D. Tatomirescu, Laser-PlasmaAccelerationatUltraHighIntensity - NumericalModeling
important to note that at the onset of the ultra-highintensity regime the improvements in maximum particle energy and angular distribution follow the trends observed in lower intensity cases previously studied in the literature [82-87]. While each of these target augmentations (curvature, microdot and cone structure) has its advantages, a composite target featuring all three attributes has the potential to produce higher quality proton and ion beams. From our simulations we can clearly see that the case of a microdotted concave tip cone target has led to the most favorable results in terms of hot electron conversion, laser pulse focusing and particle beam maximum energy. Further studies will have to be done in order to confirm that the poor proton angular distribution for the cone target is a consequence of the proportionality between cone tip size and microdot size.
By leveraging those novel high scalable numerical tools to perform large scales and massively par- allel PIC simulations, I conducted an extensive numerical and theoretical study in the context of high order harmonic generation on plasma mirrors. In particular, I was interested in the generation of isolated attosecond light pulses from plasma mirrors in the relativistic regime. In this regard, an innovative scheme to angularly separate attosecond light pulses has been proposed at CEA [ 26 ] a few years ago. However, this approach cannot be directly transposed to the relativistic regime, because the optimal setup that maximizes the harmonic generation efficiency increases the harmonic beam divergence. My work involved identifying and characterizing optimal interaction conditions to enable the angular isolation of attosecond light pulses. For this matter, I developed an analytical toy model to predict the angular separation of attosecond light pulses based on the laser and plasma parameters, and in the presence of spatio-temporal couplings. This model was validated with an extensive numeri- cal parametric study in 2D and 3D geometries, over a broad range of laser and plasma parameters. In the end, we designed two efficient all-optical schemes for generating isolated attosecond pulses, both involving specific control of the driving laser phase.
The results presented in this thesis is the outcome of collective work of many team members. Most of the credits definitely go to Jérôme Faure, whose expertise and diligence led to a very well designed project with clear path and objectives to work on. His high quality standards permitted having an output of several international level publications that may be found at the back of this book, and patience discussing various physical aspects left the author of this manuscript as well as the other group members with few knowledge loopholes on how to aim for this quality. Secondly, nothing would have been possible without a long-term dedication and leadership by Rodrigo Lopez-Martens, whose team has for many years been devotedly developing the state-of-the-art laser system used in all the described experiments, introducing numerous highly innovative solutions on the way. Of greatest importance has been Aline Vernier, the main person behind putting the ideas of the two group leaders work in practice, always in the most clean and elegant way. Her efforts have made the setup as convenient to operate as possible, and her multi-skillfulness helped to overcome many unexpected obstacles of various sorts during the project. Diego Guénot has been the main driver of experimental execution during the first half of this thesis, when the most important breakthroughs were achieved. However, among the intense moments of fixing the never-ending final bugs he has always found time to teach the newly-arrived author the very basics of optics laboratory skills, which eventually permitted successful continuation with less human resources after his departure. Benoit Beaurepaire, the first PhD student of the project, delivered a lot of first hardware design and preliminary plasmaacceleration studies, without which the later push to a higher maturity output could not have happened. Crucial has also been the major laser system upgrade by Frederik Böhle, whose hands were later just as well replaced by Marie Ouillé. The implementation of the CEP control tool by her together with Stefan Hässler allowed obtaining one of the most interesting sets of experimental data presented in this manuscript. A lot of troubles have been avoided by using microstructured plasma targets suggested and provided by François Sylla from SourceLab. Finally, it was a great pleasure to receive a lot of attention and input from Agustin Lifschitz, one of the best experts of numerical studies in the field, bringing a significantly clearer perspective into interpreting our results.
