Member of the Helmholtz Association Particle Confinement Control with Resonant Magnetic Perturbations (RMP) at TEXTOR-DED Oliver Schmitz 1 , J.W. Coenen 1 , H. Frerichs 1 , M. Lehnen 1 , B. Unterberg 1 S. Brezin T.E. Evans 3 , K.H. Finken 1 , M.W. Jakubowski 1,2 , M. Kantor 1 , A. Kraemer-Flecken 1 , V. U. Samm 1 , G.W. Spakman 1 , G. Telesca 1 and the TEXTOR Team 1 – Forschungszentrum Jülich GmbH, Institut für Energieforschung- IEF- 4:Plasmaphysik, Jülich, Germany 2 - Max Planck Institut für Plasmaphysik, IPP-EURATOM Association, Greifswald, Germany 3 - General Atomics, P.O. Box. 85608, San Diego, California 92186-5608 USA
33
Embed
Particle Confinement Control with Resonant Magnetic Perturbations (RMP) at TEXTOR-DED
Particle Confinement Control with Resonant Magnetic Perturbations (RMP) at TEXTOR-DED. Oliver Schmitz 1 , J.W. Coenen 1 , H. Frerichs 1 , M. Lehnen 1 , B. Unterberg 1 S. Brezinsek 1 , M. Clever 1 , - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Mem
ber
of
the H
elm
holt
z
Associa
tion
Particle Confinement Control with Resonant Magnetic Perturbations (RMP) at TEXTOR-DEDOliver Schmitz1, J.W. Coenen1, H. Frerichs1, M. Lehnen1, B. Unterberg1 S. Brezinsek1, M. Clever1,
T.E. Evans3, K.H. Finken1, M.W. Jakubowski1,2, M. Kantor1, A. Kraemer-Flecken1, V. Philipps1, D. Reiter1,
U. Samm1, G.W. Spakman1, G. Telesca1 and the TEXTOR Team
1 – Forschungszentrum Jülich GmbH, Institut für Energieforschung- IEF-4:Plasmaphysik, Jülich, Germany2 - Max Planck Institut für Plasmaphysik, IPP-EURATOM Association, Greifswald, Germany3 - General Atomics, P.O. Box. 85608, San Diego, California 92186-5608 USA
RMP facilitate particle transport and exhaust control in modern fusion devices
ELM control in divertor tokamaks with H-mode plasma
Please see • invited talk by R. Moyer, Wednesday, I-11• Posters P1-32, Monday by M.E. Fenstermacher, • P2-01, Tuesday by E.A. Unterberg, • P3-30, Thursday by S. Mordijck
DIII-D -> ELM suppression JET -> ELM mitigation
Please see invited talk by Y. Liang, Wednesday, I-12
Helical and island divertor in helical devices and Stellarators
Please remeber talk by M. Kobayashi O-3 and see e.g. S. Masuzaki P2-33, M. Shoji P2-02 and many more …
Magnetic islands and stochastic layers realize particle exhaust and facilitate control of particle inventory
Role and origin of particle pump out is an important topic to understand
DED at TEXTOR as flexible tool to mockup various perturbed magnetic topologies
Stochastic boundary induces controlled density reduction …
Continuous density decrease with increasing DED current
• 25% decrease in density
• temperature constant!
… with flattening of edge density gradient!
Flattening of ne(r) gradient in edge region N>0.85
In contrast, stochastic boundary also allows for spontaneous density build up …
Spontaneous density build up at moderate perturbation level
• 15% increase in density
• temperature constant!
Reported: Finken K.H. et al., PRL 98 (2007) 065001
Also observed at Tore Supra: Ghendrih Ph. et al., NF 42 (2002) 1221-1250 and Evans T.E., Word Scientific 2008
… with steepening of edge density and temperature gradient!
Steepening of ne(r) and Te(r) gradient in edge region N>0.92
Particle balance allows to quantify confinement changes in measures of P and P*
Change of number of confined particles
Influx from recycling
Influx from beams and gas inlet
How does manipulation of P relate to transport?
io is hardly available experimentally, needs 3D modeling with e.g. EMC3/EIRENE
a needs to be determined from topology
see e.g. Stangeby P., „The plasma boundary of magnetic fusion devices“, IoP 2000
n shows weak manipulation with perturbed topology established
Complex, 3D magnetic topology induced
Are density changes correlated to perturbed topology?
• valid in plasma edge region …
• … without tearing modes
Jakubowski et al., PRL 96 (2006) 035004
Spakman et al., NF (2008), submitted
For TEXTOR:
Vacuum paradigm used
external RMP field
axis symmetric plasma equilibrium
+
DED target
Reduction to 1D description
LK > Lc -> Laminar Zone, i.e. SOL pendant
LK < Lc -> Ergodic Zone with stochastic field line diffusion
Kolmogorov LK length is used to order complex 3D topologyGhendrih Ph. et al., PoP 38 (1996) 1653
Tokar M. et al., PoP 6,7 (1999) 2808
Strong simplification neglects details of 3D
structures and transport
Probing of inner resonant island chain with open field lines improves confinement stepwise
Level of ergodisation on q=5/2 surface determines increase in P
Increase in P with ergodic layer approaching q=5/2 surface
Probing of inner resonant island chain with open field lines improves confinement stepwise
Increase in P with ergodic layer approaching q=5/2 surface
Level of ergodisation on q=5/2 surface determines increase in P
Probing of inner resonant island chain with open field lines improves confinement stepwise
Increase in P with ergodic layer approaching q=5/2 surface
Level of ergodisation on q=5/2 surface determines increase in P
Probing of inner resonant island chain with open field lines improves confinement stepwise
Decrease in P with laminar layer jumping in and ergodic layer extending the q=5/2 surface
Level of ergodisation on q=5/2 surface determines increase in P
E x B shear increases and turbulent transport decreases on q=5/2 surface for IPC
Increase of E x B shear (m/n=6/2) and reduction on DRW (m/n=3/1) observed at q=5/2 surface
A. Kraemer-Flecken et al., NF 46 (2006) S730-S742
q=
5/2
TEXTOR Reflectometer
m/n=6/2m/n=3/1
P and P* decrease with raising DED current showing reduced particle confinement
Simultaneous reduction of CVI concentration in corePoster 3.81 by G. Telesca et al.
