Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
-24,
201
0rq
ue, N
M, S
epte
mbe
r 20-
ogy
Con
fere
nce,
Alb
uque
rec
raft
Cha
rgin
g Te
chno
lo11
th S
pace
1
Comparison of numerical and experimentalComparison of numerical and experimental investigations on the ESD onset in the Inverted
Potential Gradient situation in GEO
P. Sarrailh, J.-C. Matéo-Vélez, J.-F. Roussel, B. Dirassen (ONERA)
11th Spacecraft Charging Technology Conference, Albuquerque, NM, September 20-24, 2010
, , , ( )J. Forest, B. Thiébault (ARTENUM)
D. Rodgers, A. Hilgers (ESA)
Outline
• General overview on the ESD risk• Experimental setup
-24,
201
0
p p• SPIS ESD tool• Parametric study: comparison experimental/numerical results
rque
, NM
, Sep
tem
ber 2
0-og
y C
onfe
renc
e, A
lbuq
uer
ecra
ft C
harg
ing
Tech
nolo
11th
Spa
ce
3
General overview on the ESD risk
Spacecraft Charging at GEO in sunlight
• S/C structure at a negative potential several kV due to electron collection• Photoemission on solar cell cover glass:
• Dielectrics in sunlight more positive than S/C structure
-24,
201
0
g p• Barrier of potential on the front side due to rear side negative potential
• This situation is the Inverted Potential Gradient (IPG)
net photo electron current
rque
, NM
, Sep
tem
ber 2
0-
saddle point - 2010 V
- 1500 V
hυrecollected electrons
ogy
Con
fere
nce,
Alb
uque
r
S/C SA SA
- 2800 V
- 2000 Vfront side
rear side - 3000 V
electrons
ecra
ft C
harg
ing
Tech
nolo
- 2500 V
- 2200 V ibl t j t i f
2010 V
11th
Spa
ce
5
- 2010 V
- 1500 V
IPG situation at the solar cell scale
• At the solar cell edge:• Gap surface covered with (photoconductive) kapton = S/C structure voltage• Solar cell voltage ~ S/C structure voltage (strings voltage ~100 V)
-24,
201
0
• Cover Glass top surface is more positive than solar cell = differential voltage• Triple point between conductor, dielectric and vacuum
• Uses of a simplified geometry (valid for interconnects) to compare experimental and
rque
, NM
, Sep
tem
ber 2
0- numerical results
-2200 V-2400 V
-2600 V
-2800 VFront surface
barrier potential
he-
ogy
Con
fere
nce,
Alb
uque
r
Triple point
Coverglass
-2000 V
Glue
Front surface
Side surface
he
e-e-
Differential voltage
ecra
ft C
harg
ing
Tech
nolo
Dielectric layer
Triple point
Solar Cell ~ -3000 V
Glue
-3000 V~ -3000 V
Glue
metal
dielectrics
Differential voltage
11th
Spa
ce
6
Structure S/C stucture potential
Microscopic structure and electron avalanche
Simulation at mesoscopic scale
Microscopic scale model
-24,
201
0
Fowler-Nordheim e-
rque
, NM
, Sep
tem
ber 2
0-
electric field amplification due to micron tipdielectrics
ogy
Con
fere
nce,
Alb
uque
r
metal
Differential potential
Electric field onSecondaryElectron avalanche
+
ecra
ft C
harg
ing
Tech
nolo
Secondary emission by electron impact + recollection
Electric field on the tip
Fowler-Nordheim
Secondary emission > 1 Cathode spot creation
=> ESD
11th
Spa
ce
7
electron emission
Tip geometry and field enhancement factor
• Field enhancement factor from 300 to 800• Tip length: 1 m • Tip radius (from emitting area): 1 - 10 nm
-24,
201
0rq
ue, N
M, S
epte
mbe
r 20-
ogy
Con
fere
nce,
Alb
uque
rec
raft
Cha
rgin
g Te
chno
lo
• Figure 5 and figure 8 extracted from:Effect of electrode surface finish on electrical breakdown in vacuum, D W Williams and W T
11th
Spa
ce
8
Williams, J. Phys. D: Appl. Phys. (5), 1972.
