The 1-D Cryogenic Implosion Campaign on OMEGA

Transcription

1 The 1-D Cryogenic Implosion Campaign on OMEGA Yield Exp (#1 14 ) D campaign neutron yields.2 R. Betti University of Rochester Laboratory for Laser Energetics LILAC 4 8. V MLILAC imp stag Tmin 45. # 113 d n d n 4. 1 d T n max (#1 14 ) 59th Annual Meeting of the American Physical Society Division of Plasma Physics Milwaukee, WI October 217 1

2 Summary High-adiabat (a. 6 to 7) implosions achieved the highest fusion yields of on OMEGA with high predictability using a statistical mapping model The ongoing 1-D campaign is building an experimental database of high-adiabat, low-convergence implosions (CR* ~ 12 to 14) The highest fusion yield of (tripled in the past year) is achieved by increasing the target outer diameter and reducing the DT ice thickness Based on these recent results, a roadmap is being developed to find the optimum target and laser pulse TC1374a * CR: convergence ratio 2

3 Collaborators V. Gopalaswamy, J. P. Knauer, A. R. Christopherson, D. Patel, K. M. Woo, A. Bose, K. S. Anderson, T. J. B. Collins, S. X. Hu, D. T. Michel, C. J. Forrest, R. Shah, P. B. Radha, V. N. Goncharov, V. Yu. Glebov, A. V. Maximov, C. Stoeckl, F. J. Marshall, M. J. Bonino, D. R. Harding, R. T. Janezic, J. H. Kelly, S. Sampat, T. C. Sangster, S. P. Regan, and E. M. Campbell University of Rochester Laboratory for Laser Energetics M. Gatu Johnson, J. Frenje, C. K. Li, and R. Petrasso Plasma Science and Fusion Center Massachusetts Institute of Technology 3

4 The 1-D Campaign Implosions 4

5 The 1-D Campaign developed a database of more-predictable, lower-convergence, high-adiabat implosions 43 nm OMEGA CD DT ice DT gas 8 nm 5 nm 372 nm Power (TW) a Time (ns) 7768 (a 3) 882 (a 7) Density (g/cm 3 ) a Time (ns) CR = 18 TC1396b 2 CR = Radius (nm) 5

6 Systematic hydro-stability, convergence, shock-timing, and preheat tests are conducted to assess degradation mechanisms and proximity to stability cliffs Shock-timing test SSD* on/off (RT** test) Power (TW) Time (ns) Time (ns) Convergence test Hot electrons on/off Power (TW) 2 1 CR = 12 CR = 16 Power (TW) TC1395b Time (ns) Time (ns) * SSD: smoothing by spectral dispersion ** RT: Rayleigh Taylor 6

7 To date, high-adiabat, low-convergence implosions achieved the highest fusion yields on OMEGA with areal densities up to ~65% of the highest compression shots Yield Exp (#1 14 ) High a > 5 Low a # 4 tr exp (mg/cm 2 ) High a > 5 Low a # Calculated CR (LILAC*) Calculated CR (LILAC) TC13677b * LILAC is the LLE 1-D hydrodynamics code 7

8 As expected, the main dependence of the yield is on the implosion velocity Yield Exp (#1 14 ) a # 4 a > 5 Yield Exp (#1 14 ) a # 4 a > Calculated V imp (km/s) Calculated IFAR 2/3 OMEGA implosions are testing the IFAR* limits. TC13678a * IFAR: in-flight aspect ratio 8

9 The Statistical Mapping Model as a Predictive Tool V. Gopalaswamy and R. Betti, CO8.1, this conference. 9

10 A reliable predictive capability is required to find the optimum implosion Conventional predictive capability is based on using simulations to design future experiments Despite advances in simulations, codes are still not predictive enough to enable quick progress in improving implosion performance Target specifications TC13679a Yield 43 nm CD DT ice DT gas 8 nm 5 nm 372 nm Power (TW) Implosion experiments a Time (ns) Good predictive capability tr Optimum implosion Limited predictive capability Laser pulse Implosion performance figure of merit Lawson parameter * Experimental ignition threshold factor (ITFx)** ITFx and scale with yield tr 2 # mass ^DTh * R. Betti et al., Phys. Plasmas 17, 5812 (21). ** B. K. Spears et al., Phys. Plasmas 19, (212). 1

