DSMC Simulation of the Plasma Bombardment on Io

1
DSMC Simulation of the Plasma Bombardment on Ios
Sublimated and Sputtered Atmosphere

• Chris Moore0 and Andrew Walker1
• N. Parsons2, D. B. Goldstein1, P. L. Varghese1,
  L. M. Trafton1, D.A. Levin2
• 0Sandia National Labs1University of Texas at
  Austin2Penn State University
• 50th AIAA Aerospace Sciences Meeting
• 1/10/2012

Supported by the NASA Planetary Atmospheres and
Outer Planets Research Programs. Computations
performed at the Texas Advanced Computing Center.
Sandia is a multiprogram laboratory operated by
Sandia Corporation, a Lockheed Martin
Company,for the United States Department of
Energys National Nuclear Security
Administration under contract DE-AC04-94AL85000.
2
Outline

• Brief motivation and background information on Io
• Overview of physical models in our planetary DSMC
  Code
• Description of new physical models
• Particle description of the plasma
• Surface sputtering due to energetic ions
• Ion reaction chemistry
• Photo-chemistry
• Atmospheric Simulations
• Conclusions

2
3
Motivation

• Jovian plasma torus sweeps past Ios atmosphere
  causing
• Heating
• Chemistry
• Changes to the global winds
• Enhanced gas columns due to sputtering
• Observed auroral glows
• Matching obs. can be used to probe the torus
  conditions

Jupiter
Io
Plasma Torus
Io Flux Tube
Illustration by Dr. John Spencer

• Io supplies the Jovian plasma torus
• Surface and atmospheric sputtering
• Ionization
• Charge exchange

3
4
Background Information on Io
Frost patch of condensed SO2
Volcanic plume with ring deposition
Illustration by Dr. John Spencer

• Io is the closest satellite to Jupiter
• Radius 1820 km (slightly larger than our moon)
• Atmosphere sustained by volcanism and sublimation
  from SO2 surface frosts
• Dominant dayside atmospheric species is SO2
  lesser species - S, S2, SO, O, O2
• Io is the most volcanically active body in the
  solar system
• Volcanism is due to an orbital resonance with
  Europa and Ganymede which causes strong tidal
  forces in Io

4
5
Brief Overview of DSMC

• DSMC simulates gas dynamics using a large
  number of representative particles
• Position, velocity, internal state, etc. stored
• Particle collisions and movement are decoupled in
  a given timestep
• Particles are moved by integrating Fma
• Binary collisions allowed to occur between
  particles in the same collision cell

5
6
Overview of our DSMC code

• Atmospheric models
• Rotational and vibrational energy states
• Sub-stepped emission
• Variable gravity
• Simulate plasma with particles
• Chemistry neutral, photo, ion, electron
• Surface models
• Non-uniform SO2 surface frosts
• Comprehensive surface thermal model
• Volcanic hot spots.
• Residence time on the non-frost surface
• Surface sputtering by energetic ions
• Numerical models
• Spatially and temporally varying weighting
  functions.
• Adaptive vertical grid that resolves mfp
• Sample onto to uniform output grid
• Separate plasma and neutral timesteps

Time scales Chemistry 10-12 seconds Surface
sputtering 10-10 seconds Plasma Timestep 0.005
seconds Ion-Neutral Collisions 0.01 seconds -
hours Vibrational Half-life millisecond-second Cyc
lotron Gyration 0.5 seconds Neutral Time step 0.5
seconds Neutral Collisions 0.1 seconds -
hours Residence Time seconds - hours Ballistic
Time 2-3 minutes Flow Evolution Several
hours Eclipse 2 hours SO2 Photo Half-life 36
hours Io Day 42 hours
6
7
Overview of our DSMC code

• Atmospheric models
• Rotational and vibrational energy states
• Sub-stepped emission
• Variable gravity
• Simulate plasma with particles
• Chemistry neutral, photo, ion, electron
• Surface models
• Non-uniform SO2 surface frosts
• Comprehensive surface thermal model
• Volcanic hot spots.
• Residence time on the non-frost surface
• Surface sputtering by energetic ions
• Numerical models
• Spatially and temporally varying weighting
  functions.
• Adaptive vertical grid that resolves mfp
• Sample onto to uniform output grid
• Separate plasma and neutral timesteps

