Magnetic reconnection, merging flux ropes, 3D effects in RSX

Transcription

1 Magnetic reconnection, merging flux ropes, 3D effects in RSX T. Intrator P-24 I. Furno, E. Hemsing, S. Hsu, + many students G.Lapenta, P.Ricci T-15 Plasma Theory Second Workshop on Thin Current Sheets University of Maryland, College Park April 19-21, 2004 UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 1

2 Magnetic reconnection, merging flux ropes and 3D effects in RSX T. Intrator, I. Furno, E. Hemsing, S. Hsu, G. Lapenta, P. Ricci Los Alamos National Laboratory In space physics, solar coronal plasmas, and astrophysics, there are many instances of current-carrying magnetic flux ropes, which are a fundamental building block of these plasmas. Observations of our solar photosphere show magnetic field that is organized into isolated flux bundles, that continue up to the corona. The magnetic fields are wound up by photospheric motions, twisting the flux tubes into flux ropes. Locally this geometry resembles the RSX experimental configuration. A collection of flux tubes or ropes can wind, braid, and intertangle over some distance. The physical processes resulting from their mutual interactions may form the basis for the observed complex behavior of coronal plasmas; including relaxation to large scale self organized structures. It is likely that the essential physics resides in the magnetic topology, requires magnetic field annihilation, and must have the freedom to evolve in three dimensions (3D). The Reconnection Scaling Experiment (RSX) in the P-24 Group is our experimental tool. It is clear from RSX data that two or three flux ropes can twist, braid and merge, annihilating flux to form a composite flux rope. In the context of a generalized Ohm s Law, preliminary data show that truly 3D effects dominate. RSX has the unique capability to experimentally vary the guide field, collisionality, and number of flux ropes independently of each other. UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 2

3 Outline Plasma physics research P-24, T-15, X-1 Magnetic reconnection and flux rope interactions Reconnection Scaling Experiment (RSX) Pictures of 3D flux ropes Magnetic data Density, temperature, pressure data 3D effects Theory and computations CELESTE-3D, FLIP-3D - LANL Discussion and summary UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 3

4 In nature, reconnection is expected to be free-boundary and 3D Solar arcades 3D reconnection region Magnetic field generated by plasma currents with complex geometry are present in nature 3D effects important Plasma flow is generated by internal instabilities free-boundary rcxn Attraction of current channels UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 4 Need for new experiments

5 Reconnection Scaling experiment - RSX RSX is a linear machine Dynamic interactions of 2, 3 MHD flux tubes rcxn is driven by internal instabilities in a fully 3D geometry, ie no toroidal symmetry Independent scaling of key parameters Collisionality (density) and reconnection magnetic field (current) B guide /B rcxn Ion gyro radius / ion skin depth Field aligned current λ=µ 0 J/B m -1 q Research plans for RSX Free boundary - onset of instability and reconnection micro physics of reconnection Why high guide field? Tokamak limit, kink stability Quasi 2-D Good data set UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 5

6 RSX experiment schematic UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 6 Cross section of RSX device: (a) side view, (b) end view and (c) schematics of two current channels interacting along the axis of the device. RSX main elements are shown: (1) plasma guns, (2) magnet coils, (3) vacuum vessel High ionization fraction, low neutral collisionality %

7 3D cutaway view RSX Key advantages Scan guide field B guide /B rcxn Scan density n x10 14 cm -3 3D geometry Free boundary Study scaling vs β UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 7

8 Flux ropes twist individually and collectively Several guns fire axially t 0 t nsec UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 8

9 Three or more flux ropes are possible Normal operation in hydrogen Magnetic probe stalks Operation in argon for spectroscopy diagnostics of ion flows UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 9

10 Two flux tubes with X-point UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 10

11 Reconnection geometry These data: B y /B z <<1 β<<1 ρ s <<c/ω pi UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 11

