NASA Technical Reports Server (NTRS) 19930017793: Electron cyclotron thruster new modeling results preparation for initial experiments

9 Pages·0.23 MB·English

by  NASA Technical Reports Server (NTRS)

N93-26982
Electron Cyclotron Thruster
New Modeling Results
Preparation for Initial Experiments
E. Bickford Hooper
Lawrence Livermore National Laboratory
Presented at
Nuclear Propulsion Technical Interchange Meeting
NASA-LeRC Plum Brook Station
October 20-23, 1992
Whistler-Based ECRH Thruster -- Concept
A thruster using ECRH has no electrodes and, Is thus less sensitive to
materials problems than arc-based thrusters such as the Magneto-Plasma
Dynamic (MPD) arc.
• Rear wall bombardment can
be minimized, by a large
inlrror ratio between the
magnetic field
resonance and peak field.
(The flow across the mirror
Is reduced by approximately
the mirror ratio from that
downfleld.) This:
gns is_joction
._LC.10Lr0_[Qn_S.oq._.r_..c......e..... ._ ,_. .... Maximizes efficiency by
minimizing energy loss to
Wave _. the wall
propagation
o Maximizes lifetime by
minimizing material
damage
LR
Cross-field Coupling In the Helicon Approximation
Coupling Is expected to be strongest If the magnetic field has s small
gradient• Thus, we consider coupling at the peak of the magnetic mirror,
There, coc/oJ,copIoJ) 1. We Illustrate the coupling at o_+I+ = 10, (copJoj)2 = 1000.
This IS the helicon regime, with
=z o_(os-_ co_O) co_¢mO
• The wave characlerlstlcs can be seen from a plot of the squared parallel vs
perpendicular Indices of refraction
Waves In the upper-rlgh| quadrent are
propagating both along z and radially.
These are the waves o! Interest
There are two such waves at a given
..... +...._ .......... I........................
-t(,oo :+ +6o iooo parallel Index of refraction, but one Is
at very large perpendicular Index of
'il ,,) =k),:2/_ refraction and not of Interest In the
-10( finite-radius plasma column
-15C
The finite-radial geometry will plckout
particular values of nj.
EBH 1/30.3i/92
Wave propagation:
Wavequide with helix and plasma column
• Several modes with dlfferenl radial structure propagate In the system
I i _I_/0. C=tllw 11= i0lle612 em* n. J,6(l ©m m - I
2.0 1+1- -33,00 ° r_- 1,00 bl = 1'_0'."20cm ' ' ll "
• -plaml l_Im, l
- Ml,l ,_+u
b .w_gu_ i_
v-11+11._ i
!,
++• , ,
• • .• t• • .
• ..
k
• • .o \ •fi', "
"_Itl
•i,/'. •
°1,
irii -- -- -- ._-
•'"+ k '+_
05 •"" ! i
o::'.
W ._- ,,,(If"qNellO(l
O+U _L ._._¢.._,_1 ..... _. I_ 1_+..+__, +_+i _ ,L+.+.i_ J_L---..J,-. -4....4...
-tO Io
NI_: Technology 1054 NP-TIM-92
Wave structure: Low Impedance mode
• Electric field = solid lines, magnetic field ,= dashed lines
• Note Jump In magnetic field corresponding to current flow In helix
radial azimuthal axial '
;'" " _ " + I " _..... ' "" ¢" " " i i++_lk._
' i J.(r_.) I *, ._z_,) ', m,(L.) .° otirol •
.- ill m
........ . i i + • +m
+! _. =uB*
_ . /r I _.1=l
12': ' i
_. LtlR
i m_{C_r) I _{C-i) ; _((J) i,,. ILIIIM _*
,\ p !.... !
L, nip • _
ol ,
o ,
lhK'- -LI;_
• ' 'I ........ +" " "
o+oV,L+ ,,+* k
;_ i' I +t , +; •*o ,
I- .,'- -_Ilk.+l
1,0 .. :". i • ° o
, •, . ': +
0S + . _+ " .t
t . ,° , + : . .0 , , _ • ,0
B
Wave structure: Hiqh impedance mode
• Electric field = solid lines, magnetic field, = dashed lines
• Note no Jump In magnetic field corresponding small current flow
azimuthal axial
radial
••iI_+II _+
:'x:
t l._w
l.. _lrt.**. =.
-LI<+_.*_ 01,..
-I + I '. ,' I
I ". I
_2 + l I • . °
.u+__ --,.'"'%,+Ip"+--- ,_ I II++4'*=+le,S*_
. •
o.+ ,
_0 ', [
ik.it. ii*I.+i.I
00 -- . +P_-
' +'<,m' } 8 ,O , •4(+.," } , 10 , 4,(ca)' , ,O
NP-TIM-92 1055 NEP:Technology
System ,Impedance varies with plasma density ELI
• The experiment is designed to allow tuning of the microwave system
10(_
E
J¢
o
10 _
o
u
c
o
'°1
E
10 .... ' .... , • ,' • • , ....
O.Oe÷O 5.0e+17 1.0e÷18 1.5e÷18 2.0e+18
Density (m-3)
Wave Absorption at the Cyclotron Resonance
• As the whistler wave approaches the cyolotror_ resonance, the value of kll
becomes very large and the phase velocity becomes small
This has two favorable consequences for absorption:
o The direction of propagation becomes nearly along the field and at short
wavelength so that reflection Is very small
o The phase velocity becomes comparable to the thermal velocity of the .
particles, so that the Doppler-shifted resonance (_o- =,- k,,vH = O)