11 By assuming that the 1-D codes do not accurately reproduce the 1-D implosion dynamics, we look for statistical correlations between experimental and simulated data 1-D physics models (LILAC) Laser pulse and target specifications are inputs to both the code and the experiment 43 nm OMEGA CD DT ice DT gas 8 nm 5 nm 372 nm Power (TW) a Time (ns) Implosion experiment: real physics Assuming 1-D implosions Code output: LILAC imp LILAC LILAC i LILAC LILAC LILAC Yield, V, tr,t, x,r Mapping Measurable parameters: Exp Exp imp Exp Exp i Exp Exp Yield, V, tr,t, x,r TC1368a Example: YieldExp = FV _ LILAC, tr LILAC, T LILAC, x LILAC i imp i New predictive capability (also valid for 3-D systematic nonuniformities) Instead of comparing each simulated parameter with its measured value, we map each measurement onto all simulated parameters through nonlinear correlations. 11

12 The combination of 1-D simulations and mapping relations provides a predictive capability as long as its validity can be extrapolated TC13681a Yield Exp (#1 14 ) Mapping onto 1-D simulated parameters 1-D campaign a ~ 6 to 7 Shots used to generate prediction LILAC Extrapolation LILAC 65. Vimp Mstag 43. # 113 d n d 4 1. Shots used to generate prediction Predicted shots Predictive capability: 1-D simulations + mapping relations n (#1 14 ) V M LILAC imp = stag LILAC = LILAC implosionvelocity LILAC stagnation mass 12

13 The higher yields were achieved through CD ablators, thinner ice, 4/6 DT mixture, modified pulse shapes, and larger diameter shells 43 nm OMEGA 5/5 DT CH DT ice DT gas 8 nm 53 nm Power (TW) nm OMEGA 4/6 DT CD DT ice DT gas 7.5 nm 42 nm 882 Y = 3.2 #1 13 tr = 126 mg/cm 2 T i = 2.6 kev Time (ns) Y = 1.2 to 1.4 #1 14 tr = 11 to 13 mg/cm 2 T i = 4.3 to 4.6 kev TC13687b 13

14 Systematic changes to the target specifications and laser pulse resulted in the expected increase in yield 1.1 Four shot days (Feb. to July 217) Yield (#1 14 ) Measured yield Prediction from mapping VLILAC MLILAC imp stag 43. # 113 d n d n 4 1. No change Power (TW) Time shift 1 2 Time (ns) D DT = constant No change Power (TW) D DT = 3 nm 2 1 Ramp change 1 2 Time (ns) D DT = constant No change OD = +4 nm D DT = constant Foot delay 1 2 Time (ns) D DT = 48 nm Outer diameter (OD) = 87 nm TC13688b D DT = 3 nm Sequential shot number D DT = 42 nm OD = 91 nm 14

15 In the last two shot days (ten shots), an ~2% yield variability is observed between repeats and also between the measured yield and prediction of mapping model 1.4 Without T i asymmetries T max /T min = With T i asymmetries TC1394 Yield Exp (#1 14 ) T max /T min = 1.2 OD = 98 nm repeats T max /T min = 1.23 T max /T min = 1.18 Sept. and Oct. 217 shots (not included in mapping) (#1 14 ) (#1 14 ) LILAC V MLILAC imp stag V MLILAC 43. # 113 imp stag d n T d n min 47. # 113 d n d n d T n Yield Exp (#1 14 ) LILAC Large temperature asymmetries suggest nonsystematic distortions possibly caused by the large DT ice rms (~1.5 nm) or target offsets max 15