Reduction of P ~ 20% and of CVI concentration ~ 25% during stochastic pump out
P and P* decrease with raising DED current showing reduced particle confinementLevel of ergodisation of resonant surfaces determines decrease in P
Decrease of P with ergodic layer extending q=6/2 surface
P and P* decrease with raising DED current showing reduced particle confinementLevel of ergodisation of resonant surfaces determines decrease in P
Decrease of P with ergodic layer extending q=5/2 surface
P and P* decrease with raising DED current showing reduced particle confinement
Further reduction in P with laminar layer penetrating, i.e. extension of SOL
Level of ergodisation of resonant surfaces determines decrease in P
E x B shear is reduced on q=5/2 surface for particle pump out
Effective radial outward transport is enhanced and overcomes improvement of particle confinement
Decrease of E x B shear at q=5/2 surface in case of particle pump out
Radial electron loss flattens shear
Extending laminar zone displaces SOL shear layer inside
Unterberg B. et al., JNM 363-365 (2007)
0
200
400
600
800
1000
1200
L-modeDED, 1 kADED, 2.5 kADED, 4 kA
pedestal width(no DED)
0
2000
200
LK
LK
0.8 0.85 0.9 0.95 1 1.050
200
N
LK
IDED=1.0 kA
IDED=2.5 kA
IDED=4.0 kA
H-mode
pe [
Pa]
Lc [
m]
Application of PO to limiter H-mode shows correlated reduction of density pedestal
laminarergodic
Increasing stochastic layer width allows for pedestal control
Stronger reduction of pe in pedestal as soon as ergodic layer exceeds
pedestal width
Destruction of pedestal as soon as laminar layer exceeds pedestal
Dedicated control of density pedestal in TEXTOR limiter H-modes achieved
Particle pump out and connected reduction of P is driving term
Poster 1.03 by B. Unterberg et al.
Summary and conclusion
Perturbed magnetic topology determines confinement stage reached
Improved particle confinement
• Shot cuts to wall change radial electric field and improve particle confinement
Resolution of localized particle source distribution and fuelling mechanism is important to conclude on changes in radial particle diffusion coefficient
Particle pump out
• Stochastic field line diffusion becomes dominant and enhanced outward transport is indicated
• Confinement loss due to open field lines is overcompensated
• Radial electron loss reduces electrical field gradients
EMC3/EIRENE will help to resolve source distribution vs. magnetic topology
At TEXTOR both regimes can be achieved on demand and therefore studied in detail
Thank you!
Particle balance allows to quantify confinement changes in measures of P and P*
Change of number of confined particles
Particle eflux
Influx from recycling
Influx from beams and
gas inlet
Pumped particle balance
Particle confinement time Effective particle confinement time
How does change of P relate to transport?
io is hardly available experimentally, needs 3D modeling with e.g. EMC3/EIRENEa needs to be determined from topology
see e.g. Stangeby P., „The plasma boundary of magnetic fusion devices“, IoP 2000
Particle balance allows to quantify confinement changes in measures of P and P*
Tangential CCD camera with D filter
Calibrated against gas inlet
Particle balance allows to quantify confinement changes in measures of P and P*
Magnetic topology in m/n=6/2 base mode in geometrical coordinates
Imprint of homoclinic tangles as direct proof for stochastization
Proves penetration of RMP field in accordance to vacuum magnetic topology and shows non-linear deviation in case of plasma feedback!
m/n=12/4c
m/n=12/4c
M. Jakubowski et al., JNM (2007)
Direct validation of vacuum approach
Imprint of homoclinic tangles as direct proof for stochastization
However, transition to TM unstable regime leads to deviation!
m/n=12/4c
m/n=12/4c
m/n=6/2
m/n=3/1
M. Jakubowski et al., JNM (2007)
Direct validation of vacuum approach
Electron temperature and density fields
Important role of open, perturbed field lines resolved!
Laminar field lines imprint characteristic poloidal modulation!
O. Schmitz et al., NF 48 (2008) 024009
Direct validation of vacuum approach
Electron temperature and density fields
Important role of open, perturbed field lines resolved!
Laminar field lines imprint characteristic poloidal modulation!
O. Schmitz et al., NF 48 (2008) 024009
Direct validation of vacuum approach
Electron temperature and density fields
Impact much more pronounced in electron density!
Ergodic domain showed enhancement of radial particle transport by 30%
reduction by 40%
reduction by 20%
O. Schmitz et al., NF 48 (2008) 024009
Direct validation of vacuum approach
Identification of reconnected magnetic islands and implication to transport
Occurrence of edge island causes sudden drop in P by 50%
Magnetic islands in source region are able to drive particle transport