will be irradiated from two sides by 192 laser beams, arranged into inner and outer cones. The relative power in these cones is tuned to provide the time-dependent x-ray symmetry required for the implosion. The hohlraum is filled with a low-Z plasma with a quasi-uniform density of ~1021 cm−3, which is needed to prevent radiation asymmetry due to excessive motion of the hohlraum walls. Before collisionally absorbing in the high-Z wall plasma, the laser beams propagate through up to 5 mm (>104 free space wavelengths) of the low-Z fill plasma at a peak intensity of order 2×1015 W/cm2 (and much larger near the laser entrance holes). Excellent absorption of the laser beams is desired, and excellent temporal and spatial control of the absorption is required for the requisite implosion symmetry. The interaction physics is extraordinarily rich. As the laser beams propagate through the high-temperature (electron temperature ~5 keV) plasma within the hohlraum, they can filament and spray out in angle, developing additional spatial and temporal incoherence. The beams can undergo enhanced bending in places where the plasma flow is near sonic, where significant energy transfer among crossing beams can also occur. The laser beams can scatter and/or generate high-energy electrons via a variety of instabilities involving either electron plasma waves (the stimulated Raman instability and the two-plasmon decay) or ion sound waves (the stimulated Brillouin instability). These instabilities were controlled in previous
OCR for page 128
FIGURE 4.3 Images of the type of hohlraum used for ignition experiments. The schematic on the right shows the hohlraum irradiated by 192 NIF beams. Courtesy of Lawrence Livermore National Laboratory. millimeter-scale experiments at the Nova facility by using significant laser beam smoothing. Understanding and controlling their evolution in new regimes and over the centimeter scales and higher energies at NIF pose a significantly greater challenge. Rarely have questions of such a fundamental physics nature been so directly coupled to a programmatic and societal need. Fast Ignition In the past decade, the development of short-pulse, ultrahigh-power lasers has motivated another approach to inertial fusion energy called Fast Ignition. In this case, cold deuterium-tritium (DT) fuel is first compressed to high density, and fusion burn is then initiated by a rapid heating of a portion of the fuel to high temperature. Significantly higher gains are possible compared to the conventional approach.
OCR for page 129
Furthermore, the Fast Ignition scheme is less sensitive to hydrodynamic instabilities and mix, since the processes of fuel compression and hot spot creation are separated. The critical issue for Fast Ignition is the efficient generation and transport of a short, ultraintense energy flux to the precompressed fuel, a relatively unexplored topic involving many extremely rich nonlinear physical phenomena and many potential applications. The following estimates illustrate the challenge: The basic requirement is to heat to about 10 keV a volume of cold fuel with a radius equal to the alpha particle range, which will enable a propagating burn into the remaining fuel. For a compressed DT fuel density of 300 g/cm3, the required radius is about 10 microns and the disassembly time is about 10 ps. To heat this amount of fuel to 10 keV requires delivery of at least 3 kJ in a time less than the disassembly time, which represents a power in excess of 4×1014 W and an energy flux of order 1020 W/cm2. The incident laser energy needs to be roughly an order-of-magnitude larger, depending on the efficiency of energy transfer from the laser to the dense fuel and on the size of the hot spot created. Laser-accelerated electrons and/or protons with energies of order 1 MeV and 5 MeV, respectively, offer an attractive means of coupling this energy because their range is similar to that of the alpha particles. Just how efficiently such beams can be generated and transported looms as a key fundamental question. Intense laser pulses made with chirped pulse amplification technology provide a key tool for investigating physics in this regime. Although at lower energy/pulse and shorter pulse lengths than required for Fast Ignition, CPA lasers have already achieved the requisite powers (up to 1015 W) and focused light intensities (up to 1021 W/cm2). These enable important features of this ultraintense laser plasma regime to be explored. The features include relativistic self-focusing and filamentation of the laser light, pronounced hole boring into overdense plasma, and a variety of acceleration processes. Other notable effects include self-generated magnetic fields with an amplitude up to 109 G and multi-MeV ion generation. Computer simulations of this ultraintense regime using two- and three-dimensional electromagnetic, relativistic particle codes have illustrated many of these effects and shown significant absorption (>30 percent) into electrons with energies above 1 MeV. They illustrate several acceleration mechanisms, such as heating due to the oscillating ponderomotive force, conversion of transverse laser fields to longitudinal fields at overdense plasma layers, and electron acceleration at the betatron resonance in relativistic laser channels (an inverse free electron laser (FEL) process). Many of these effects have now been observed in experiments, but much more work is needed for a quantitative understanding. The transport of ultraintense energetic particle fluxes over distances of hundreds of microns from the laser absorption region to the compressed fuel is another very challenging issue at the forefront of high energy density physics. For example, if a
OCR for page 130
power of 1015 W is carried by electrons with an average energy of 3 MeV, the associated current is ~300 MA. Such a current far exceeds the Alfvén critical current (~100 kA) and so requires neutralization by return currents in the dense plasma. Many physical processes then come into play, including strong return current heating, excitation of the Weibel instability, formation and coalescence of current filaments and magnetic channels, as well as various instabilities that can disrupt the beam propagation (see Figure 4.4). Especially encouraging have been experiments at Osaka University in which cold fuel has been assembled to high density and a significant heating of the fuel by a short, very intense, low-energy (60-J) laser pulse has been observed. An energy transfer efficiency from the short laser pulse to the fuel of approximately 20 percent has been inferred. FIGURE 4.4 Three-dimensional particle-in-cell simulations of energetic electron generation and transport with laser pulses in overdense plasmas. Contours of magnetic-field structure due to the Weibel instability are shown. Courtesy of Y.Sentoku, General Atomics.
