Laser Fusion Experiments at 4 TW

Abstract

DT-filled glass microspheres have been imploded at power levels exceeding 4 TW using the Lawrence Livermore Laboratory 1.06-/um Argus laser. Thermonuclear neutron yields in excess of 1.5 x lo^ have been observed implying a DT burn efficiency of 1.6 x lo" ^ Neutron and a time-of-flight measurements indicate DT burn temperatures of 4-8 keV, implying that a DT gain of approximately 10"^ and a nr of 10^^ were obtained, As part of the effort to understand the physics of laser imploded targets, a series of experiments examining both basic laser-plasma interaction phenomena and the parameter space of exploding-pusher experiments have been performed over the last three years, both at Lawrence Livermore Laboratory and other laboratories.^"^ A laser irradiated target is said to operate in the exploding-pusher^ mode when the pusher significantly decompresses in the process of compressing the fueL This is characteristic of a high rate of energy addition to the pusher. Laser absorption mainly by collective processes producing super thermal electrons, early energy deposition in the hsell by these super thermal electrons, a near supersonic electron thermal wave driven by electron thermal conduction from the laser-absorption region, and significant shock compression of the fuel causing a large entropy change are other characteristics of the exploding-pusher mode. This results in a limited density increase, but significant heating of the fuel. Early exploding-pusher experiments with DT-filled glass microshells indicated that for fixed DT fill, the target performance, measured in terms of neutron yield, should increase^° as ^^10/3^2, HcJv)T -1/2 Here t*o is the target radius, w the shell wall thickness, T the time-and spaceaveraged final DT fuel temperature, and {uv) the Maxwell averaged DT cross section. It is assumed thatr is proportional to the useful specific absorbed energy, SQ, The useful fraction is essentially the absorbed energy corrected for any temporal mismatch between the input laser pulse length and the characteristic target-implosion time scale, and is thus the time in the laser pulse beyond which further absorption can no longer influence the final implosion phase. This time is found empirically to be roughly determined by the amount of energy absorbed until the pusher has traversed '^ 30% of the initial target radius, with the instantaneous pusher velocity assumed proportional to the energy absorbed up to that time. Since pusher velocities are -(2.5 -^.5)x 10^ cm/sec,^'^^ a 90-/im-diam target would find nearly all the energy ^'useful" for laser pulses with a full width at half-maximum (FWHM) ^ 40 ps. Experiments performed at KMS Fusion, Inc. also gave results consistent with this description.^^ Targets irradiated at a fixed peak power, with pulse lengths increasing in steps of 40 ps up to 240 ps, showed no increase in neutron yield for pulse lengths beyond 80 ps. Both to confirm the "useful energy'* hypothesis and to extend our data 1570

Assessment of ion kinetic effects in shock-driven inertial confinement fusion implosions using fusion burn imaging

“…In this case, ion diffusion across the fuel-shell interface also drastically reduces (by a factor of ∼30) the overall predicted yields, and preferentially reduces the number of reactions near the shell. The D and 3 He ions, which have mean free paths in the D 3 He fuel of 460 µm and 140 µm, respectively, readily escape the fuel region (of radius 90 µm) and penetrate into the shell, where fusion reactions are largely suppressed.…”

“…As in dued, physical ion viscosity was used. To account for ion transport effects that are expected to be significant in these implosions, some lasnex simulations were also run with a model of classical ion diffusion included, 37 which models diffusion of D, 3 He, Si, and O ions across the D 3 He fuel/SiO 2 shell interface. The comparison of experimental burn profiles with lasnex simulations excluding and including ion diffusion, which provides strong evidence of the significance 31 While emissivity is the preferred and more physically intuitive quantity, and is more directly calculated in the PCIS analysis, some simulations give instead surface brightness (the line-of-sight integral of the radial emissivity profile) for comparison to the experimental data.…”

“…This differential in the higher-density implosions is likely indicative of ion temperature gradients, which give rise to differences between D 3 He and DD reaction profiles due to the different temperature sensitivities of the two reactions (D 3 He being more strongly weighted by the hotter regions of the fuel). In lowerdensity implosions, for which purely hydrodynamic codes predict ∆R 50 ∼0, the persistence of a differential between burn radii could be a signature of ion diffusion or other kinetic ion transport effects, which are expected to be quite significant in this N K >1 plasma and which allow for deuterium ions to be transported farther from the center of the implosion than 3 He ions. These ion species separation effects 19 may also contribute to ∆R 50 in the higher-density implosions.…”

“…Shock-driven exploding-pusher implosions [2][3][4][5] are an ideal platform to isolate and probe ion kinetic and multiple-ion-fluid effects in ICF implosions, 6,7 as the bulk of fusion reactions (and diagnosis of implosion conditions) occurs when the plasma is at kinetica) Electronic mail: mros@lle.rochester.edu; Present institution: Laboratory for Laser Energetics, University of Rochester. like conditions.…”

“…Previous experimental and theoretical work has explored deviations from hydrodynamic behavior in ICF implosions. Yield anomalies observed in mixed-fuel implosions such as D 3 He, 8 DT, 9 and DT 3 He 10 have been partially explained by multipleion models that include barodiffusion, electrodiffusion, and thermodiffusion. [11][12][13][14] Models allowing for Knudsenlayer losses of suprathermal ions and a deviation from Maxwellian ion distributions 15,16 have produced better agreement with the results of shock-driven implosions, and kinetic simulations have been found to predict weaker shock-front gradients and shock-induced fusion yields than in hydrodynamic simulations, 17,18 in better agreement with experimental results.…”