1
A Brief History of Results from
the Rotating Wall Machine
David Hannum
W. Bergerson, M. Brookhart, C.B. Forest, R.D. Kendrick, C.P. Soldan
UW-Madison
2
Motivation/Outline:
Can Moving Metal Walls Stabilize the RWM?
RWM: chief performance limit in toroidal devices
Study ext. kink mode in unique cylindrical device
Change wall around plasma to change growth rate
Identify several current-driven, line-tied modes
RWM, FWM, NWM
Characterize plasma from unique multi-gun source
Creates plasma, drives current, controls safety factor
q
a
determines mode onset
MHD changes when plasma current rises
Finally, add a rotating wall to stabilize RWM
(Gimblett/Hegna)
3
Two Walls Can Stabilize
RWM by Rotation
RWM lies between No Wall, superconducting Ideal Wall limits
Spinning outer wall expands stable safety factor range
FWM stability reduced by magnetic permeability of wall
r = a: Plasma
r = b: Stat. wall
r = c: Rot. wall
4
Rotating Wall Machine
Line-Tied Screw Pinch
5
Look Inside to See Plasma Source and Walls
6
Access Different Modes by
Changing Static Wall
7
Flux Loop Array Outside Wall
Sees Spatially-Locked RWM
8
RWM: Predictions Roughly Match Data
FWM: Faster Growth Than Predicted
9
Characterize Plasma Inside Machine, then
Add Rotating Wall
Plasma gun array
Axial probe: B-dot or Langmuir (singletip)
Segmented Bullseye anode
10
Electrostatic washer guns
7 guns, t
pulse
~ 25 msec
Each gun can be individually fired, biased
For this talk, all seven guns fire in tandem
Plasma Gun Array Controls n
e
, J
z
Profiles
11
Plasma Gun Array Pictures
12
Axial Probes Take B-dot,
Langmuir Measurements
Probe moves from z = 10-120 cm (z = distance from guns)
Rotate to trace arc, derive radial profile (r
eff
= pseudo-radius)
Segmented
anode in back
View from guns
down the column
13
Segmented Anode Gives Rough J
z
Profile,
Confirms B-dot Probes Measurement
Each segment connected to ground, with Rogowskis
to monitor current flow
Can approximate q profile:
2 2
0
4
z
r B
q
LI
t
=
14
As current rises
Density rises:
T
e
stays uniformly cold (3 eV) in bulk plasma
E field rises, affecting ExB drift and resistivity
Raise the Plasma Current:
Internal Probes Show Merging Streams
V I P n
e
2
=
15
Current Streams Visible at z = 10 cm,
Merge Further Down the Column
Contour map data: 5 axial x 21 radial
16
Density Rises With Plasma Current
Large radial expansion as current increases
Big densities get even bigger with plasma current
17
Injected Power (in form of Plasma Current)
Increases Electron Density by 500%
T
e
~ 3 eV throughout bulk plasma as I
p
rises
Spectrometer sees high neutral density (~10
18
m
-3
) from
97% ionization for least-ionized case
o
H
18
2D Electrostatic Field (E
r
, E
z
) Defined by
Voltage Drop Along Column
Wall Floats
Grounded
Anode
Negative
Bias
Larger E
r
close to guns
19
Rigid rotation in core, with max v
ExB
/ v
A
=.16
Slows down as plasma approaches anode
E
r
xB
z
Rotation is Radially, Axially Sheared
20
= E
z
/J
z
: Direct measurement from Langmuir/B-dot
Parallel Resistivity Twice the Spitzer Value
//
q
21
Summary
Modular wall system shows current-driven kink
mode for several boundary conditions
NWM, RWM, FWM all detected
Turning RWM into FWM reduces q
a
stability range
Gun array drives kink mode, other MHD effects
Current streams reconnect into tight laminar column
Large density streams expand radially
ExB drift with radial, axial shear
Parallel resistivity: close to Spitzer
Final step: Stabilize RWM with rotating wall
Assembled, tested, installing it in July
Questions? [email protected]
22
Preliminary m=1 amplitude data
Still subtracting error fields, calculating growth rate
Completely unexpected
RWM Stabilized, and Destabilized,
by Rotation
23
Large Rotation (Hence R
m
) Drives RWM Into
Stability
t
-1
pulse
24
Temperature Flat, Filled with Errors
25
Clear plasma radius a,
vacuum region a<r<b
Heaviside J
z
profile
IMHD:
Alfven timescale
Lundquist # S is infinite
Observed Physics Call for
Revisions to Growth Rate Model
Model
0 = |
Variable a, approaches
wall in some cases
Gaussian J
z
profile
(Diamagnetic loop)
RMHD:
Fast reconnection
S ~ 1-10
m
Reality
0 = q O ~ q 100
B B B J
~ ~ ~
2
+ =
o
o
o o
e R + V V H p
Add resistive & pressure gradient terms
2 . ~ |
Revise momentum balance
26
Spectrometer looks across plasma column
Plasma almost fully ionized in each case
Neutral density fraction:
H-Neutrals Dominate Impurities
) /(
e n n
n n n + = _