OCR for page 138
FIGURE 4.9 Simulated time-averaged electron-cloud density in a field-free region in the arcs of the PEP-II positron ring. The beam (not shown) occupies a small region at the center of the chamber, of order 1 mm in diameter. Courtesy of M.A.Furman, Lawrence Berkeley National Laboratory. Short-Lifetime Particle “Factories” Laser wakefield acceleration in high-density plasmas produces longitudinal electric fields comparable to the laser transverse field. For next-generation CPA lasers at 1023 W/cm2, this suggests longitudinal fields of the order of ~1 TV/cm. With such fields, particles could be accelerated to relativistic energies in such short distances as to significantly extend their lifetime. Particle-in-cell simulations by A.Pukhov have shown that 5-GeV protons could be produced as the result of the interaction of an ultraintense pulse at the 1023 W/cm2 level with a metallic target. These protons can be used to produce pions, which have a lifetime at rest of 20 ns and are well synchronized to the 10-fs laser pulse. By using a laser-driven acceleration mechanism, one could quickly increase the pion energy and lifetime by roughly 100 times to 15 GeV and 2 µs, in only a few picoseconds, a time much shorter than the pion lifetime. This long lifetime would make it possible to accelerate the pions to higher energy, if necessary, with conventional (i.e., lower gradient) accelerating structures. At the gigaelectronvolt energy level, the pions will decay into a well-collimated beam of muons and neutrinos. Sources of muons are of interest for a possible muon collider that offers cleaner particle physics interactions than in a proton-antiproton collider (because the muons lack the quark substructure of protons) but lower synchrotron radiation than circular electron-positron accelerators. Another elegant application that combines high-
OCR for page 139
energy accelerators and either high-power lasers or magnetic wigglers is a gamma-gamma collider. By colliding photons instead of charged particles, limits to luminosity arising from beam-beam interactions (so-called bremsstrahlung) can be avoided. OPPORTUNITIES This section identifies some illustrative examples of research opportunities for high energy density laser-plasma and beam-plasma physics using existing infrastructure, existing infrastructure with modifications, and future infrastructure. “Infrastructure” includes experimental laser and beam facilities and computational facilities. A number of these were identified in the facilities matrix in Chapter 3 (Tables 3.1 through 3.4), and selected examples are briefly described here. Beam Facilities: The Stanford Linear Accelerator Center Fundamental research into the formation, injection, and propagation of short and possibly shaped bunches in long plasmas appears to hold great promise for advancing toward a future device that can benefit high energy physics and other areas of science. There are relatively few beam facilities for high energy density plasma physics. The Accelerator Test Facility at Brookhaven National Laboratory and the Advanced Wakefield Facility at Argonne National Laboratory are significant user facilities with 50-MeV beamlines. Only SLAC has facilities in the gigaelectronvolt class and with positrons. This subsection is limited to describing some of the possibilities available at SLAC. In order to increase particle beam-driven wakefields to amplitudes similar to the laser wakefield case, the bunch length of the electron beam drivers must be shortened from their present length of a few picoseconds to the subpicosecond range. Presently, work at SLAC is directed toward reducing the bunch length of the beam to 20 fs (a factor of 200 shorter than the present bunch length) for applications such as the Linac Coherent Light Source (LCLS). If this can be done, it becomes possible to consider ultrahigh-gradient acceleration at several gigaelectronvolts per meter over multimeter distances in plasma modules referred to as plasma afterburners. In this concept, the energy of an e+e− collider may be doubled in a few meters by splitting each of its bunches into two, with the first bunch exciting a plasma wake that accelerates the second bunch just before the collision point. Appropriate phasing and shaping of the second bunch will be key to keeping the energy spread small, and plasma lenses are key to maintaining the luminosity at the collision point. High energy density plasma experiments at SLAC thus far have been parasitic with other laboratory programs sharing a single beamline, but the upcoming LCLS may be incompatible with parasitic operation. A dedicated beamline that could be
OCR for page 140