Experimental setup
Experimental Setup
• ONERA CEDRE vacuum chamber• Pressure: 3.10-6mbar• Barrier potential controlled by Vbias
-24,
201
0
Barrier potential controlled by Vbias
• Differential voltage by photoemission (as in flight)
Rotatingcubic holder
UV sourceElectrostatic probe
rque
, NM
, Sep
tem
ber 2
0-
HV Feedthroughs Electrostatic probecubic holder
Metal
Dielectric
ogy
Con
fere
nce,
Alb
uque
r
UV sourceSample
tank walls
Csat =
I
ecra
ft C
harg
ing
Tech
nolo
CCD Camera
CEDRE chamber
VbiasCsat 10nF
11th
Spa
ce
10
Nominal case – Perpendicular TP with Aluminum and Teflon
• The triple point is irradiated by the UV source• Only the chosen triple point line can generate
3mm
Al
UV
Irradiated zone
-24,
201
0
y p p gESDs thanks to a protecting mask 40mm
Dielectric FEP with aluminization on the rear side Protecting mask
rque
, NM
, Sep
tem
ber 2
0-
Al
g
ogy
Con
fere
nce,
Alb
uque
r
UV Irradiation zone
FEP
ecra
ft C
harg
ing
Tech
nolo
sample
UV protection
11th
Spa
ce
11
200 mmUV protection mask
Nominal case – Perpendicular TP with Aluminum and Teflon
• The voltage on the FEP is close to 0V • 8 ESDs have been obtained at 1 5
2
2,5
3
A
ESD #1ESD #2ESD #3
-24,
201
0
different locations of the triple point edge
• The ESD threshold is about -800V -0,5
0
0,5
1
1,5
curre
nt, A ESD #4
ESD #5ESD #6ESD #7ESD #8
rque
, NM
, Sep
tem
ber 2
0- -1-5 0 5 10 15 20
Time, µs
2,5
ogy
Con
fere
nce,
Alb
uque
r
1,5
2
curr
ent,
A
ecra
ft C
harg
ing
Tech
nolo
0
0,5
1Pe
ak
11th
Spa
ce
12
00 200 400 600 800 1000
Differential voltage threshold, V
Results of the parametric study
6Nominal caseMetal length 15mm
-24,
201
0 4
5
nt, A
Metal length 15mmCMX AR dielectricChange of incidence angleGold
rque
, NM
, Sep
tem
ber 2
0-
2
3
eak
curr
e Co planar (5nF)
ogy
Con
fere
nce,
Alb
uque
r
1
2P
ecra
ft C
harg
ing
Tech
nolo
00 500 1000 1500 2000 2500
Differential voltage threshold, V
11th
Spa
ce
13
Differential voltage threshold, V
SPIS ESD tool
The mock-up geometry and the boundary conditions
UV• Simulation conditions:
• Teflon (127 m) / aluminum (1.5 mm)• External box : 4 cm x 4 cm x 0.6 cm
x
z
40mm
UV
Irradiated zone
-24,
201
0
• TP zone : 1 m• tip length = 1 m• Barrier potential from 500 V to 2000V• from 800 to 300;
3mm
40mm
Al
rque
, NM
, Sep
tem
ber 2
0-
Barrier potential(Dirichlet condition) Refinement
boxTriple point with field
enhancement
;Dielectric FEP
ogy
Con
fere
nce,
Alb
uque
r
Symmetry conditions
enhancement
ecra
ft C
harg
ing
Tech
nolo
YZ
YX
Z
PerpendicularX
Y
Z
z
x
y
x
z
x
y
11th
Spa
ce
15
Y p
Metal plane + triple point(Dirichlet; V = 0)
Dielectric(implicit solver)
SPIS ESD tool
• ESD prediction scenario, two imbricate loops:• Barrier potential prediction• Tip geometry evaluation: field enhancement factor
loop (for a fixed tip length – 1m standard)
Tip and potential P t ti l
ESD prediction scenario
-24,
201
0
Data from UI
Init simulationFN current calculation
TipRecession interactor
FN current calculation
TipRecession interactor
no
Simulation Loop
yesno
Next time-step
Same potentialNext tip (lower
p pre-initializationPotential sweep
rque
, NM
, Sep
tem
ber 2
0-
Plasma integration
SC/Plasma interactions simulationDt
ScenarioPower balance on tip
Tip temperature
Mass losses by evaporation
Neutral density
Power balance on tip
Tip temperature
Mass losses by evaporation
Neutral density
noESD ?