The aim of this study is to improve the performance lev- els of gratings used as compressors in high power lasers. In the optimization process, particular attention is paid to the limitations of the fabrication process. The optimization method selects the best proﬁles in the whole domain of the fabrication ﬂuctuation. Firstly, a numerical optimiza- tion of the eﬃciency in the 1 order in reﬂection is per- formed as a function of the thickness of the dielectric layers making up the reﬂecting stack, and as a function of the grating proﬁle. As far as the mirror is concerned, it has been evidenced that the use of a metal substrate enables a substantial reduction in the number of dielectric bilayers in the stack, and that the mirror stack must be cen- tered at the median angle of the 0 and the 1 orders. As far as the grating proﬁle is concerned, the slope angle must be increased up to 90 which corresponds to a lamellar proﬁle. Averaged eﬃciencies over the entire surface of the grating higher than 99% can be obtained with only 7 bilayers. Sec- ondly, for each selected proﬁle, the maximum of the square of the electric ﬁeld value is calculated inside the solid mate- rial. Plots of these maximums as a function of the grating proﬁle reveal a decrease by a factor close to 2.5 when groove thickness and width values are increased. As a con- clusion, among the proﬁles selected for their good diﬀrac-
driving wavelength and x is the coordinate in the target normal direction. Technically this requires a highly focusable terawatt-class driver laser with a temporal contrast of > ∼ 10 10 . These conditions are typically met by Joule-class amplifier chains with
dedicated contrast filters [9, 10] and operating at ∼ 10 Hz repetition rate [4–9]. Many applications as well as parametric studies of this regime would benefit from a higher repetition rate. At LOA, we have developed a unique laser chain with power- scaled hollow-core-fiber postcompression system  operating at 1 kHz repetition rate. Using this kHz-laser, which achieves ultra-high intensties with few-mJ pulse en- ergy and few-cycle pulse duration, we have demonstrated laser-plasma interaction in the relativistic regime through laser-wakefield acceleration of electrons both in under- dense gas jets [11, 12] and in the underdense part of a smooth plasma density gradient on a plasma mirror . Here we report on the first experimental demonstration of relativistic SHHG at kHz-repetition rate, the arguably most demanding application in terms of laser performance as it depends critically on the spatio-temporal pulse quality and the temporal contrast.
3.4. SUPA, Department of Physics, University of Strathclyde
An Advanced Laser-PlasmaHigh-energy Accelerators towards X-rays (ALPHA-X) accelerator beam line has been commissioned [ 12 ]. A Titanium:Sapphire laser pulse centered at a wavelength of 800 nm with full-width at half-maximum duration of 36 fs and peak intensity of 2 × 10 18 W · cm −2 is focused to a 20 µm waist at the leading edge of a 2 mm diameter Helium gas jet to form a relativistic self-guided plasma channel. The electron beams produced are initially collimated using a triplet of miniature permanent magnet quadrupoles of fixed gradients of 500 T/m. A triplet of electromagnetic quadrupoles then focuses the beam through the undulator with gradient ∼ 2.4 T/m. Undulator output radiation is detected using a vacuum scanning monochromator and a CCD camera. The energy distribution measured has mean central energy of 104 MeV, with a 5% relative energy spread and contains a mean charge of 1.1 ± 0.8 pC. The mean spectral bandwidth of the radiation is 69 ± 11 nm corresponding to a relative bandwidth of 32 ± 7%, decreasing to as low as 16%.
out significant lengthening of the pulse due to numerical dispersion for relatively low spatial resolution (see, e.g., Ref. [ 15 ] for a study of dispersion characteristics of high order FDTD solvers). We note that our probe pulse in- tensity is four orders of magnitude higher than typical experimental values [ 7 ] in order to have a better signal-to- noise ratio in the simulated shadowgrams (in PIC simula- tions the noise level in the electromagnetic fields is much higher than in experiments due to the discretization by macroparticles). However, we verified that varying the probe intensity by one order of magnitude does not al- ter the shadowgrams, i.e., the results can be considered linear in the probe intensity (as one would indeed expect for non-relativistic probe pulses). We note that one has to be careful when using increased probe intensity if ion- ization effects are included in the simulations (neglected in the present study).
Figure 2 | Experimental evidence of vacuum laseracceleration. Panel a, Typical experimental angular distribution of electrons emitted from plasma mirrors into the vacuum, measured with the LANEX screen. It consists of a broad emission cone (blue disk), which is strongly modulated by two main patterns. One is a well defined hole (in white) around the reflected laser beam (whose size and position in the detection plane are indicated by the dashed circle), due to the ponderomotive scattering of electrons after their ejection from the plasma mirror. The other is a bright peak (in red), right on the edge of this hole, due to VLA of a fraction of these electrons. Line-outs of the distribution along the dashed lines are plotted in the side panels, and the direction normal to the plasma mirror surface corresponds to θ '960 mrad (not shown). b, Electron spectra measured at two different locations in the beam (the horizontal error bars represent the spectrometer resolution). These locations are indicated by the blue circle and the red square in a, that respectively correspond to the blue and red curves of b. All the features of a,b were very robust experimentally, being observed on all shots performed in similar experimental conditions (see Supplementary Information). c,d, Same quantities as in a,b, now obtained from numerical simulations based on a 3D test particle model. The dashed curve in d shows the initial electron energy distribution used in this model. As can be seen from the green and red spectra in d, the model shows that electrons are accelerated by VLA from 1.5 to 10 MeV, resulting in a sevenfold energy gain.