created at modest cost would be a significant opportunity for high energy density plasma physics and other science. Coupling Physics on Long-Pulse Laser Facilities As discussed in the subsection entitled “Backscatter Laser Amplifer,” above, fundamental research into the interaction of intense laser light with long-scale-length plasmas is very important for the optimal use of high-power lasers for inertial fusion and high energy density physics in general. Indeed, the time is ripe for significant advances. In recent years, the plasma and laser beam conditions have become increasingly well characterized and controlled in experiments. Furthermore, the diagnostics have become quite sophisticated. For example, Figure 4.10 shows measurements via Thomson scattering of the temporally and spatially resolved amplitudes of both the electron plasma wave due to stimulated Raman scattering and the ion sound wave due to stimulated Brillouin scattering in a laser-plasma experiment. Such detailed measurements provide remarkable tests of the fidelity of simulation models and new insight into relevant behavior. Development of quantitative models of laser-matter interactions (including, for example, experimentally tested saturation models for laser-plasma instabilities) requires an improved understanding of many basic plasma phenomena and is ideal for university research. University-scale experiments with near-kilojoule lasers can identify and quantify key physical processes. These less expensive facilities enable more numerous experiments to explore parametric dependencies and provide a very important testbed for concept development and discovery. As an example, a new electron-acoustic wave has been discovered in recent experiments with the Trident facility, a 250-J laser at Los Alamos National Laboratory. Unfortunately, there is currently a paucity of kilojoule-class long-pulse length lasers at universities in this country. Experiments on higher-energy lasers, such as OMEGA, and in the future, the NIF (see Chapter 3 for more about the large laser facilities), allow many important features of coupling physics to be explored in larger and/or hotter plasmas, more relevant to future ignition experiments. For example, on the OMEGA laser new techniques for laser beam smoothing, such as smoothing by spectral dispersion and polarization smoothing, were developed and demonstrated. Reduction of stimulated scattering by these techniques was observed. Beam phasing has been demonstrated as a technique to control radiation symmetry in hohlraums. In ongoing experiments, the interaction physics of multiple crossing laser beams is being studied, and new instability control mechanisms are being developed. The National Laser Users Facility (NLUF) at the University of Rochester provides facility time on the OMEGA laser for outside users. For several decades, NLUF has
OCR for page 141
FIGURE 4.10 Measurements in space and time of the amplitudes of both electron plasma and ion sound waves in a laser-irradiated plasma. SOURCE: Reprinted, with permission, from C.Labaune, H.A.Baldis, N.Renard, E.Schifano, and A.Michard, 1997, “Interplay Between Ion Acoustic Waves and Electron Plasma Waves Associated with Stimulated Brillouin and Raman Scattering,” Phys. Plasmas, 4:423–427, copyright 1997 by the American Institute of Physics. enabled valuable laser experiments on topics ranging from basic plasma and atomic physics to implosion physics. Another national facility providing a unique resource, an ultrasmooth few-kilojoule laser, is NIKE, at the Naval Research Laboratory. It enables instability studies to be done without worrying about beam seeding of specific modes. By far the most energetic laser for high energy density physics will be the NIF, described in the Chapter 3. The first quad of beams is anticipated to be available for experiments in fiscal year (FY) 2003, and the entire system is projected for comple-
OCR for page 142
tion later in the decade. As more beams are activated, more sophisticated experiments can be carried out. NIF experiments on the coupling physics will be vital for the development of quantitative models for the interaction of multiple, crossing laser beams with very long scale length plasmas. In turn, this understanding will enable more efficient use of NIF for applications ranging from inertial fusion to high energy density physics. As a bonus, the concomitant advances in the understanding of nonlinear, kinetic plasma phenomena will benefit other applications of plasma science. About 10 percent of the NIF laser shots have been allocated for basic science, which is also represented on the facility’s Experimental Planning Advisory Committee. A symposium on frontier science with NIF was held in 1999. Finally, a Stewardship Science Academic Alliances program for university users was launched in FY 2002. This program builds on and significantly extends the existing Inertial Fusion Science in Support of Stockpile Stewardship grant program. Chirped Pulse Amplification Laser Facilities The generation and transport of ultrastrong energy flows in matter are clearly a very promising frontier of high energy density physics. Applications, many of which have been described above, extend beyond fusion energy, to fast ignition, to stockpile stewardship, and to astrophysics. The