Vapor Condensation Study

1
Vapor Condensation Study for HIF Liquid Chambers
by Patrick Calderoni UCLA Fusion Engineering
Sciences 15th International Symposium on
Heavy Ion Inertial Fusion Princeton Plasma
Physics Laboratory Princeton, NJ, June 7-11, 2004
2
Experimental research work major accomplishments
Developed an innovative and inexpensive scheme to
generate flibe vapor in conditions relevant to
fusion technology design studies involving a
liquid protection scheme (HIF, IFE, Z-pinch)
Measured flibe vapor clearing rates suggest that
high repetition rates in HIF power plants are
feasible provided that high purity of the molten
salt is ensured
Found that for flow conditions characterized by
high kinetic energy flibe vapor condensation is
partially inhibited on metal surfaces
perpendicular to the main component of the vapor
velocity
3
HYLIFE-II parameters relevant to vapor
condensation studies (Moir, 1994)
Total mass generated from x-ray absorption on
liquid surfaces 14 kg Total volume for vapor
expansion 280 m3 Initial n 0.9x1018 /cm3
(0.5x1018 /cm3) Recovered n 3x1013 /cm3
(2x1015 /cm3)

Energy from explosion coupled with x-ray and
debris 110 MJ Energy density 7.85 kJ/g (7.5
kJ/g)
Structural surface for condensation 40
m2 Surface from droplet injection 1060 m2 Ratio
of surface area per unit mass of generated vapor
785 cm2/g (4300 cm2/g for LiF and 10205 cm2/g for
flibe)
Droplet spray injection design T 843 K Spray
flow rate 2.4x103 kg/s (1.6 of main flow rate)
4
Experimental approach staged design of vapor
generation facility
Characterization of superheated vapor source in
comparison to other experiments using
electro-thermal sources
Limited availability, cost and toxicity of
materials demonstrate efficiency, repeatability
and reliability before using flibe
Reduce residual non-condensable gases
Diagnostic development
5
time 0
Stage 1 Lexan with Argon background (1 Torr)
time 820 ?s
Typical vapor parameters in the source
n 1019 - 1020 / cm3 T 1-3 eV
time 1640 ?s
Argon background is ionized (10-100 ns) forming
initial plasma column
Energy stored in cap banks maintains plasma at
1-3 eV for 100 micros
Injected electrical power radiated to surface,
ablates material of interest
Pressure gradient drives injection, ablation
balances axial mass loss
6
Stage 2 Teflon and LiF in vacuum chamber
current in
vapor source
current out
time of flight view ports
expansion chamber volume 4000 cm3 surface area
1720 cm2
7
Cathode ¼ D 10 long W rod
Stage 3 Flibe vapor generation
Total length 34 cm
Expansion volume 400 cm3 Surface area 420 cm2
SS witness plates for SEM and EDX analysis
Insulation high-vacuum ceramic breaks
Pressure sensor, water cooled (Tmax 260 C)
Flibe liquid pool 1.6 cm3 volume
Anode Nickel crucible with embedded high density
cartridge heater
8
High-speed camera frames sequence from flibe
discharge
220 ?s
100 ?s
120 ?s
240 ?s
260 ?s
140 ?s
160 ?s
280 ?s
300 ?s
180 ?s
320 ?s
200 ?s
9
Pressure data and residual gas composition flibe
condensation is completed in 30 ms no residual
BeF2 traces at 47 amu
t500 6.58 ms
t300 4.27 ms (4.22 ms at 1.44 kJ)
300 C 500 C
H2
28,16 amu hydrocarbons
44 amu CO2
10
Comparison with HYLIFE-II chamber clearing models
p(T) ? n(T) assumes thermodynamic equilibrium
with a liquid surface
n n0 x e-t/T n0 0.9x1018 /cm3 nend 3x1013
/cm3 Clearing period for HYLIFE-II 68 ms
11
Measured composition of flibe vapors in
equilibrium with a liquid surface
hydrocarbons
CO2
H2
BeF2
at 460 C
Heating sequence (linear) from 460 C to 700 C
(about 30 min)
12
Additional data post-analysis of side collecting
plates
Collecting plates parallel to the radial
direction
Surfaces are gold plated for SEM analysis
Film forms during first expansion phase (100 ?s)
when vapor velocity is highly directional
At low T film is thinner and breaks due to quick
cooling and solidification
460 C
300 C
Drops condense in the chamber volume after the
velocity has become uniform and deposit on the
liquid film without merging
13
Additional data post-analysis of front
collecting plates
Collecting plates perpendicular to the radial
direction
Film condensation is inhibited at 300 C
300 C
460 C