In our contribution, we develop a numerical study with the aim to simulate the occurring phenomena when a sample is moved under a laser beam, and to obtain a cutting kerf at given laser operating parameters, under normal atmospheric conditions. A 3D modeling is con- ducted by implementing the finite volume method through Ansys-Fluent CFD softwares to solve the hydrodynamic governing set of equations. The volume-of-fluid (VOF) method is applied for liquid–gas interfaces tracking, and suitable boundary conditions and source terms are intro- duced in the Fluent calculation process. By solving the unsteady Navier–Stokes, energy, and the VOF equations, we obtain the kerf profile formation, and we observe the generated periodical movement (humps) on the cutting front for given process velocity conditions. The model involves the implementation of a local absorbed laserintensity depending on the tilt angle. At low cutting speeds, the shelf structure formation is obtained by using our model, in accordance with the experimental obser- vations of Hirano and Fabbro , and as predicted by Golubev . The main data used in the simulation concern the temperature-dependent physical properties of the workpiece, the laser beam parameters, and the sur- rounding air conditions. A quite complex problem is thus considered, where standard feature of the CFD Fluent code must be enhanced, and for that, we have developed numerical procedures called user-defined functions (UDF), working interactively with Fluent in order to solve numerically the governing equations.
2 Dipartimento di Fisica “G. Occhialini,” Università degli Studi di Milano-Bicocca, Piazza della Scienza 3, 20126 Milano, Italy 3
CEA, DAM, DIF, 91297 Arpajon, France
共Received 23 March 2009; revised manuscript received 9 December 2009; published 11 March 2010 兲 Experimental measurements of proton acceleration with highintensity and high-contrast short laser pulses have been carried out over an order of magnitude range in target thickness and laser pulse duration. The dependence of the maximum proton energy with these parameters is qualitatively supported by two- dimensional particle-in-cell simulations. They evidence that two regimes of proton acceleration can take place, depending on the ratio between the density gradient and the hot electron Debye length at the rear target surface. As this ratio can be affected by the target thickness, a complex interplay between pulse duration and target thickness is observed. Measurements and simulations support unexpected variations in the laser absorption and hot electron temperature with the pulse duration and laserintensity, for which density profile modification at the target front surface is the controlling parameter.
The aim of this work is to analyze the effect of the turbulent gaseous cooling of a laser amplifier on the wavefront of the laser beam using TrioCFD, a code developed by the CEA . The mean phase distortion could be compensated by adaptive optics, but not the fluctuations. It is therefore crucial to assess their impact on the laser beam quality in order to provide some guidance for future design and operation. As the statistical approach (i.e. Reynolds-Averaged Navier-Stokes simulations) is inadequate to investigate the phase distortion fluctuations, Large Eddy Simulations are carried out. To represent the cryogenic cooling of a high power and high repetition rate laser, a plane turbulent open channel flow is first simulated. The fluid, gaseous helium at 80 K, cools two solid slabs made of Yb:YAG amplifying crystal. The resulting temperature distribution is analyzed from the optical point of view. The paper is organized as follows: the section 2 describes the nu- merical methods, both for the thermo-hydraulics and the optics aspects; the section 2.2 assesses the numerical model by comparing dynamical and thermal profiles to the literature; the section 3.1 presents the temperature field in the amplifier and the section 3.2 investigates its effect on the laser beam; finally the section 3.3 deals with the consequences of an increased optical heating of the slabs.
soumis ` a son champ ´electrique y oscillent ` a des vitesses relativistes (´equation 2.9-p.13). ´
Etant donn´ee la bri`evet´e de l’impulsion (τ < 1ps), la composante du plasma la plus active au cours de l’interaction est celle ´electronique (les composantes ioniques ont des fr´equences plasmas – ´eq. 3.1-p.15 – inf´erieures). Lors de l’interaction laser - plasma, diff´erents m´ecanismes collectifs se produisent et permettent le couplage de l’´energie laser en ´energie cin´etique communiqu´ee aux ´electrons (absorption collisionnelle, chauffage j × B, absorption r´esonante, chauffage de Brunel): leurs importances relatives est li´ees aux param`etres du plasma pr´esent sur la surface ´eclair´ee. Les ´electrons les plus chauds ainsi cr´e´es ` a la surface traverse la cible. Les courants associ´es sont si forts que les ´electrons peuvent se propager dans la cible non perturb´ee de fa¸con tr`es collimat´ee. Lors de leur ´emergence en face arri`ere, celle-ci devient rapidement ionis´ee et transform´ee en plasma qui va se d´etendre dans le vide: son expansion est responsable de l’acc´el´eration ionique et de la formation de faisceaux d’ions tr`es collimat´es dont les ´energies augmentent du fait de la s´eparation de charge produite ` a l’interface plasma/vide du fait de la s´eparation de masse. Plus la temp´erature ´electronique est grande et plus la s´eparation de charge est importante, et, par l` a mˆeme la valeur du champ ´electrique acc´el´erateur. Ce m´ecanisme, est connu sous le nom du “TNSA” (Target Normal Sheath Acceleration, acc´el´eration de gaine ` a la normale ` a la cible) et il est, avec les param`etres explor´es dans ma th`ese, le m´ecanisme le plus efficace d’acc´el´eration ionique.
