Modelling of plasma tank and related langmuir probe calibration MATEO-VELEZ J.-C, ROUSSEL J.-F., SARRAIL D, BOULAY F., INGUIMBERT V.

Transcription

1 Modelling of plasma tank and related langmuir probe calibration MATEO-VELEZ J.-C, OUSSEL J.-F., SAAIL D, BOULAY F., INGUIMBET V. PAYAN D. ONEA CNES

2 Objectives Initial: Validation of SPIS modelling (LEO type conditions) Have a simulation almost ready to simulate our routine experiments (just by changing mock-up, polarisations...) Incidental: Improve LP I-V characteristics interpretation method 2

3 The Study Ionosphere simulation tank JONAS Experiments achieved in 2002 SPIS development numerical simulations CNES &T funding Plasma source (Argon + + electrons) 1850 Pressure ~ mbar Plasma density ~ cm -3 Ion energy ev Electron temperature ~ 1500 K Test on materials, Devices, etc

4 The Experiment Plasma I-V characteristics using Langmuir probes (4) plasma density electron temperature plasma potential Wake effect using a plate Fast ions (10-25 ev) emitted from the source can not reach the region behind the plate, Slow ions created by charge exchange reactions (CEX) are present in the whole tank LEO 4

5 Langmuir probe I-V characteristics interpretation Classical method, method 1 (M1) Plasma potential I Electron branch N e Floating potential φ s - φ p Classical handling of ion branch: linear interpolation to plasma potential Extra difficulty: separate fast and slow ions => need several probes (simultaneous interpretation) Spherical probe I + Planar probe (// or to the source) Plasma source (Fast ions) Slow ions S 2 S 1 Plasma source (Fast ions) Slow ions S 1 S 2 5

6 Guarded planar probes results without the plate (method M1) Fast ion density profiles at various distances from the source: 1/d 2 decay (m-3) 1,2E+13 1,0E+13 8,0E+12 6,0E+12 4,0E+12 2,0E+12 X=230mm X=800mm X=1230mm log(n+r) 13, , ,5 pente -2,03 0,0E Y (cm) Slow ion density profiles (noisy) 11-0,8-0,6-0,4-0,2 0 0,2 log(x) (m-3) 5,0E+12 4,0E+12 3,0E+12 2,0E+12 X=230mm X=800mm X=1230mm 1,0E+12 0,0E Y (cm) 6

7 Guarded planar probes results WITH the plate (method M1) No fast ions in the plate wake 3,0E+11 2,5E+11 2,0E+11 X=38mm de la plaque X=200mm de la plaque X=510mm de la plaque (m-3) 1,5E+11 1,0E+11 5,0E+10 0,0E Y (cm) Slow ions in the plate wake (m-3) 4,0E+12 3,0E+12 2,0E+12 X=38mm de la plaque X=200mm de la plaque X=510mm de la plaque 1,0E+12 0,0E Y (cm) 7

8 Discussion of LP data interpretation Ion branch linear interpolation to plasma potential: Only an approximation Quite valid : fast ions: small influence of potential, and not far from linear in OML conditions: theory => linear for a sphere But not really: in non-oml conditions (larger r / λ d ratio) fast + slow ions: more complicated Improvements: General idea: use a realistic ion branch model(ling) Other ~ analytical theory when available: Langmuir-Blodgett instead of Langmuir-Mott-Smith (OML): cf. below, in the wake Numerical method, interesting in more complicated situations (SPIS), as e.g. fast + slow ions 8

9 Spherical probe in a non-drifting plasma (method M2) Application = wake: slow ions only, far from linear Only slow ion current collection S In a Maxwellian non-drifting plasma, sheath size S (Langmuir et Blodgett Parker 1980) : S S = = D D where D is 1D Child-Langmuir sheath size Then current = D D H ( 0,2) ev D = 1.26 D kt I = en 4πS for for D 3 D 3 19 > ε 0 λ λ D = 2 2 kbt 2πm + k B T Ne total current (A) 2,0E-07 1,5E-07 1,0E-07 5,0E-08 0,0E+00-5,0E-08-1,0E-07-1,5E-07-2,0E-07 measurements Te = Ti model Probe potential (V) 9

10 Other approach: probe modelling with SPIS (not used here) spherical probe 15 mm in diameter Maxwellian plasma at rest. N = m -3, Te = Ti =1eV OML theory valid total current (A) 2,0E-07 1,0E-07 0,0E+00-1,0E-07 adial motion model for hot ions SPIS simulation OML model 10 spherical probe 15 mm in diameter Maxwellian plasma at rest N = m -3, Te = Ti =0.09eV Langmuir-Blodget theory valid total current (A) -2,0E Probe potential (V) 2,0E-07 adial motion model for hot ions 1,0E-07 SPIS simulation OML model 0,0E+00-1,0E-07-2,0E-07-3,0E Probe potential (V)

11 JONAS tank numerical modelling (SPIS) CAD model and mesh using Gmsh Pump orifice Tank nodes tetra Source plate Collecting sphere 11

12 Numerical simulations (SPIS Model) Plasma volume - Electrons : Boltzmann distribution - CEX reaction Supposes a uniform neutral density (SPIS improvement) obtained by exp. Enables the prediction of slow ion production Pump orifice (Plasma boundary) Ion source φ= 3 cm I = 1 ma E = 15 ev PIC method Satellite 12

13 Numerical simulations (SPIS) Parameter Simulation time Fast ion current esidual gas pressure CEX cross section σ Tank voltage Electron temperature Electron density N 0 Value s 1,1 ma Pa m 2-2 V 0,09 ev m -3 Numerical results Fast ion density : m -3 ealistic decrease of ion density (~ 1/r 2 for fast ions) Wake effect clearly demonstrated for fast ions Important amount of slow ions > m -3, which is in agreement with measurements 13

14 Numerical simulations (SPIS) Parameter Simulation time Fast ion current esidual gas pressure CEX cross section σ Tank voltage Electron temperature Electron density N 0 Value s 1,1 ma Pa m 2-2 V 0,09 ev m -3 Numerical results Electron density and potential : Boltzmann distribution N e N 0 exp((v-v 0 )/k B T) Quasi neutral plasma 14

15 Comparison of simulations to experimental data Fast ion density Drifting ion density (m-3) 1,4E+13 1,2E+13 1,0E+13 8,0E+12 6,0E+12 4,0E+12 2,0E+12 Experiments Method M1 Numerical results Drifting ion density (m-3) 1,0E+14 1,0E+13 1,0E+12-2,03 slope Exp. method M1 Numerical results 1,0E+11-0,8-0,6-0,4-0,2 0 0,2 Log. of distance to the source (m) 0,0E Distance from the source axis (cm) 15

16 Comparison of simulations to experimental data Slow ion density 1,0E+13 Exp. OML model (M1) Exp. radial motion models (M2) Numerical results CEX ion density (m-3) 1,0E+12 1,0E+11 plate location 1,0E ,5 1 1,5 2 2,5 3 distance to the source (m) Much better agreement with M2 method 16

17 Conclusions and perspectives Validation of the numerical approach for plasma tank modelling The final model will help to simulate the experiments to be conducted in the ONEA tank Future possible improvements of the approach Using SPIS for generic LP I-V characteristics interpretation Improved tank model: a DSMC for CEX would allow predicting the neutral pressure, and possible inhomogeneities (here taken from measurements) 17