Length scales Atomic interactions 10-9
m Sputtering radius 10-7 m Debye Length lt1
m Electron Larmor radius 3 m Dayside neutral
m.f.p. 10 m Volcanic plume vents 0.110
km Ion-neutral m.f.p. 500 m Electron-ion
m.f.p. 1 km Ion Larmor radius 3
km Atmospheric scale height 10100 km Nightside
neutral m.f.p. 100 km Volcanic plumes 100500
km Ios radius 1820 km Jovian plasma
torus 105 km
7
8
3D / Parallel

• 3D
• Spherical grid northern hemisphere
• 33 latitude/longitude cells
• Non-uniform radial grid
• Parallel
• MPI, 900 CPUs

• Parameters
• 360 million molecules instantaneously
• Simulated 10 hours to quasi-SS
• 25,000 computational hours

8
9
Surface Sputtering

• Ion energies in the collision cascade regime
• Little sputtering contribution from electronic
  excitation
• Sputtering yield proportional to incident ion
  energy

9
10
Surface Sputtering

• Ion energies in the collision cascade regime
• Little sputtering contribution from electronic
  excitation
• Sputtering yield proportional to incident ion
  energy
• Sputtering yield exponential with surface frost
  temperature

SO2 sputtering yield, S, versus SO2 frost
temperature. Lanzerotti et al. (1982)
10
11
Surface Sputtering

• Ion energies in the collision cascade regime
• Little sputtering contribution from electronic
  excitation
• Sputtering yield proportional to incident ion
  energy
• Sputtering yield exponential with surface frost
  temperature
• Sputtered particles leave with Thompson energy
  distribution

Sputtered SO2 energy distribution. Boring et al.
(1984)
11
12
Charged Particle Motion

• Acceleration during move
• Use predictor-corrector integrator
• Pre-computed (MHD) fields used
• Electrons are assumed to move with the ions
• Debye length ltlt m.f.p.

Simulate simple ion motion and impact onto
surface
B-Field
E-Field (Out of the page)
12
13
Heavy Interactions MD/QCT1

• SO2 O collisions simulated using Molecular
  Dynamics/Quasi-Classical Trajectories (MD/QCT)
• RK-4 integration of Hamiltonian equations
• Particles interact via their potentials
• Cases run for range of collider velocities and
  initial SO2 internal energies

• Each case consists of 10,000 separate
  trajectories Microcanonical sample unique
  impact parameters and initial SO2 component
  coordinates
• Potential Energy Surface
• Total potential of SO2 O system is the
  summation of the collisional interaction
  potential and molecular potential of the SO2
  molecule
• Collisional interaction Lennard-Jones 6-12
  potential
• SO2 molecular potential Murrell 3-body
  potential
• Allows for accurate dissociation of SO2 molecule
  to SO O, O2 S, or S 2O

13
1Parsons, N. and Levin, D., 50th AIAA Aerospace
Sciences Meeting 2012-0227
14
Heavy Interactions

• MD/QCT (fast neutrals/ions) or theoretical cross
  section data vs. translational and internal
  energy
• Linearly interpolate between nearest cross
  section data points
• If no MD/QCT data, use Arrhenius coefficients
  TCE

• Always use the total cross section to determine
  the reaction rate (number of selections and
  fraction accepted)
• VHS cross section Total cross section above 20
  km/s

14
15
Photo-chemistry

• Rate constants, kreact,s,i, assume quite sun
• Assume gas is optically thin
• Optical depth over photo-dissociation wavelengths
  less than 0.1
• Give dissociation products an average excess
  kinetic energy
• Accurate below the exobase where products are
  collisionally equilibrated

1 Io Day
0-D box initialized with only SO2 particles.
Lines are analytic, diamonds from DSMC.
15
16
Simulation Conditions

• Io just before ingress ? Plasma incident onto
  dusk terminator
• Assume uniform SO2 frost ? No rock surface or
  residence time
• Assume simple radiative equilibrium surface
  temperature model
• Do not account for Ios rotation, thermal inertia