yes
Is the tip melted ?
ogy
Con
fere
nce,
Alb
uque
r
x
RR
Circuit solver
simulationDuration
Scenario end ?
simulationLoop
Tip length
evaporation
New
Ionization probability
ESD ?
Tip length
evaporation
New
Ionization probability
ESD ?
ecra
ft C
harg
ing
Tech
nolo L
Scenario end ?
Results to UI
• Tip model (tipRecession interactor):• Tip model based on a cylindrical geometry
11th
Spa
ce
16
• Tip model based on a cylindrical geometry• Zero dimensional thermal model (compared to Rossetti ….)
Barrier potential and emitted currentste
ntia
l (V
)
1500
2000
-24,
201
0
Bar
rier p
ot
0
500
1000
ESD T i
ESD Trig. = 650
ESD Trig. = 600
rque
, NM
, Sep
tem
ber 2
0-
)
10-4
1ESD Trig. = 800
ESD Trig. = 750
ESD Trig. = 700
650
Ph t i i
ogy
Con
fere
nce,
Alb
uque
r
Cur
rent
(A)
10-12
10-8
Fowler-Nordheim emission
Photoemission
ecra
ft C
harg
ing
Tech
nolo
0 50 100 150 200 250 300 350 40010-20
10-16
11th
Spa
ce
17
0 50 100 150 200 250 300 350 400
Time (s)
Electron emission : densities in volume-2
4, 2
010
rque
, NM
, Sep
tem
ber 2
0-
Dielectric Metal Dielectric Metal
ogy
Con
fere
nce,
Alb
uque
r
S d l t d itF l N dh i l t d it
DielectricMetal
DielectricMetal
ecra
ft C
harg
ing
Tech
nolo Secondary electrons densityFowler-Nordheim electrons density
• F-N primary electrons:• directed toward the barrier potential
M d it h 1020 3 (S h ff t)
11th
Spa
ce
18
• Max density can reach 1020 m-3 (Space charge effect)• Secondary electrons are present on the top side of the dielectric due to hoping
Parametric study: comparison of the numerical and experimental results
Parametric study – reference case
6
-24,
201
0
4
5 Reference caseAluminum-Teflon 127 m3 mm Gap
3mm
40mm
rque
, NM
, Sep
tem
ber 2
0-
3
4
ak c
urre
nt (A
)
Irradiated Gap
ESD Trig. = 800
ESD Trig. = 600
3mm
ogy
Con
fere
nce,
Alb
uque
r
2
Pea 800 600
ecra
ft C
harg
ing
Tech
nolo
0
1
500 1000 1500 2000 2500
Experimental resultsNumerical results
11th
Spa
ce
20
Differential voltage (V)
Parametric study – Sun incidence angle
6
Not Irradiated Gap
-24,
201
0
4
5 Reference caseAluminum-Teflon 127 m3 mm Gap
Not Irradiated GapAluminum-Teflon 127 m3 mm Gap
rque
, NM
, Sep
tem
ber 2
0-
3
4
k cu
rren
t (A
) Irradiated Gap
400V 3mm
40mm
ogy
Con
fere
nce,
Alb
uque
r
2Peak
350V
3mm
ecra
ft C
harg
ing
Tech
nolo
0
1
500 1000 1500 2000 2500
Experimental resultsNumerical results
11th
Spa
ce
21
Differential voltage threshold (V)
Parametric study – Gap length
6
15mm Gap
-24,
201
0
4
5Reference caseAluminum-Teflon 127 m3 mm Gap
15mm GapAluminum-Teflon 127 mIrradiated Gap
rque
, NM
, Sep
tem
ber 2
0-
3
4
ak c
urre
nt (A
)
Irradiated Gap
900V 3mm