16 What are the Results Telling Us? 16

17 Hydro stability SSD on/off tests show modest degradation in performance but no proximity to a stability cliff Yield (SSD off) Yield (SSD on) tr (SSD off) tr (SSD on) Yield (#1 13 SSD on) Yield (#1 13 SSD on) 15 < higher IFAR < < higher IFAR < 34 TC13685a 17

18 Including lower-adiabat implosions shows the yield increase from convergence is less than 1-D predictions while the areal density scales as 1-D 1-D theory 17. CR 1 b 13 l tr (MRS*) (mg/cm 2 ) Areal density CR 1 c 13 Exp 15. m Yield Exp (#1 14 ) Neutron yield: experiment versus LILAC param +CR Exp a # 4 a > (#1 14 ) LILAC 4 7. V MLILAC imp stag CRExp 4# 113 d n d n c m D theory CR 2 In experiments, a higher CR leads to a very modest increase in yield. TC13739a * MRS: magnetic recoil spectrometer 18

19 The shape of the hot-spot self-emission profile is used as a measure of proximity to 1-D behavior; the 1-D Campaign targets are more 1-D than high-cr targets y (nm) D DRACO simulations* : Y = : Y = 3.9 # 1 13 tr = 13 mg/cm 2 tr = 225 mg/cm 2 GMXI** GMXI** TC1394a Normalized intensity x (nm) x (nm) 1. Core self-emission lineouts Radius (nm) 2 Arbitrary untis SGfit exp = GMXI lineouts SGfit exp = 2.1 Radius (nm) 2 4 * R. Nora, Laboratory for Laser Energetics, private communication (214). ** F. J. Marshall and J. A. Oertel, Rev. Sci. Instrum. 68, 735 (1997). GMXI: gated monochromatic x-ray imager 19

20 OMEGA Roadmap and Extrapolation to NIF* Laser Energies * NIF: National Ignition Facility 2

21 A systematic approach is being used to find the optimum implosion on OMEGA Hydro-stability or laser-energy boundary Larger yield by increasing OD, better coupling, larger V imp, larger IFAR Velocity campaign Increase DT thickness Optimum implosion Tuning campaign increase tr October 3 shots: as predicted Y = 1.4 # 1 14, tr = 11 to 13 mg/cm 2 April 4 shots: as predicted Y = 7.8 # 1 13, tr = 11 mg/cm 2 TC13691a Ice thickness 21

22 OMEGA highest-yield shots extrapolate to fusion yields L23 kj at 1.9 MJ of symmetric illumination Density (g/cm 3 ) 1-D profiles from OMEGA/NIF-symmetric used in the extrapolation TC R/R() Pressure (Gbar) t (OMEGA) t (NIF) P (OMEGA) P (NIF) Shot 8726 OMEGA 28 kj.86 mm T i (kev) Hydrodynamic scaling R 17 (nm) x (ps) GPH (Gbar) Direct drive NIF 1.9 MJ 3.6 mm Yield Experiment Simulation (degraded** in 2-D to YOC =.75) 1.9-MJ extrapolation Hydro-equivalent extrapolation* V imp = const, a = const, mass ~ volume ~ E L no a kj * R. Nora et al., Phys. Plasmas 21, (214). ** A. Bose et al., Phys. Plasmas 24, 1274 (217). YOC: yield-over-clean 22

23 Summary/Conclusions High-adiabat (a. 6 to 7) implosions achieved the highest fusion yields of on OMEGA with high predictability using a statistical mapping model The ongoing 1-D campaign is building an experimental database of high-adiabat, low-convergence implosions (CR* ~ 12 to 14) By mapping the experimental results onto the simulation database, an accurate predictive tool is developed that enables the design of targets with improved performance The highest fusion yield of is achieved by increasing the target outer diameter and reducing the DT ice thickness Based on these recent results, a roadmap is being developed to find the optimum target and laser pulse Hydro-equivalent extrapolation to 1.9 MJ symmetric illumination shows fusion yields $23 kj for best performing implosions TC1374b * CR: convergence ratio 23

