charged particle transport in turbulent media
play

Charged particle transport in turbulent media F. Spanier A. - PowerPoint PPT Presentation

Charged particle transport in turbulent media F. Spanier A. Ivascenko S. Lange C. Schreiner Center for Space Research, North-West University Astronum 2013, Biarritz Motivation Particle transport in heliosphere and ISM What is the


  1. Charged particle transport in turbulent media F. Spanier A. Ivascenko S. Lange C. Schreiner Center for Space Research, North-West University Astronum 2013, Biarritz

  2. Motivation Particle transport in heliosphere and ISM What is the microphysics of transport? Turbulent magnetic fields ⇒ charged particle scattering Felix Spanier (NWU) Astronum 2013, Biarritz 2 / 26

  3. Motivation Particle transport in heliosphere and ISM What is the microphysics of transport? Turbulent magnetic fields ⇒ charged particle scattering Felix Spanier (NWU) Astronum 2013, Biarritz 2 / 26

  4. Fokker-Planck Equation Particle transport is described by the Fokker-Planck equation Vlasov equation in gyrocenter coordinates Fokker-Planck-Equation ∂ F T + v µ∂ F T ∂ Z − ǫ Ω ∂ F T ∂φ = S T ( X σ , t ) + 1 � ∂ F T � ∂ p 2 D X σ X ν p 2 ∂ t ∂ X σ ∂ X ν Diffusion-convection equation Pitch angle diffusion coefficient D µµ particularly important Mean free path λ � derived from that Felix Spanier (NWU) Astronum 2013, Biarritz 3 / 26

  5. MHD Simulations Felix Spanier (NWU) Astronum 2013, Biarritz 4 / 26

  6. MHD Simulations 1e+12 1e+10 E(k) [numerisch] 1e+08 t = 17 s t = 34 s t = 51 s t = 68 s 1e+06 t = 85 s Kolmogorov-Spektrum 10000 100 1 1 10 k L / 2 π Felix Spanier (NWU) Astronum 2013, Biarritz 4 / 26

  7. Test particle Simulations Felix Spanier (NWU) Astronum 2013, Biarritz 5 / 26

  8. Test particle Simulations 1500 initial distribution distribution after >20 gyrations 1000 particle number 500 0 -1 -0.5 0 0.5 1 pitch angle Felix Spanier (NWU) Astronum 2013, Biarritz 5 / 26

  9. Wave-particle resonance Testing the interaction of particles with a single wave Inject isotropic, monoenergetic particle distribution Assume background plasma with one Alfvén wave Plot ∆ µ ( t ) vs. µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 6 / 26

  10. Simple wave-particle resonance 2 gyrations, wave amplitude δ B / B 0 = 0 . 01, QLT prediction Felix Spanier (NWU) Astronum 2013, Biarritz 7 / 26

  11. Simple wave-particle resonance 2 gyrations, wave amplitude δ B / B 0 = 0 . 01 Felix Spanier (NWU) Astronum 2013, Biarritz 7 / 26

  12. Simple wave-particle resonance 10 gyrations, wave amplitude δ B / B 0 = 0 . 001 Felix Spanier (NWU) Astronum 2013, Biarritz 7 / 26

  13. Simple wave-particle resonance 2 gyrations, wave amplitude δ B / B 0 = 0 . 1 Felix Spanier (NWU) Astronum 2013, Biarritz 7 / 26

  14. Simple wave-particle resonance 50 gyrations, wave amplitude δ B / B 0 = 0 . 001 Felix Spanier (NWU) Astronum 2013, Biarritz 7 / 26

  15. Turbulent transport Testing the interaction of particles with turbulence Undisturbed turbulence Excited turbulence S I E mag (t 1 ) 1e+14 S II E mag (t 2 ) 1e+15 S III E mag (t 3 ) S IV Gaussian fit Kolomogorov spectrum Gaussian fit 1e+12 Gaussian fit 1e+10 E(k || ) [numerical units] E(k) [numerical units] 1e+10 1e+08 100000 1e+06 1 10000 1e-05 100 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 1 10 k || L / 2 π k L / 2 π Not only µ 0 − ∆ µ plot, but additional � α 2 � D αα = lim 2 ∆ t t →∞ scattering angle diffusion coefficient Felix Spanier (NWU) Astronum 2013, Biarritz 8 / 26

  16. Scattering in MHD turbulence 0.14 T = 1 gyr T = 5 gyr T = 10 gyr 0.12 T = 30 gyr 0.1 0.08 D αα [s -1 ] 0.06 0.04 0.02 0 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 9 / 26

  17. Scattering in MHD turbulence Felix Spanier (NWU) Astronum 2013, Biarritz 9 / 26

  18. MHD excitation Felix Spanier (NWU) Astronum 2013, Biarritz 10 / 26

  19. Scattering with excited modes I k � 0.3 T = 1 gyr T = 5 gyr T = 10 gyr 0.25 n = -1 n = 0 n = 1 0.2 D αα [s -1 ] 0.15 0.1 0.05 0 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 11 / 26

  20. Scattering with excited modes I k � Felix Spanier (NWU) Astronum 2013, Biarritz 11 / 26

  21. Scattering with excited modes II k � , k ⊥ 1.4 T = 1 gyr T = 5 gyr T = 10 gyr 1.2 1 0.8 D αα [s -1 ] 0.6 0.4 0.2 0 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 12 / 26