At 460 C thin film starts at fixed r from plate
side center (flow stagnation point)
EDX analysis confirms qualitative results
14
Additional data post-analysis of collecting
plates
Further evidence of volumetric condensation
300 C
300 C
side plate
front plate
460 C
Evidence of liquid displacement by the pressure
front generated during the discharge large
liquid drops are entrained in the flow and
deposit around the crucible and the collecting
plates
Large drop is flibe
15
Conclusion
Condensation rates of flibe vapor in conditions
relevant to IFE power plant studies have been
measured experimentally - Vapor density decays
exponentially with a time constant of 6.58 ms in
the range between 5x1017 cm-3 and 2x1015 cm-3
Extending to HYLIFE-II expected density cycles
the vapor clearing rate is 68 ms, compatible with
the desired 6 Hz repetition rate
Data suggest that for flow conditions
characterized by high kinetic energy flibe vapor
condensation is partially inhibited on surfaces
normal to the main component of the vapor
velocity
Control of the impurities dissolved in the molten
salt is a fundamental issue for applications that
require recovery of vacuum conditions in the 1013
/cm3 range
16
Extra slides
17
Discharge parameters
Teflon I and V for 10 and 5 disk configuration
LiF I for different energy experiments
I for flibe experiments
18
Liquid protection for Inertial Fusion Energy
power plant chambers
HYLIFE-II concept
Neutron damages and activation of flowing liquid
accumulate only in the short residence period -
no blanket replacement required, increases
availability
Thick liquid pockets or thin liquid layers
shield chamber structures fluid mechanics
questions replace materials questions
19
The molten salt flibe
For fusion system design
For small scale experiments
Limited material availability
Beryllium safety hazard
Uncertain composition and purity level of
available material
Lack of data on physical and chemical properties
20
Flibe composition is a compromise between melting
temperature (structural costs) and viscosity
(pumping cost)
2 LiF 1 BeF2 in moles
21
Flibe low vapor pressure is a key property for
efficient driver coupling with the fusion target
at high repetition rates
Reference data for flibe vapors are based on the
assumption of thermodynamic equilibrium at the
liquid interface and ideal gas equation
In equilibrium with the liquid at 600 C, flibe
vapor pressure is about 1 mTorr, corresponding to
a density of 1.2 1013 cm-3
22
Weight loss measurements
18 experimental runs with 10 Teflon disks as
sleeve material 1 with 5 disks
5.76 kJ
2.54 kJ
Volume ratio between source and chamber 125
Evidence of C soot deposition on source component
23
Dynamic sensor mounted at center of back plate
(high noise from vibration)
Pressure data Lexan and Teflon
about 5 times decay
Teflon CF2 chain Residual gases (20-70
Torr) 28 amu (C2H2 - C2H4 - C2H6) 69 amu
(CF4) 20 amu (HF)
24
Jet velocity optical measurement system
Diode axis separation 7.62 cm Peak time delay 6
microseconds Estimated initial vapor
velocity 10110 m/s
25
Jet velocity optical measurement system
Sensor closer to center of chamber sees a second
peak in the light, about 2ms after
triggering Peak due to first reflection of the
vapor jet at the chamber bottom Vapor cools and
expands in the chamber, emitting front does not
reach the upper sensor Estimated average
velocity of vapor in the chamber after first
reflection 320 m/s - 4 kV 210 m/s - 3 kV
26
Flibe experiments results
Exponential decay fits data points good
Decay time constant t1 is a measure of the
condensation rate
t1 4.22 ms
27
Flibe experiments results
Exponential decay fits data points good
Decay time constant t1 is independent from
initial particle density
t1 4.27 ms
28
Additional data emission spectroscopy
Strong lines from neutral and first ionization
state of both lithium and beryllium atoms are
present in the spectrum

• Be vapors diagnostic development
• measure at different times the ratio of line
  intensity of Li I and Li II transitions
• steady-state emission calibration with langmuir
  probe will associate line ratio with
  thermodynamic vapor parameters (pressure and
  density)