Introduction of temporal sub-stepping in the Multi-Level Multi-Domain semi-implicit Particle-In-Cell code Parsek2D-MLMD

Innocenti, Maria Elena; Beck, Arnaud; Ponweiser, Thomas; Markidis, Stefano; Lapenta, Giovanni

doi:10.1016/j.cpc.2014.12.004

Cited by 19 publications

(25 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in this particular case, we use sufficiently high resolutions (verified in previous work (yu Wang, Jiang & nian Wang 2010;Jiang et al 2011;Peng et al 2018)) to be able to resolve the fast spatial and temporal oscillations of the ambipolar electric field. Semi-implicit, adaptive methods for kinetic plasma simulations have been developed to be able to self-consistently include higher resolution areas in low resolution domains (Innocenti et al 2013(Innocenti et al , 2015 and will be used in the future for similar simulations. To make sure that the configuration we reach is stable, we run our simulations up to 25 ms, compatibly with experimental data.…”

Section: Simulation Resultsmentioning

confidence: 99%

On the breakdown modes and parameter space of ohmic tokamak start-up

et al. 2018

View full text Add to dashboard Cite

Tokamak start-up is strongly dependent on the state of the initial plasma formed during plasma breakdown. We have investigated through numerical simulations the effects that the pre-filling pressure and induced electric field have on pure ohmic heating during the breakdown process. Three breakdown modes during the discharge are found, as a function of different initial parameters: no breakdown mode, successful breakdown mode and runaway mode. No breakdown mode often occurs with low electric field or high pre-filling pressure, while runaway electrons are usually easy to generate at high electric field or low pre-filling pressure (${<}1.33\times 10^{-4}$ Pa). The plasma behaviours and the physical mechanisms under the three breakdown modes are discussed. We have identified the electric field and pressure values at which the different modes occur. In particular, when the electric field is $0.3~\text{V}~\text{m}^{-1}$ (the value at which ITER operates), the pressure range for possible breakdown becomes narrow, which is consistent with Lloyd’s theoretical prediction. In addition, for $0.3~\text{V}~\text{m}^{-1}$, the optimal pre-filling pressure range obtained from our simulations is $1.33\times 10^{-3}\sim 2.66\times 10^{-3}$ Pa, in good agreement with ITER’s design. Besides, we also find that the Townsend discharge model does not appropriately describe the plasma behaviour during tokamak breakdown due to the presence of a toroidal field. Furthermore, we suggest three possible operation mechanisms for general start-up scenarios which could better control the breakdown phase.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

On the breakdown modes and parameter space of ohmic tokamak start-up

et al. 2018

View full text Add to dashboard Cite

show abstract

“…The jumps in spatial and temporal resolution between the levels are indicated as Refinement Factor, RF = ∆x/δx, and Time Ratio, T R = ∆t/δt, where ∆ and δ label the resolution on the coarse and refined grid. Refinement Factors and Time Ratios as high as RF = 14 and T R = 10 have been used in Innocenti et al 35 . The refined grid in a two level MLMD system has a spatial extension L x /RF × L y /RF , with L x and L y the dimensions of the coarse grid.…”

Section: The Multi Level Multi Domain Methods and Its Application mentioning

confidence: 99%

“…The aim is to reduce the computational cost of the simulations while retaining fundamental characteristics: a fully kinetic description of both ions and electrons, realistic simulation parameters and a wide range of wavelengths. The method proposed, the Multi Level Multi Domain method [33][34][35] , is demonstrated through the simulation, at realistic mass ratio, of turbulence generated by the Lower Hybrid Drift Instability (LHDI). In the case of the LHDI, the use of high mass ratios is essential to ensure a clear separation between the electron and the ion scales 36 .…”