  22. Scattering with excited modes II k � , k ⊥ Felix Spanier (NWU) Astronum 2013, Biarritz 12 / 26

  23. Quasilinear comparison 1 0.01 0.0001 D αα [s -1 ] 1e-06 1e-08 SQLT background SQLT driven stage SQLT decay stage Particle 30gyr background 1e-10 Particle 30gyr driven stage Particle 30gyr decay stage 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 13 / 26

  24. Turbulent transport Good agreement of testparticle and QLT results in D αα SQLT misses Cherenkov resonance ( n = 0) Limited spectrum yields resonance gap Finite simulation time results in broadened resonances Felix Spanier (NWU) Astronum 2013, Biarritz 14 / 26

  25. Derivation of coefficients µ − ∆ µ plots show the physics of scattering For further use D µµ or D α α is needed! Determination is - as seen - flawed, especially for strong turbulence. Felix Spanier (NWU) Astronum 2013, Biarritz 15 / 26

  26. Running coefficient Derivation of D µµ via its definition: (∆ µ ) 2 ≈ (∆ µ ) 2 t ≫ t 0 D µµ = lim 2 ∆ t 2 ∆ t t →∞ 0.14 T = 1 gyr T = 5 gyr T = 10 gyr 0.12 T = 30 gyr 0.1 0.08 D µµ [s -1 ] 0.06 0.04 0.02 0 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 16 / 26

  27. Kubo-Formalismus Integration along trajectories: t 1 � � D µµ = ∆ t ˙ µ ( t 0 ) ˙ µ ( t ) N T T t 0 = 0 0.4 T = 1 gyr T = 5 gyr 0.3 T = 10 gyr T = 30 gyr 0.2 0.1 D µµ [s -1 ] 0 -0.1 -0.2 -0.3 -0.4 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 17 / 26

  28. Matrix inversion Discretisation of the diffusion equation for each µ n : ∂ t f = D n + 1 µµ − D n − 1 µµ ∂ µ f + D n µµ ∂ µµ f 2 · ∆ µ Tridiagonal matrix equation:  ∂ µ f 0  ∂ µµ f 0 0 0 D 0 ∂ t f 0     2 ∆ µ µµ  ...  − ∂ µ f 1 D 1 ∂ t f 1 ∂ µµ f 1   0  µµ     2 ∆ µ  ·  =  .   .    . . ... ...     ∂ µ f n − 1 . .   0      2 ∆ µ D n ∂ t f n   − ∂ µ f n µµ ∂ µµ f n 0 0 2 ∆ µ Matrix inversion with standard methods! Felix Spanier (NWU) Astronum 2013, Biarritz 18 / 26

  29. Integration method Ensemble averaging over several simulations would be statistically correct, but expensive Fitting or smoothing is usually required Integration of the diffusion equation over µ smoothes time derivatives � µ ∂ f T ( µ, t ) d µ = D µµ ( µ ) ∂ f T ( µ, t ) = − j µ ( µ ) ∂ t ∂µ − 1 Diffusion coefficients are calculated via the integration of µ -stream Felix Spanier (NWU) Astronum 2013, Biarritz 19 / 26

  30. Results 0.15 0.1 D µµ [s -1 ] 0.05 0 MIV background turb polynomial fit MIV background turb integration -0.05 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 20 / 26

  31. Results 0.25 0.2 0.15 D µµ [s -1 ] 0.1 0.05 0 MIV peaked turb polynomial fit MIV peaked turb integration -0.05 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 µ 0 Felix Spanier (NWU) Astronum 2013, Biarritz 20 / 26

  32. Wave-Particle Interaction with PiC Physical Background: magnetized plasma in heliosphere / solar wind thermal background plasma non thermal component of energetic particles resonant scattering of particles on plasma waves Fermi-II-acceleration Felix Spanier (NWU) Astronum 2013, Biarritz 21 / 26

  33. Wave-Particle Interaction with PiC Physical Background: magnetized plasma in heliosphere / solar wind thermal background plasma non thermal component of energetic particles resonant scattering of particles on plasma waves Fermi-II-acceleration Numerical Setting: magnetized thermal background plasma one excited wave mode population of relativistic test particles Felix Spanier (NWU) Astronum 2013, Biarritz 21 / 26

  34. PiC simulations E y in x direction (transverse) Simulation Setup: 0.1 L-Mode R-Mode 10 1 excitation of low frequency 0.08 10 0 wave (ideal case: Alfvén 0.06 ω ( ω pe ) wave ) 10 -1 0.04 → huge number of cells and 10 -2 0.02 timesteps required 0 10 -3 -0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02 k (1/cm) use resonance condition to determine the parallel component of the test particles’ velocities → parameters k w , ω w and Ω i give constraints → resonant pitch angle µ res is free initialize monoenergetic test particles ( | v | = v � ) with isotropic angular distribution → resonant scattering only for particles with µ = µ res Felix Spanier (NWU) Astronum 2013, Biarritz 22 / 26

  35. Pitch Angle Diffusion Scatter Plots: peaks at ± µ res left peak: lefthanded wave right peak: righthanded wave ballistic transport (smaller peaks) QLT approximation Felix Spanier (NWU) Astronum 2013, Biarritz 23 / 26

  36. Pitch Angle Diffusion with PiC lefthanded wave mode Felix Spanier (NWU) Astronum 2013, Biarritz 24 / 26

  37. Pitch Angle Diffusion with PiC righthanded wave mode Felix Spanier (NWU) Astronum 2013, Biarritz 25 / 26

Recommend


More recommend