24 Backup 24

25 The time history of the hot-spot radius R 17% monitors deviations from 1-D behavior 6 5 Low modes (, ~2) increase the hot-spot size Intermediate modes (, ~1) decrease the size 1-D bang time, = 2 R 17% (nm) Clean, = 2, YOC = 6%, = 1, YOC = 6% Clean, = Time (ps) The time history of the hot-spot radius reveals which modes are dominant. TC1393 * A. Bose, R. Betti, and D. Shvarts, JO7.7, this conference. 25

26 The hydrodynamic scaling holds in three dimensions In-flight scaling: R+ E 13 / P + E 23 / x + E / L L L pulse V imp + const a + const RT growth factors+ const 13 L Stagnation scaling: P + const T + R 2. V + R x burn + R tr tot + R hs 3 z (nm) YOC NIF = 6% YOC X = 61% 3 NIF 2 1 OMEGA t (g/cm 2 ) Yield Y Y no a no a a no a + +. E E 145. L 35. L 1 ^ no a h TC1975e r (nm) r (nm) A. Bose et al., Phys. Plasmas 22, 7272 (215); R. Nora et al., Phys. Plasmas 21, (214). 26

27 The hydro-equivalent extrapolation uses the results of 1-D simulations combined with 2-D degradation to match the experimental observables P n /P n 1-D , = 1, = 2 T n /T n 1-D V GhsH /V GhsH 1-D Y/Y 1-D.8 1. x/x 1-D Y/Y 1-D TC1325 * A. Bose, Ph.D. thesis, University of Rochester, 217; A. Bose et al., Phys. Plasmas 24, 1274 (217). 27

28 The yield predictive capability of the mapping relations applies over a wide range of extrapolations 1. Yield Exp (#1 14 ) Shots used to generate prediction Predicted shots (#1 14 ) LILAC 43. V MLILAC imp stag 43. # 113 e o e o TC13682a 28

29 The yield s unique dependence on 1-D simulated parameters persists in 3-D as long as the 3-D dynamics are affected by dominant systematic nonuniformities Y exp = Y 1-D (1-D parameters) YOC (distortion) A = initial nonuniformities (target and/or laser) f (1-D) = amplifications of distortion caused by implosion (RT, RM*, BP**) systematic YOC= YOC A f 1D Arandom 8u ^ - h+ u f ^1-DhB s If systematic nonuniformities are dominant: Au r Au systematic random & systematic Y = Y ^1-DparametershYOC 8Au f ^1- D parametershb exp 1-D If Au systematic = constant, the yield depends only on 1-D parameters even in distorted implosions s Y exp = Y 1-D (1-D parameters) For 1-D implosions or 3-D with dominant systematic nonuniformities TC13683a *RM: Richtmyer Meshkov ** BP: Bell Plesset 29

30 A convergence test on the thin 42-nm ice targets did not show proximity to stability cliffs for low and mid modes TC13686a 43 nm 8566 CR = 12 CD DT ice Power (TW) DT gas nm 42 nm 8566 CD E L = 27.7 kj V LILAC imp = 45 km/s Time (ns) 43 nm 8529 CH E L = 26.4 kj VLILAC imp = 397 km/s CR = CH DT ice DT gas 7.5 nm 42 nm Higher-CR targets performed better than low-cr targets at the same predicted velocity. Yield Exp (#1 14 ) Shot 8529 CR = 16 tr = LILAC 42. Vimp Mstag 43. # 113 e o e 4 1. LILAC Shot 8566 CR = 12 tr = 113 o (#1 13 ) 65. 3

31 Simulated trajectories compare favorably with experimental measurements 5 Shot Shot Shot R (nm) P (#1 2 TW) t (ps) t (ps) t (ps) TC13725a 31