40mm
ogy
Con
fere
nce,
Alb
uque
r
2
Pea
800V
ecra
ft C
harg
ing
Tech
nolo
0
1
500 1000 1500 2000 2500
Experimental resultsNumerical results
11th
Spa
ce
22
Differential voltage threshold (V)
Parametric study – Dielectric material
6
Aluminum-CMX 100 m3 mm Gap
-24,
201
0
4
5
)
Reference caseAluminum-Teflon 127 m
3 mm GapIrradiated Gap
rque
, NM
, Sep
tem
ber 2
0-
3
4
k cu
rren
t (A
) 3 mm GapIrradiated Gap
CMX-AR □ = 2.6×1016
CMX-AR □ = 1016
CMX-AR □ = 5×1015
ogy
Con
fere
nce,
Alb
uque
r
2Peak
Teflon □ = 1016
3mm
40mm
ecra
ft C
harg
ing
Tech
nolo
0
1
500 1000 1500 2000 2500
Experimental resultsNumerical results
11th
Spa
ce
23
500 1000 1500 2000 2500Differential voltage threshold (V)
Lessons to be taken from the experiments and numerical simulations
• The numerical results qualitatively agree with the experiments• It is definitely more difficult to trig ESDs in the following cases
-24,
201
0
• Low photon incidence angle• Longer metallic plate• Co-planar geometry
More conductive dielectric
rque
, NM
, Sep
tem
ber 2
0- • More conductive dielectric
• Strong effect of • surface resistivity of the front edge
ogy
Con
fere
nce,
Alb
uque
r • surface resistivity of the front edge• Tip geometry (radius and shape)
• Need for complementary experimental data (surface resistivity measurement,
ecra
ft C
harg
ing
Tech
nolo
p y p ( y ,surface distribution and type of microscopic structures)
11th
Spa
ce
24
Conclusion
Conclusion
• Large experimental parametric study
-24,
201
0
• SPIS ESD tool:• Complete ESD scenario:
• 3D Electron avalancheThermal model of the tip
rque
, NM
, Sep
tem
ber 2
0- • Thermal model of the tip• ESD threshold estimation
• Simulation of the charging and electron avalanche well described• Automatically adapted time steps: from 1s to 10-12s
ogy
Con
fere
nce,
Alb
uque
r Automatically adapted time steps: from 1s to 10 s• 3 length scales simulated
• Good qualitative agreement between experiments and numerical results
ecra
ft C
harg
ing
Tech
nolo
q g p
• Need of further experimental data to provide design engineers with quantitative/predictive ESD risk tool
11th
Spa
ce
26
Perspective
• Experimental investigation on surface resistivity to get more data
-24,
201
0
• Photo conductivity modeling improvement
• More complex geometry:
rque
, NM
, Sep
tem
ber 2
0-
p g y• solar cell edge • here, it is close to the interconnects
ogy
Con
fere
nce,
Alb
uque
r
• Effect of multiple tips (tip distribution on the surface)
• Emission from semi-conductors
ecra
ft C
harg
ing
Tech
nolo
• Model the plasma generation during the transition to arc
11th
Spa
ce
27
Questions ?