Section: Introductionmentioning

confidence: 99%

Study of electric and magnetic field fluctuations from lower hybrid drift instability waves in the terrestrial magnetotail with the fully kinetic, semi-implicit, adaptive multi level multi domain method

et al. 2016

Self Cite

View full text Add to dashboard Cite

The newly developed fully kinetic, semi-implicit, adaptive multi-level multi-domain (MLMD) method is used to simulate, at realistic mass ratio, the development of the lower hybrid drift instability (LHDI) in the terrestrial magnetotail over a large wavenumber range and at a low computational cost. The power spectra of the perpendicular electric field and of the fluctuations of the parallel magnetic field are studied at wavenumbers and times that allow to appreciate the onset of the electrostatic and electromagnetic LHDI branches and of the kink instability. The coupling between electric and magnetic field fluctuations observed by Norgren et al. [“Lower hybrid drift waves: Space observations,” Phys. Rev. Lett. 109, 055001 (2012)] for high wavenumber LHDI waves in the terrestrial magnetotail is verified. In the MLMD simulations presented, a domain (“coarse grid”) is simulated with low resolution. A small fraction of the entire domain is then simulated with higher resolution also (“refined grid”) to capture smaller scale, higher frequency processes. Initially, the MLMD method is validated for LHDI simulations. MLMD simulations with different levels of grid refinement are validated against the standard semi-implicit particle in cell simulations of domains corresponding to both the coarse and the refined grid. Precious information regarding the applicability of the MLMD method to turbulence simulations is derived. The power spectra of MLMD simulations done with different levels of refinements are then compared. They consistently show a break in the magnetic field spectra at k⊥di∼30, with di the ion skin depth and k⊥ the perpendicular wavenumber. The break is observed at early simulated times, Ωcit<6, with Ωci the ion cyclotron frequency. It is due to the initial decoupling of electric and magnetic field fluctuations at intermediate and low wavenumbers, before the development of the electromagnetic LHDI branch. Evidence of coupling between electric and magnetic field fluctuations in the wavenumber range where the fast and slow LHDI branches develop is then provided for a cluster magnetotail crossing.

show abstract

“…Thanks to the semi-implicit IMM temporal discretization, the temporal and spatial resolution of the code can be adapted to the scale of the processes under investigation rather than being tied to the strict stability constraints of explicit discretizations. This property is massively exploited in the Multi-Level Multi-Domain evolution of iPic3D, where several grid levels are resolved with different spatial and temporal resolutions to catch progressively smaller scale processes (Innocenti et al 2013(Innocenti et al , 2015(Innocenti et al , 2016. Other, non semi-implicit, fully kinetic implementations of similar EB formulations are Sironi & Narayan (2015) and Ahmadi et al (2017), where the compression (rather than expansion) of the plasma mimics collisionless accretion flow and magnetopause compression respectively.…”

Section: Introductionmentioning

confidence: 99%

Onset and Evolution of the Oblique, Resonant Electron Firehose Instability in the Expanding Solar Wind Plasma

Innocenti

Tenerani

Boella

et al. 2019

ApJ

View full text Add to dashboard Cite

A double-adiabatically expanding solar wind would quickly develop large parallel to perpendicular temperature anisotropies in electrons and ions, that are not observed. One reason is that firehose instabilities would be triggered, leading to an ongoing driving/saturation evolution mechanism. We verify this assumption here, for the first time, for the electron distribution function and the electron firehose instability (EFI), using fully kinetic simulations with the expanding box model. This allows the self-consistent study of onset and evolution of the oblique, resonant EFI in an expanding solar wind. We characterize how the competition between EFI and adiabatic expansion plays out in highand low-beta cases, in high and low speed solar wind streams. We observe that, even when competing against expansion, the EFI results in perpendicular heating and parallel cooling. These two concurrent processes effectively limit the expansion-induced increase in temperature anisotropy and parallel electron beta. We show that the EFI goes through cycles of stabilization and destabilization: when higher-wavenumber EFI modes saturate, lower-wavenumber modes are destabilized by the effects of the expansion. We show how resonant wave-particle interaction modifies the electron eVDF after the onset of the EFI. The simulations are performed with the fully kinetic, semi-implicit Expanding Box code EB-iPic3D.

show abstract

Introduction of temporal sub-stepping in the Multi-Level Multi-Domain semi-implicit Particle-In-Cell code Parsek2D-MLMD

Cited by 19 publications

References 59 publications

On the breakdown modes and parameter space of ohmic tokamak start-up

On the breakdown modes and parameter space of ohmic tokamak start-up

Study of electric and magnetic field fluctuations from lower hybrid drift instability waves in the terrestrial magnetotail with the fully kinetic, semi-implicit, adaptive multi level multi domain method

Onset and Evolution of the Oblique, Resonant Electron Firehose Instability in the Expanding Solar Wind Plasma

Contact Info

Product

Resources

About