Simulation of Plasma Wakefields and Weibel Instability of Electron Beams in Plasma Using Codes LSP and OSIRIS

Transcription

1 Simulation of Plasma Wakefields and Weibel Instability of Electron Beams in Plasma Using Codes and OSIRIS 15 Electron density Beam density Plasma density A. Solodov, C. Ren, J. Myatt, and R. Betti University of Rochester Laboratory for Laser Energetics W. B. Mori, UCLA 47th Annual Meeting of the American Physical Society Division of Plasma Physics Denver, CO October 25

2 Summary We benchmarked two PIC codes and OSIRIS simulating two physical problems Excitation of plasma wakefields by electron beams Both codes correctly simulate the plasma wakefield excitation provided a sufficient spatial and temporal resolution. correctly simulates the collisional plasma wave damping. Weibel instability of electron beams in plasma Both codes simulate qualitatively similar the Weibel instability. The most agreement is found for the OSIRIS and particle mode. The degree of plasma heating is different in particle and hybrid simulations. The total energy is not conserved in hybrid simulations. TC723

3 Codes and OSIRIS OSIRIS (developed at UCLA) Explicit PIC (product of MRC, Albuquerque) Explicit or implicit Particle (PIC) or fluid (hybrid PIC) Courant condition for the electromagnetic fields (Dt < Dx/c) Numerical heating if Dx > 3m D TC724 Courant condition is unnecessary in the implicit mode; nonresolved temporary modes are damped, remaining limitation Dt < Dx/o te Implicit PIC numerical heating or cooling if Dx > m D Implicit hybrid PIC no numerical heating or cooling

4 We simulate a linear plasma wakefield excited by a Gaussian electron beam in a plasma (2-D case) Electron density 15 6 Electron beam with a maximum density n b =.1 n p, -1 width w=. 5 k p `kp = ~ p cj, and velocity close to c Plasma temperature T p = 51 ev Theory & dn/n =.79* TC725 *See, for instance, E. Esarey et al., IEEE Trans. Plasma Sci. 24, 252 (1996).

5 and OSIRIS correctly simulate the plasma wakefield provided a sufficient spatial and temporal resolution One-component electron plasma Dx =. 25k -1 p, Dy =. 63k -1 p bbeam width. 5k -1 p l, Dt =. 3 ~ -1 p. dn n eex m ~ p c ee y m ~ p c hybrid 15 6 particle OSIRIS TC

6 The plasma wave collisional damping is simulated correctly in (except for the Coulomb logarithm which should be corrected) ~ Two-component plasma, n e = n i = 1 22 cm 3, Z i = / 2-3/ 2 o ei ~ pe = # 1 Zine Te ln K hybrid 15 dn n eex m p c ee y m ~ p c 6 T e, ev particle TC

7 We have performed 2-D simulations of Weibel instability of an electron beam as a FI-relevant benchmarking problem 4 Beam density Plasma density 4 TC7212 Simulation parameters n p = 1 22 cm 3, n b =.1 n p, cb b = 2.8, T e = 5 kev Dx = Dy =.4 c/~ p Dt min = Dx/2c

8 and OSIRIS simulations with immobile ions show similar electron-beam density profiles t = ~ - p 1 t = 47 ~ - p 1 t = 188 ~ - p 1 t = 752 ~ - p 1 t = 154 ~ - p 1 hybrid 4 4 particle OSIRIS TC

9 and OSIRIS simulations with immobile ions show similar plasma-density profiles t = ~ - p 1 t = 47 ~ - p 1 t = 188 ~ - p 1 t = 752 ~ - p 1 t = 154 ~ - p 1 hybrid 4 4 particle OSIRIS TC

10 hybrid simulations show poor energy conservation hybrid particle OSIRIS ~ ~ ~ TC726

11 Implicit simulations with large time steps show more filaments and distortions at the late stage of Weibel instability: Dt < Dx/v e is not satisfied particle D - t = 2 ~ p 1 4 t = ~ - p 1 t = 47 ~ - p 1 t = 188 ~ - p 1 t = 752 ~ - p 1 t = 154 ~ - p 1 4 D - t = ~ p 1 D - t =. 2 ~ p 1 TC

12 and OSIRIS simulations with mobile ions (H + ) show qualitatively similar beam-density profiles t = ~ - p 1 t = 47 ~ - p 1 t = 188 ~ - p 1 t = 752 ~ - p 1 t = 154 ~ - p 1 hybrid 4 4 particle OSIRIS TC

13 hybrid simulations with mobile ions predict stronger plasma density compressions at the late stage of the instability t = ~ - p 1 t = 47 ~ - p 1 t = 188 ~ - p 1 t = 752 ~ - p 1 t = 154 ~ - p 1 hybrid 4 4 particle OSIRIS TC

14 hybrid simulations with mobile ions predict stronger ion-density compressions at the late stage of the instability t = ~ - p 1 t = 47 ~ - p 1 t = 188 ~ - p 1 t = 752 ~ - p 1 t = 154 ~ - p 1 hybrid 4 4 particle OSIRIS TC

15 The energy is not conserved in hybrid simulations hybrid particle OSIRIS ~ ~ ~ TC728

16 Summary/Conclusions We benchmarked two PIC codes and OSIRIS simulating two physical problems Excitation of plasma wakefields by electron beams Both codes correctly simulate the plasma wakefield excitation provided a sufficient spatial and temporal resolution. TC723a correctly simulates the collisional plasma wave damping. Weibel instability of electron beams in plasma Both codes simulate qualitatively similar the Weibel instability. The most agreement is found for the OSIRIS and particle mode. hybrid simulations with mobile ions show stronger density compressions at the late stage of Weibel instability than particle simulations. The degree of plasma heating is different in particle and hybrid simulations. The total energy is not conserved in hybrid simulations. implicit simulations with a time step exceeding the Courant limit show more filaments at the late stage of Weibel instability than simulations with a small time step.

17 Even better spatial and temporal resolution is necessary in the fluid mode to overcome numerical damping if the particle momenta are well averaged on the grid Larger values of the averaging parameter have a stabilizing effect on grid noise and are recommended in documentation. dn n eex m ~ p c ee y m ~ p c hybrid fluid electron streaming factor = hybrid fluid electron streaming factor = TC729

18 hybrid simulations show poor energy conservation hybrid Fluid electron streaming factor =.1 hybrid particle OSIRIS ~ ~ ~ ~ TC721