couples to the bulk electrons
• Furthermore, there Is no electromagnetic plasma mode at high density and
> mc, sO the wave cannot tunnel through the resonance
• Absorption is consequently nearly 100% for the whistler wave at the
cyclotron resonance
• Absorption at high power will generally generate a nonthermel electron
velocity distribution, Calculations are needed to quantify this and Its
consequences
EBH In0-31 m2
1056 m,-'r_.92
Flow sensitivity to electron distribution functi0q
The Isothermal and adiabatic limits Illustrate the sensitivity of the
flow to the thermal conductivity and thus to the electron
distribution function
For ECRH the electron distribution may be ansiotropic and
nonthermal in nature, with significant consequences for thermal
conductivity, particle and energy flow, plasma recycling at the rear
wall, etc.
Understanding the distribution resulting from the heating, as a
function of plasma density and microwave power, Is thus key to
predicting performance.
Comoarino isothermal and adiabatic Dlattma flow
Magnetic field (loop model) Density
/ ", !
/ \ 8/8.. n/no !
0,5
,' \ O.8
/ \ o.,
0,2
_..X,/_dt/ba_lc
'O.!
/ t
I 0.4
"Q%.2., O.02 Isolhcrmal_
O+Ol
-3 -2 -1 ! 2 z/_ 3 -3 -2 -t # ! 2 z/a 3
Flow veloclly Electron temperature (adiabatic)
II/C_ I
2
zl
t 2 z/= 3
-3 -2 -t l 2 z/,' 3
NP-TIM-q2 1057 NEP: T_xJ_ok_u
ECR thruster modelinq: heatin_q and plasma flow
• A particle-in-cell code - ICEPIC - has been used to model the
thruster plasma heating and motion along the magnetic field
• Individual particles are followed In the guiding center
approximation
o Electrons are heated by rf with velocity-space diffusion In
the quaalllnear approximation
o For the present cases, the electrons are weakly colllsional
o The ion mass Is 10Ome to speed up calculations
• Plasma Is injected on the side of a magnetic hill and heated up the
hill from the injection point
• Two cases are compared
Inlected Te In!ected TI ECRH
NoECRH 100 eV 5 eV None
ECRH 5 eV 5 eV Er! = 320 V/cm
Geometry for PIC code model
Maanetic field strenaths
z(cm) 0 2 3.5 10
B(gauas) 3650 2350 1250 125
B(O)/B 1 1,5 2.9 29
sticky ECRH plasma sticky
wall inje,ction wail
$
I
I
g
I
I
!
$ |
% ! , .! ,.
flxJslPosl#J0n [cm)
NEP:Tvchnolofy 1058 i_-'I1M-92
Ele r n " r e°°mom n in thefl w
• The electrons are highly anlsotroplc even without ECRH
• The electron temperature is highly nonuniform along B
• Strong electron heating by ECRH is evident perpendicular to B
No ECRH ECRH
Parallel
Te
"t2[ I I_
> .is
"2BI •
Perp
i'_/
Te
.aS
1
itxiel Posntlon (cml
Den i nd otential are stron I affe t ECRH i_
• Note the rise In potential upfleld of the ECRH. It reduces the flow ol
Ions to balance the p.aB/as force on the electrons and maintain
quaslneutrallty
ECRH
No ECRH
/I\
I..
°/
Density
lgle
c: 101g'
o I
o
Potential - I
_. -50 "
e
-10C
0 _ --, _.
flxnel pos_l_on (cm] Rxlel Posm'tlon[cml
NP-TIM-92 1059 NEP: Tochnology
Ele ron ener is conv ed in ion flow ELaJ
No ECRH
ECRH
i.¢f "l
Ion
._ .S.
s.catter
plot
-,6
_ t_!6!f_.".,
f9 ¢_1 (D ¢O ,I
=, _'_,'_
.E
Ion
velocity
>^ _, / =
" I0
-o!
IJ
tlxJol Posl?,on (c=l
Rxael Pos,'t_0n [ca)
fl w h fi I issu r e E RH [_=J
• The total energy flow Is proportional to the flux bundle area, which
Is a factor of 29 larger at the exit than at the magnetic field peak
NO ECRH
ECRH
• /
I!
Electron 'r' e.
energy
flux
_ .._.._,,
=- _,' _.- _ _.
C3 , , n , , , . . _
2.I C_
•. ,/ : q.
Ion
energy 22
flux >_I_]6
li,
_, :,,' _, _ , _
Rx,al Posil,on loll rlxool Pos=?,on tc_!
NEP:Technology 1060
NP-TIM-92
Initial experimental tests: preparation
Initial experiments will be conducted at NASA LeRC (tank 7)
o Space has been provided; magnets and SCR controller for pulsing
microwave power have been sent to LeRC
o Microwave components have been delivered to LeRC
o Vacuum vessel, helical coupler, and gas box have been
constructed and are undergoing final bench tests at LLNL
• First experiments will be directed to forming the plasma and
making preliminary measurements of density, electron temperature
• Subsequent experiments will explore the details of the plasma for
comparison with modeling
o Electron anlsotropy
o Suppression of flow to rear wall
o Efficiency
• Measurements will also be made of the separation of the plasma
plume from the magnetic nozzle