Slide 1

Dynamical plasma response during driven magnetic reconnection in the laboratory Ambrogio Fasoli* Jan Egedal MIT Physics Dpt & Plasma Science and Fusion Center Ackn.: W.Fox, J.Nazemi, M.Porkolab *Now at EPFL Physics Dpt & Centre de Recherche en Physique des Plasmas Slide 2 Change of magnetic field topology in the presence of plasma Reconnection rate: value of E-field along X-line, perpendicular to plane over which flux annihilates Our definition of magnetic reconnection Slide 3 Driven reconnection in the VTF open cusp Conditions to create a current channel Dynamical evolution of j and E j || and dE/dt linked through ion polarization current Size of diffusion region (E B 0) Orbit effects Future work on VTF New diagnostics Closed cusp configuration Outline Slide 4 Family of Reconnection Experiments (from H.Ji, PPPL) Slide 5 2 m The VTF device Magnetic Reconnection on VTF Origin of fast time scale for reconnection, mechanisms behind breaking of frozen-in flux Particle orbits ? Instabilities / waves ? . Slide 6 Diagnostic workhorses on VTF 40 Channels B-probe 45 heads L-probe Slide 7 VTF configuration mfp >>L, coll > orbit,t A ; i 1 Plasma production by ECRH separate from reconnection drive Ex. of target plasma profiles B cusp = 50mT, B guide = 87mT; P ECRH ~ 30 kW J.Egedal, A.Fasoli et al., RSI 71, 3351 (2000) Slide 8 Reconnection drive Ohmic coils driven by LC resonant circuit Flux swing ~ 0.2 V-s, duration ~ 6 ms (>>t reconnection ) V loop ~ 100 V, v ExB ~ 2km/s ~ v A /10 Slide 9 Sketch of poloidal flux during reconnection drive No reconnection as in ideal MHD Fast reconnection as in vacuum Slide 10 Plasma response to driven reconnection (1) Slide 11 Plasma response to driven reconnection (3) Current layers develop for l 0 =B guide / B cusp ~3m No steady-state Questions: How much max current / min anomalous resistivity (though ratio E/J is not constant!)? What determines the size and time evolution of the diffusion region where ideal MHD breaks? Slide 12 Anomalous resistivity First observation of strongly anomalous parallel resistivity ( meas = E /J max ) - Current sustained in plasma determines reconnection rate - For l 0 =B guide / B cuspSlide 13 Electrostatic potential (away from the X-line) B=b 0 ( x x y y + l 0 z) E = E z ( x + y + z) -l 0 l 0 2x 2y -- EzEz =E z l 0 log(|x/y|) Ideal region : E + vB = 0 E B = 0 E= E z z - Slide 14 Observation of self-consistent e.s. potential E z 0 Poloidal Drift w/o e.s. potential: charge separation No charge separation if E tot B =E z l 0 log(x/y) E z l 0 log(x/y) ExperimentTheory Deviations from E B=0 are observed close to separatrix diffusion region J.Egedal and A.Fasoli, PRL 86, 5047 (2001) Slide 15 The size of the diffusion region (1) Frozen in law is broken where E B 0 Experimental measurement = 3.5 cm E B=(E - ) B Fit extending form valid in ideal region Slide 16 The size of the diffusion region (2) NeonNitrogen KryptonXenon The size of the diffusion region is clearly independent of ion mass and n e It cannot be related to c/ pi,e or i,s Slide 17 The size of the diffusion region (3) The size depends on [cm] Slide 18 Temporal evolution of the current channel Time response of the toroidal current Time in steps of 12 s f ~20-30 kHz H2H2 Ar Slide 19 Plasma response to an oscillating drive (1) Slide 20 0 1.2 kA/(Vm 2 ) In phase withV loop 0 20 mAs/(Vm 2 ) 90 0 ahead of V loop Plasma response to an oscillating drive (2) The current profile can be separated in two parts: What causes the out of phase current? Slide 21 fits exp. Data with = 3.5 cm Ion polarization currents due to d /dt Ion polarization current: Quasi neutrality: Slide 22 Interpretation with polarisation current predicts time evolution and shape of current channel MEASURED change in and j dt between t=0 and 30 s THEORY / FIT Slide 23 Circuit model for VTF plasmas (1), ; Slide 24 Circuit model for VTF plasmas (2) Total current is measured in each shot by a Rogowsky coil Values of R j1 and C j2 obtained by curve fitting Deviation? As the observed dependence implies C j2 V loop [As] Slide 25 What breaks the frozen in condition? The plasma frozen in condition is violated where: Generalized Ohms law: too small ( Slide 26 Breaking the frozen in law Orbit width, cusp = ( g l 0 ) 0.5 All electrons are trapped limiting macroscopic current channel Electrons short circuit electric fields along their orbits The frozen in law is broken in areas where the orbits do not follow the field lines: EB 0 J.Egedal, PoP 9, 1095 (2002) [cm] Slide 27 Conclusions Fast, collisionless driven reconnection directly observed B guide > B cusp Dynamic evolution of current profile and self-consistent potential Classical collisions: not important Ion polarisation current (analogy with RLC circuit) explains observed reconnection dynamics Key parameter is Diffusion region does not scale with el./ion Larmor radius, el./ion skin depth, but with characteristic particle orbit size Direct measurements of different terms in generalised Ohms law suggest that p e (off-diagonal) and/or dJ/dt terms are needed (kinetic effect) Slide 28 Future developments Energy and velocity distributions Laser Induced Fluorescence f i (v) at one position; planar LIF f i (v, x) with intensified CCD E.s. energy analyzer Systematic analysis of e.s. and e.m. fluctuations Spatial and temporal correlations; effect on plasma effective resistivity Combined reconstruction of (x,t) and f e,i (v,x) around X-point particle energization mechanism Machine upgrades Increase strength of reconnection drive (reduce ECRH frequency) Installation of in-vessel coils Reduction in direct plasma losses: from collisionless to collisional regime Slide 29 VTF Diagnostics: LIF E.g. planar set-up: f i (v klaser,x,y) Pulsed dye laser (Lambda Physik Scanmate pumped by Nd:Yag) pumps 611.5 nm line E laser ~ 20 mJ in 10 ns LIF detected at 461 nm (intensified CCD?) 1 3 2 Slide 30 IF PMT Present set-up for LIF on VTF laser beam Slide 31 First observation of LIF on VTF ArII plasma Broad band, E laser ~ 5 mJ/pulse, t laser ~ 15 ns Slide 32 Freq (MHz) 0 250 Freq (MHz) 0 250 E.S. fluctuations during reconnection, weak I p J(r) vs.time Increased turbulence close to current channel, where gradients are large f < 100 MHz