12 Movie of curling, kinking, of 2 flux ropes UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 12

13 Internal and external measurements z y x Probe insertion for in situ measurements Internal measurements over many shots in the reconnection plane (x,y) => magnetic field, electron temperature and density. External fast CCD camera to externally monitor the dynamics. Fast camera Visible light images of coalescence UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 13

14 Experimental set-up and diagnostics Density Guide field Reconnection magnetic field electron temp scale size ion skin depth electron skin depth ion gyro radius electron gyro radius n e ~10 13 cm -3 B Z = Gauss B rec ~10-20 Gauss T e ~3-12eV L~6-10cm δ i =c/ω pi ~2 cm δ e =c/ω pe ~1-2 mm r Gi ~ cm r Ge ~1-4mm Reconnection in high guide field (B z / B rec = 20-60) 2 plasma guns z = 0, 4cm spacing column length=1m Many reproducible shots in x,y reconnection plane z Multi-2D magnetic probe (2.5mm spacing B field Triple probe ( 2 mm space resolution T e, n e UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 14

15 Typical time history of shot Long quiescent time Magnetic probe data on this expanded time scale Gun arc turns on first Bias between external anode & gun 1 msec later Current rise time 30 µsec Reconnection data for first 40 µsec of current ramp Shots are very reproducible for first 28 µsec Columns start to thrash around subsequently UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 15

16 B field jump, induced current layer UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 16

17 current sheet width shrinks to small size Need to vary B guide and R Gi / δ e ratio to find which scale length is relevant Interesting result: Current sheet width 5mm<<ion skin depth UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 17

18 Magnetic data movie B 0z =200G Flux line contours B vector arrows Current density color contours Note current reversal at X point UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 18 Bz reverse

19 Reconnection of magnetic field lines is observed during coalescence B z / B rec = 25 B field Current channels X- Point UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 19

20 3D effects density contours UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 20

21 3D effects pressure contours UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 21

22 magnetic contours A z line contours B vector arrows E z color contours UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 22

23 cut between density maxima, current channels Middle of cut X point for reconnection E z reverses during rcxn Look at time history at points along this cut UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 23

24 Radial density propagation Anode 0V gun -250V startup v G 10 6 cm/sec rcxn E z direction UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 24

25 Time history near X point wave propagates in electron J e direction Wave propagation reversal n e >n i n i >n e z=5cm <=> closer to gun z=0cm <=> farther from gun UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 25 initial gun pulse: electrons move away from gun towards external anode X point forms, reconnection and axial E z field reversal occurs In rcxn region, electrons propagate towards gun

26 Some plasma and wave parameters Plasma T e 2-4eV B z =200G B rcxn 5G n e cm -3 Time scales f 150kHz f ci 300kHz ω ci 2e6rad/sec ω pi -1 10nsec length c/ω pe 2mm k z 0.25cm -1 ρ S 0.5cm λ z 25cm dimensionless k z ρ S 0.1 c S /v A 0.15 ω/ω ci 0.5 speeds v wave x10 6 cm/sec =20-30km/sec upstream towards gun v A 1.3e7cm/sec (guide Bz) v A 3x10 5 cm/sec (rcxn Bx) c S 2e6cm/sec v inflow 10 6 cm/sec Bulk flow 3x10 6 cm/sec (away from gun) 3D ( P z /en)xb x /B x 2 v y (inflow) 2x10 6 cm/sec v y xb x Ez 10-12V/m UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 26

27 KAW wave could fit the bill Assume complex k_perp, ω/ω ci, calculate kz Dispersion KAW couples Alfven, ion acoustic waves [I 0 (λ i )exp(- λ i ) -(ω/k z c S ) 2 ] {1- (ω/k z v A ) 2 [1-I 0 (λ i )exp(- λ i )]/ λ i } = (ω/k z v i ) 2 [1-I 0 (λ i )exp(- λ i )] UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 27

28 Could be ion density hole Potential reflected electrons dip passing electrons Electrons in reconnection reversed current region collide with potential hill from ion hole. Slow ones reflect, fast ones pass UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 28