studies can also enable improved understanding of many basic relativistic plasma phenomena, such as relativistic shocks. Many relevant experiments can be carried out with university-scale, 100-TW-class lasers. Such lasers are available at a number of universities, including the University of Michigan, the University of Maryland, and the University of Texas at Austin. The first petawatt laser used a beamline of the Nova laser at LLNL. Petawatt-class lasers are currently under construction at Osaka University in Japan and at the Rutherford Appleton Laboratory in the United Kingdom. Plans are under way for petawatt-class lasers at the University of Nevada at Reno using parts from the LLNL laser, and at both Sandia National Laboratories and the University of Rochester. However, it is important to develop high-energy, high-power lasers beyond a petawatt. That can be accomplished, for example, by configuring the NIF to provide a chirped pulse amplification pulse in one or more beamlines. In this way, one can move into the multipetawatt, even exawatt, class with 10 to 100 kJ of energy. This would allow x-ray radiography of imploding capsules, testing of fast ignition, and exploration of extreme field science. This development should be of high priority, for it clearly is needed for further advance of this important field.
OCR for page 143
Novel Light Sources To date, the study of ultrafast processes has largely relied on femtosecond optical pulses. Since x rays interact with core electronic levels and hence are effective structural probes, the availability of femtosecond x-ray pulses from laser sources and the intrinsic synchronization between laser and x-ray pulses would make it possible to directly probe changes in atomic structure on ultrafast time scales. Such sources could be complementary to recent light source facilities such as that at the Jefferson Laboratory infrared free electron laser. The vast potential of these short-pulse, multicolor operating scenarios for materials and other research and development is only just beginning to be explored. The Linac Coherent Light Source (LCLS), which is proposed to be constructed in the 2004–2007 time frame, will utilize the last third of the existing SLAC linac. This linac produces a high-current 5- to 15-GeV electron beam that is bunched into 230-fs slices with a 120-Hz repetition rate. When traveling through an ~100-m-long undulator, the electron bunches will lead to self-amplification of the emitted x-ray intensity in the 800- to 8,000-eV energy range and will function as the first x-ray free-electron laser. The emitted coherent x rays will have unprecedented brightness and hence offer a new window on beam-matter interactions. Two general classes of experiments are proposed for the LCLS. The first class consists of experiments in which the x-ray beam is used to probe the sample without modifying it, as in most synchrotron source experiments today. In the second class, the LCLS beam is used to induce nonlinear photo-processes or to study matter in extreme conditions. The latter experiments will include pump-probe studies of so-called warm dense matter. High-Performance Computing When simulation models were first applied to the study of laser-plasma interactions in the 1960s, lasers were nanoseconds long and computing power was measured in megaflops. The parallel development of femtosecond lasers and teraflop computing has led to a convergence of these factors by more than 10 orders of magnitude! As a result, it has recently become possible for the first time to perform detailed simulations of high energy density beam-plasma experiments using algorithms with very low level approximations in three dimensions with unscaled parameters. It is even possible to begin to use the actual number of particles in simulations that are used in some experiments (e.g., 1010 electrons in a beam experiment). This is leading to a significant change in the way that simulations are used to study high energy density beam-plasma interactions. While scaled simulations were previously used to test theoretical concepts that could then be applied to experi-
OCR for page 144
ments, simulations are now able to unveil the relevant physics in a short-pulse experiment directly. Moreover, by benchmarking the codes against real data, it is possible to build and test a future device at a fraction of the cost otherwise required. The modeling of laser-plasma interactions is also becoming more powerful and detailed, owing to ongoing advances in model development. Three-dimensional computer models now span the gamut from particle-electron, particle-ion codes, and hybrid codes, in which ions are particles and the electrons are a fluid, to various reduced descriptions that can follow the light-wave propagation over experimental time and space scales. In the modeling of the interaction of subpicosecond laser pulses with plasmas, novel techniques such as moving grids and massively parallel particle codes have been able to