16
17
3D Results SO2

• SO2 number density peaks near the subsolar point
• Day-to-night near surface flow develops from
  subsolar point
• Retrograde wind forms and high density finger
  extends past the dawn terminator due to plasma
  pressure
• Slight increase in the polar atmosphere due to
  preferential polar sputtering

Direction of Ios rotation
17
18
3D Results O2

• O2 produced via photo-dissociation on dayside
• Non-condensable O2 gas dynamics very different,
  but day-to-night flow still present
• O2 finger extends much further onto the
  nightside, to the dusk terminator
• Retrograde flow across nightside meets
  day-to-night flow at dusk terminator
• O2 diffuses towards the poles where it is
  stripped away or destroyed by the plasma

Dawn Terminator
18
19
3D Results O

• O density contours 4 km above Ios surface
• High altitude ions stream along field lines to
  surface
• On the nightside, ions stream to the surface
• Upstream torus O
• density 2400 cm-3
• Dense dayside atmosphere prevents plasma
  penetration
• Enhancement on the dayside from plasma flow

19
20
Surface Sputtering of SO2 Frost

• Sputtering primarily on the nightside and at high
  latitudes
• Dense atmospheric columns (gt 1015 cm-2) block
  energetic ions from reaching the surface
• Obs. show green auroral glow only on Ios
  nightside
• Sodium is believed to be sputtered off Ios
  surface
• Simulated SO2 sputtering map suggests Na is the
  source of green aurora with sputtering blocked on
  dayside

20
21
Discussion
Current simulation

• Direction of plasma flow relative to subsolar
  point important
• Subsolar point changes during Ios orbit ?
  Atmospheric dynamics will change as Io orbits
  Jupiter
• Sputtering only occurring near night time
  temperatures implies preferential scouring of
  surface by plasma from 270360
• Eclipse inhibits formation of dayside atmosphere
• Plasma directly impacts this quadrent
• Ios surface frost poor in this region

21
22
Conclusions

• The interaction of the Jovian plasma torus with
  Ios atmosphere was simulated using the DSMC
  method.
• A sub-stepping method was used to time-resolve
  the movement and collisions of energetic ions and
  electrons from the Jovian plasma torus
• MD/QCT simulations were used to compute the
  cross-sections for heavy reactions
• Sputtering from Ios surface by energetic ions
  and fast neutrals was included
• Formation of high density finger onto the
  nightside near the dawn terminator due to plasma
  pressure
• Interesting O2 flow feature generated at the dusk
  terminator
• Non-condensable O2 pushed across the nightside to
  the dusk terminator where it meets the opposite
  day-to-night flow
• O2 stagnates and forced to diffuse slowly towards
  the pole until it is stripped away and/or
  dissociated
• Sensitivity of sputtering on surface temperature
  can lead to sharp gradients in sputtering column
  density Sputtering blocked by large columns gt
  1015 cm-2
• Concentrated at high latitudes and on low density
  nightside
• Possible cause of observed (Voyager, Galileo)
  frost-poor region of Ios surface

22
23
Electron Interactions

• Used measured (and theoretical) cross sections
  versus relative velocity (energy)
• For SO2, SO, O2, S2, O, and S
• Because of the large number of reactions,
  precompute reaction probabilities vs. finite set
  of energies
• Account for anisotropic scattering

SO2 cross sections
Trace molecular cross sections
O cross sections
24
Surface Sputtering

• Physically, sputtered particles shouldnt collide
  with each other
• Sputtering occurs over area of 10-13 m2 ? point
  source
• Collisions between sputtered particles tends to
  reduce the expansion velocity parallel to the
  surface
• Solution Place sputtered particles at cell
  corner, weighted by the inverse distance from
  impact point to each corner

0
3
Lc,0
Lc,3
Lc,2
Lc,1
2
1
Surface Cell
2
1
12
25
Column Densities

• Due to flow, SO2 column density no longer
  hydrostatic near dawn terminator on nightside
• Nightside SO2 column density increases slightly
  towards pole
• SO2 condensable, SO partially condensable, O2
  non-condensable
• O2 dominates nightside
• SO extends further than SO2 onto nightside
• O2 shows large buildup in column density near
  dusk terminator

19