29 Current sheet thickness does not depend on guide field Guide magnetic field scan 3 magnetic fields: B z =100, 200,400 Gauss Peak current density in the current sheet Gun pressure adjusted to have same n e No strong dependence of the current sheet thickness on B z is observed J 0 2ΔB x /Δ sheet in the current sheet increases and saturates with B z This observation may be interpreted in terms of increased electron mobility (meandering electron orbits) in the z direction due to a reduced Larmor radius [Ricci P. et al, Physics of Plasmas 10, 3554 (2003)]. UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 29

30 Current sheet shrinks between δ i and δ e size Ion skin depth c/ω pi 9cm Electron skin depth c/ω pe 2mm Ion Gyroradius ρ i 0.7cm Ion sound radius ρ s 1.7cm UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 30

31 3D effects: reconnection dynamics depends on the direction of the guide field B_z // J_z B_z // J_z Electron heating T e increase is explained by ohmic heating Dynamics depends on whether background magnetic field is parallel or anti parallel to J rcxn UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 31

32 reconnection and electron scattering 3D effects Consider Ohm s Law η J = E + vxb + P /(en e ) - JxB/(en e ) + η =(η Spitzer +η unknown ) Identify the physics then η => η Spitzer, and the rest can be quantified on the right hand side At the X point E z (from 2 A=µ 0 J) 2 P/(en e ) Appears to be driven by n/ z, a 3D effect Wave propagation with electrons UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 32

33 Required diagnostic resolution Ion skin depth (cm) Electron skin depth (mm) Alfven time (1µsec) Reconnection time (3-10µsec) Wave activity bandwidth > 20MHz Data of interest Reconnection rate vs B guide Is there an electron layer? Diagnostics Sufficient diagnostics are on line now B-field and J profiles; across rcxn layer, z axis B-coil arrays [1 mm (~ c/ω pe ) and µs (~ τ A ) resolution] State of the art: 250 µm probe development with Sandia/NHMFL Rogowski probe (1 cm, 0.1 µs) Electron density/temperature: triple Langmuir probe [2mm, 0.2µsec] Global imaging 2-frame CCD camera, 200 ns gate time Photo multiplier tube array, 16 chords, 70 ns under development 3D probe drive Ion flow/temperature Line-averaged Doppler spectroscopy Insertable spectroscopy probe [1cm, 1µsec] Mach probe [2mm, 0.1µsec] UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 33

34 Example of computational 3D results y z = axial out of plane direction CELESTE3D predicts kink formation. This view is in the y(vertical) -z(axial) plane. This requires fully kinetic modeling. (unpublished). 3D Hall MHD code results from Huba. Field lines warp out of reconnection plane. Field line tension in reconnection plane & z direction (new prediction). Huba 3D Hall rcxn PoP 9, 4435 (2002) UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 34

35 3D code CELESTE-3D UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 35 X-y(movie) = x-z (RSX) ions, electrons, Hall, LHDI, density waves propagating upward (lower left, Ni) LHDI waves (lower right, Ey) Kinking, late time Ni plot at flanks of rcxn channel flow structure with negative ion flowing in the reconnection region (electrons bending field lines => 4-polar structure, out of plane B)

36 What is next? Experimental study of more complex geometry which are relevant to solar dynamics 3 plasma guns for flux rope bundle Heat up T e to vary Lundquist number Measurement of reconnection rate and comparison with different reconnection theory support from theory side (T-15, X-1 collaboration) Development of a ultra miniature magnetic and electric field 3D probes (SANDIA collaboration) Access to unprecedented small scale length. UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 36

37 Three flux tubes attract each other, repel induced current UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 37

38 Conclusions RSX Unique experiment magnetic evolution of flux tubes two current channels, twin island topology 3D free boundary reconnection current sheet much thinner than the ion skin depth Ion gyro radius can be adjusted independently from electron skin depth Experimental collaboration with numerical simulations at LANL UnivMaryland 19-21Apr2004 Intrator 3D flux tubes 38