Progress on Target Survival

1
Progress on Target Survival

• Presented by A.R. Raffray
• Other Contributors
• B. Christensen, M. S. Tillack
• UCSD
• D. Goodin
• General Atomics
• HAPL Meeting
• UCLA
• Los Angeles, CA
• June 2-3, 2004

2
Outline

• Benchmark analysis with U. Roch. LLE (D. Harding)
• - Purchase more advanced version of DSMC
• - Number flux and heat flux analysis
• - Effect of accommodation and sticking
  coefficients
• Modeling experimental results from LANL
• (J. Hoffer/D. Geller)

3
DS2V was Purchased for Modeling the Thermal
Loading from the Background Gas

• Capabilities
• Axisymmetric flow.
• Adjustable sticking (condensation)
  coefficient.
• Adjustable accommodation coefficient.
• Output
• Heat flux, number flux, drag force,etc
• Injected Target Modeling
• Simulated by flow over stationary target
  (hydrodynamic similarity).
• Could not find a correct way of modeling
  moving target in stationary gas with this
  version.

• Figure Above Shows the Temperature Field Around a
  Direct Drive Target.
• Xe flowing at 400 m/s in the positive x-dir.
  4000 K Xe stream temperature.
• 3.22x1021 m-3 Xe stream density.
• Sticking coefficient 0.
• Target surface temperature 18 K.

4
The Number Flux and Heat Flux at the Target
ReachQuasi-Steady State in a Short Time
Figure Above Shows the Number Flux and Heat Flux
Around a Direct Drive Target. Xe stream
flowing at 400 m/s. 4000 K stream temperature.
3.22x1021 m-3 stream density. Sticking
coefficient 0. Target surface temperature
18 K.
5
As the Stream Density Is Increased the Sticking
Coefficient (sigma) Has a Greater Effect

• The number flux is not a function of the sticking
  coefficient (sigma) when the stream density is
  low.
• The number flux decreases with increasing sigma
  when the stream density is high.
• Kinetic theory and DS2V show good agreement
  (sigma1, no shielding effect).

Low Density Stream, n 3.22x1019 m-3
High Density Stream, n 3.22x1021 m-3
6
The Heat Flux is Significantly Affected by the
Stream Density, Temperature, and Sticking
Coefficient

• The effect of latent heat is not included in
  DS2V needs to be included in post processing.
• By neglecting the latent heat the shielding
  effect of a non-condensing gas (sigma 0) is
  seen.
• Virtually no shielding for the low density
  stream.
• Significant shielding for the high density
  stream.
• The rapid change in heat flux with position
  suggests that the average max. heat flux could be
  reduced by tumbling the target.

Low Density Stream, n 3.22x1019 m-3
High Density Stream, n 3.22x1021 m-3
7
Conclusions from DS2V Study

• Simulate injected target situation by flow over
  stationary target (hydrodynamic similarity)
• The number flux and heat flux at the target
  reach quasi-steady state in a relatively
  short time
• (no need to run longer except if outside
  conditions (gas) change)
• The effect of latent heat is not included in
  DS2V needs to be included in post
  processing.
• Shielding effect dependent on sticking
  coefficient for high density gas
• - Virtually no shielding for the low density
  stream (1 mTorr).
• - Significant shielding for the high density
  stream ( q reduced by a factor of 2 or more
  when sigma changes from 1 to 0 for example case
  at 100 mTorr)
• Experimental determination of the sticking
  coefficient is needed (U. Roch.)
• The accommodation coefficient should also be
  determined if the sticking coefficient is
  found to be significantly less than one.

8
Initial Modeling of Direct Heating Experiments at
LANL (J. Hoffer/D. Geller)

• 1-D spherical numerical model.
• Constant heat flux.
• Initial temperature 18 K.
• DT thickness 400 mm.

9
The Time to Triple Point, as Predicted by the Two
Numerical Models, is Generally Consistent with
Experimental Results
10
There are Large Differences in the Melt Layer
Thickness Results
11
Summary
Encouraging that melting time seems to be
predicted quite accurately, Some question
marks on melt layer thickness experimental and
modeling results Modeling these experimental
results can be improved - Create 1-D
cylindrical model. - Allow for variable heat
flux (for melt layer computations) - Code
optimization meshing, time-steps, assumed
temperature range over which melting
occurs - Modeling experimental
set-up Experimental uncertainties need to be
better understood - Measurement how to specify
melt layer boundary - Heat flux changes when
melting starts Working with our LANL colleagues
on how to produce experimental results more
amenable for our model and on how to improve
model to simulate a wider range of experimental
conditions
12
Please Refer to Brian Christensens Poster for
More Details on our 1-D Target Thermomechanics
Modeling (Including Phase Change) and DS2V
Modeling

• Brian has completed his MS Thesis on this - a
  summary of which will be submitted for journal
  publication
• Thesis defense next week
• His results has shed much light on the different
  processes affecting target survival
• He has included recommendation on future work
  (2-D or quasi 2-D modeling experiments)
• We have identified a new student to continue this
  work as from the Fall (after the Olympics!)