3.3.4. Modified VIP limiter
The Modified VIP limiter is constructed to add a new constraint (see Fig. 3.11 (c)) for the minimal vector magnitude, however, only in situations, where it is suitable. At first, suppose one–dimensional situation with only two vectors in the opposite directions having the same magnitude. In this case, the VIP set of these vectors is a line segment and after a restriction on the minimal magnitude (absolute value in 1D), it would not allow any contribution of the high–order part of piecewise linear reconstruction. However, standard limiters in 1D, e.g. (3.13) are applied for the value itself instead of its magnitude (absolute value), allowing the absolute value of limited reconstruction to be smaller than the absolute value in the neighboring nodes. The similar situation occurs in 2D when e.g. the origin of the coordinate system falls inside the CH of the vectors in the neighboring cell’s. In this case, any restriction in on minimal vector magnitude does not make a good sense. Therefore, we choose the following condition for application of the minimal magnitude correction of VIP. If there exist any rotated coordinate system, in which all limiting vectors are located in one quadrant, then these vectors have a small angular discrepancy and the minimal magnitude condition is applied.
Commissariat a` l’e´nergie atomique, DAM, DIF, Arpajon F-91297, France (Received 20 March 2013; published 23 August 2013)
Laser-wakefield acceleration constitutes a promising technology for future electron accelerators. A crucial step in such an accelerator is the injection of electrons into the wakefield, which will largely determine the properties of the extracted beam. We present here a new paradigm of colliding-pulse injection, which allows us to generate high-quality electron bunches having both a very low emittance (0:17 mm mrad) and a low energy spread (2%), while retaining a high charge (100 pC) and a short duration (3 fs). In this paradigm, the pulse collision provokes a transient expansion of the accelerating bubble, which then leads to transverse electron injection. This mechanism contrasts with previously observed optical injection mechanisms, which were essentially longitudinal. We also specify the range of parameters in which this new type of injection occurs and show that it is within reach of existing high- intensitylaser facilities.
reached the plasma surface, after which it gradually damps. The maximum value reached is ⬃4.
Actually, the relatively low value of the instantaneous amplification factor obtained here for the surface plasma wave results from a saturation effect of the field amplitude due to the very efficient particle heating mechanism 共as it will be seen below兲, which corresponds to a strong damping of the surface plasma wave. Inspection of the density profile and comparison with motionless ion simulations in the same conditions indicate that, in the time scale considered here, a possible saturation mechanism for the surface plasma wave through any modification of the density profile due to the hydrodynamic expansion of the plasma toward the vacuum can be ruled out. In the same way, relativistic detuning, as mentioned above, cannot explain such a low amplification factor. 18 , 19 , 27 In fact, we can infer that the wave damping is mainly due to the strong acceleration of the electrons in the field of the surface plasma wave. This statement can be veri- fied by calculating the total energy acquired by those elec- trons that have left the surface inside and outside the plasmaat maximum amplitude of the laser pulse that occurs at time t = 438 0 −1 . The average kinetic energy per particle is ap- proximately given by 具E典⬃0.01m e c 2 . Thus, the total kinetic energy is E Tot ⬃具E典n eSL pl , where S is the transverse size of
September 17, 2010 9:28 WSPC - Proceedings Trim Size: 9.75in x 6.5in main
As lasers actually are the most powerful sources of electromagnetic energy on earth, we present here analytical estimates and numerical simulations of generation of the HFGW in interaction of high power laser pulse with a medium in different geometries. First, during the laserplasma interaction, a strong shock driven by the ablation pressure is generated in the bulk material. In this configuration, material is accelerated in the shock front and in the ablation zone. Because of a short laser pulse duration (ns) GW are generated in the GHz domain. During the laser interaction with a planar thick foil (more than 100µm thickness) the laser launch a shock with a velocity V s , this shock accelerates the medium along the z axis and produces a
Vd) Use for near Coulomb Barrier Physics.
This intermediate construction step with a 40MV acceleration potential, leads for A/q=3 to 14.5 A MeV, an energy sufficient for all near-barrier physics. This would allow users to do fusion-evaporation physics, like high spin physics. Remember that beam intensities of 1mA would be available with appropriate ion sources. This would permit research in the domain of low cross-section phenomena, such as formation of super heavy elements. The very highintensity would allow the production of rare fusion-evaporation products in thin targets. Coupled to with an appropriate separator, this could provide a mean of producing secondary beams for decay properties and reaction studies.
W x 共I兲 苷 W y 共I兲 ¿ W z 共I兲. In this intermediate case,
ionization is caused only by the component of the field in a molecular plane. In both cases, linear behavior with the same slope as the anisotropic case is predicted athighintensity (Fig. 1). Extrapolation back to the ln共I 0 兲