Ei dian otsikkoa - Helsingin yliopisto
Download
Report
Transcript Ei dian otsikkoa - Helsingin yliopisto
CMS
Plasma-Wall Interactions –
Part II: In Linear Colliders
HIP
Helga Timkó
Department of Physics
University of Helsinki
Finland
Plasma-Wall Interactions – Outline
Part I: In Fusion Reactors
Materials
Science Aspect
- Materials for Plasma Facing Components
- Beryllium Simulations
Arcing
in Fusion Reactors
Part II: In Linear Colliders
Arcing
in CLIC Accelerating Components
Particle-in-Cell
Future
Helga Timkó, University of Helsinki
Simulations
Plans for a Multi-scale Model
Laudatur Seminar, 16th Sept. 2008
2
Last Week:
Arcing in Fusion Reactors
Arcing = continuous gas discharge, between electrodes or
within the plasma sheath
Causes in fusion reactors
Erosion,
Impurities
And
thus, plasma instabilities harder to reach confinement
Research on arcing has been done since 1970’s
Search
for arc-resistant materials, ideal surface conditions
Theoretical
and experimental modelling of arcing in simplified
geometries
All
in all, in fusion reactors arcing not so critical any more
But for future linear colliders it is!
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
3
CLIC = Compact Linear Collider
‘only’ 47.9 km
A proposed e- – e+ linear collider, with a CM energy of
up to 3 TeV in the final design (cf. LEP max. 209 GeV)
Linear
Can
colliders more effective than circular ones
reach higher energies
With CLIC, post-LHC physics can be done, e.g. for
Higgs physics this means:
LHC
CLIC
should see Higgs(es), should rule out some theories
would be able to measure particle properties
To be built in
three steps
Two-beam
acceleration
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
4
CLIC accelerating components
Under testing in the CTF3 project at CERN
Too high breakdown rates, 10-4, aim: 10-7 for final design
Different setups have been tested:
Geometries
Materials:
Cu and Mo best
Frequencies:
main linac fRF
was lowered 30 → 12 GHz
Most challenging is the high
accelerating gradient to be
achieved, already lowered too
150 → 100 MV/m
Need: a theoretical model
of breakdown to systemise
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
5
What is PIC and
what can we simulate with it?
PIC = Particle-in-Cell method
Basic idea: simulate the time evolution of macro quantities
instead of particle position and velocity (cf. MD method)
Need
superparticles
Restricted
to certain regime of particle density given by
reference values (those define dimensionless quantities)
Kinetic
approach of plasma, but can be applied both for
collisionless and collisional plasmas
Many application fields: solid state and quantum physics
as well as in fluid mechnics
Has become very popular in plasma physical applications
Esp.
Helga Timkó, University of Helsinki
for modelling fusion reactor plasmas (sheath and edge)
Laudatur Seminar, 16th Sept. 2008
6
The PIC Algorithm
Setting up the
simulation:
Grid
size, timestep,
superparticles, scaling
Solving the equations of motion » particle mover «
Moving particles, taking collisions & BC’s into account
Calculating plasma parameters, macro quatities
Solving Maxwell’s equations, (Poisson’s eq. in our case)
this can be done with different » solvers «
Obtaining fields and forces at grid points
In PIC, everything is calculated on the grid, interpolation
to particle positions is done by the » weighting scheme «
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
7
Solvers for
the Particle Mover and the Poisson’s Equation
Discretised equations of motion:
In 1D el.stat. case, with the leapfrog method, in
the Boris scheme:
Poisson’s equation determining the electric field
from charge density values at grid points:
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
8
Scaling in PIC –
Grid size and timestep
In the code, everything is scaled to dimensionless
quantities → easier to analyse physically, faster code
Initial
values give the scale for the simulations, only a few
orders of magnitudes can be captured
- Need a good guess: n0 = 1018 cm-3, Te = 5 keV
- Determines λD = 5.3×10-7 m and ωpe = 5.6×1013 1/s, the
internal units of the code
- For an arc, densities are only rising! model is limited
Stability conditions:
Compromise
btw. efficiency and low noise:
Δx = 0.5 λD, Δt = 0.2× 1/ωpe
Amazing: whole set of equations can be rescaled
universal results; only the incl. of collisions gives a scale
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
9
Our Model
In collaboration with the Max-Planck-Institut f.
Plasmaphysik, Greifswald
1D electrostatic, collision dominated PIC scheme
Simplistic surface interaction model:
Assuming
Const.
Cu+
const. electron thermoemission current (cathode)
flux of evaporated neutral Cu atoms, Icu=0.01Ith,e
ions sputter Cu with 100% probab., neutral Cu is
reflected back when hitting the walls
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
10
Including collisions
Arcing highly collision dominated, so is our model
Including only 3 species: electrons, neutral Cu, Cu+ ions
Multiply
ionised species ignored
Most important collisions are taken into account:
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
11
A Typical Output
Macro quantities as a function of time
Flux and energy distributions, currents
Note the sheath!
Animations by K. Matyash:
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
12
The Plasma Sheath
Sheath = a thin layer of a few Debyes near the wall
All physics happens in the sheath:
Field
& density gradients, collisions
Outside,
the potential is constant, field is zero: Doesn’t really
matter what the dimensions of the system are (nm or μm)
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
13
Future plans: Integrated Modelling of Arcing
Multi-scale model aimed: an integrated
PIC & MD model of arcing
Collaboration
between:
- Max-Planck-Institut für Plasmaphysik
MPI Greifswald
K. Matyash
R. Schneider
HIP, Helsinki
H. Timko
F. Djurabekova
K. Nordlund
- Helsinki Institute of Physics
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
14
Thank You!
Bibliography:
D. Tskhakaya, K. Matyash, R. Schneider and F. Taccogna: The ParticleIn-Cell Method, Contributions to Plasma Physics 47 (2007) 563.
Computational Many-Particle Physics, Springer Verlag, Series: Lecture
Notes in Physics, Vol. 739 (2008)
Editors: H. Fehske, R. Schneider and A. Weiße
Information:
http://clic-study.web.cern.ch/clic-study/
http://beam.acclab.helsinki.fi/~knordlun/arcmd/
Helga Timkó, University of Helsinki
Laudatur Seminar, 16th Sept. 2008
15