directly simulate a number of experiments on laser-plasma acceleration and related phenomena. The codes are in excellent agreement with much data. For long laser pulses, the modeling is a special challenge, since the key processes occur on widely disparate space and time scales. As an example, consider laser-plasma interactions for inertial fusion. In the mainline approach, laser beams with a pulse length of order 10 ns interact with a target plasma with scale lengths of order 1 cm. The absorption and reflection of the beams are modified by the excitation of plasma waves with maximum frequencies comparable to the laser light frequency (~6×1015 s−1) and with scale lengths of order the electron Debye length. (For a plasma with density ~1021 cm−3 and an electron temperature of 5 keV, the Debye length is ~10−6 cm.) Kinetic modeling using particle or Vlasov codes as well as various reduced descriptions is used to study the growth and saturation of the plasma waves. Ideally, these simulations need to consider a volume of plasma at least equal to that of a laser speckle (~3×104 λo3, where λo is the free space wavelength of the laser light and f/8 optics is taken) and to follow the evolution for many instability growth times (many thousands of laser light periods). These simulations of the microphysics are used to generate saturation models. The models are then input into wave propagation codes that follow the evolution of laser beams with realistic structure over experimental time and space scales and can be coupled with hydrodynamic codes. Accurate incorporation of the fine time and space scale physics is a grand challenge shared with other areas of science. Although the challenge is great, impressive progress is being made owing to the rapid development of high-performance computing. Figure 4.11 illustrates the remarkable increase in floating point operations per second since 1993 and includes projections to 2005. Note also the increase in the size of the plasma that can be modeled in F3D, the wave propagation code at LLNL. If projected improvements in algorithms are included, it will soon be possible to simulate the propagation of an entire NIF laser beam, including realistic models of laser beam smoothing as well as models for the microphysics due to plasma wave excitation and saturation.
OCR for page 145
FIGURE 4.11 The increase in computer power parallels the increase in laser peak power, (a) The increase in computer power in Mflops from 1993 to 2005 (projected) and (b) the increase in laser peak power from 1960 to 2010. SOURCES: Images (a) courtesy of C.H.Still, Lawrence Livermore National Laboratory; and (b) courtesy of W.Kruer, Lawrence Livermore National Laboratory.
OCR for page 146
FIGURE 4.12 Five top supercomputer sites as of November 2001. (a) ASCI Blue-Pacific at Lawrence Livermore National Laboratory is an IBM SP 604e machine capable of 2.14 Tflops; (b) the Terascale System at the Pittsburgh Supercomputing Center is a Compaq Alpha Server machine capable of 4.05 Tflops; (c) the National Energy Research Scientific Computing (NERSC) center at the Lawrence Berkeley National Laboratory is an IBM SP Powers machine capable of 3.05 Tflops; (d) ASCI Red at Sandia National Laboratories is an Intel processor-based machine capable of 2.37 Tflops; and (e) ASCI White at Lawrence Livermore National Laboratory is an IBM SP Power3 machine capable of 7.22 Tflops. Courtesy of (a) Lawrence Livermore National Laboratory, (b) The Pittsburgh Computer Center; (c) Lawrence Berkeley National Laboratory, (d) Sandia National Laboratories, and (e) Lawrence Livermore National Laboratory.
OCR for page 147
It is important to note that research in new computer architectures is leading to additional significant performance advances that also are not accounted for in Figure 4.11. These advances are now being realized on the basis of research both within and outside the United States, most notably in Japan, where the Earth Simulator machine has demonstrated, and has re-emphasized, the importance of optimizing memory bandwidth relative to floating point performance. Physicists can undertake massive calculations in a number of ways. Perhaps the most direct, besides using the National Science Foundation centers, is to access the national laboratory supercomputers. The supercomputers developed by the ASCI program represent a powerful enabling technology for the modeling of high energy density science. Figure 4.12 shows the top five supercomputer sites as of November 2001. The fastest computer is the Advanced Strategic Computing Initiative (ASCI) White machine at LLNL, achieving in excess of 7 Tflops. Supercomputers reaching 100 Tflops are projected for 2005. The ASCI Alliance program is an avenue for unclassified access to the supercomputers at the national laboratories. This program has three levels, ranging from centers of excellence to individual investigator grants.
OCR for page 148
This page in the original is blank.
Representative terms from entire chapter: