Modeling Carbon Diffusion in W-Armor

1
Modeling Carbon Diffusion in W-Armor Shahram
Sharafat and Nasr Ghoniem Graduate Students
Jaafar El-Awady Michael Andersen Qyiang Hu
Jennifer Quan Andrew Chen Wen Guo (Alice
Ying) University of California Los Angeles, CA.
The High Average Power Laser (HAPL) Program
WorkshopRochester, New YorkNovember 8 and 9,
2005
2
OUTLINE

• Brief Summary of UCLA Activities
• Roughening Modeling
• Interface Bond Strength Measurements
• Ion Implantation
• Carbon Implantation Profile
• Multi-physics Simulation
• Buildup ?

3
Modeling Surface Roughening
POSTER
Michael Andersen, (Ph.D. Thesis)

• Goal Predict crack formation by modeling surface
  roughening
• Surfaces ROUGHEN to alleviate stresses.
• Solid buildup decreases elastic energy by growing
  tips (low stress).
• Deepening valleys (material withhigh elastic
  energy is removed).
• Sources of Stress
• Thermo-mechanical (Biaxial)

4
Measurement of W - F82H Bond StrengthUsing Laser
Spallation Interferometry
POSTER
Jaafar El-Awady, Jennifer Qua, et. al.

• Goal Quantify the Interface bond strength
  between tungsten and F82H
• Approach Laser Spallation interferometry
  measurements
• Experiment ITER (JAERI) HIPd W/F82H samples
  were tested. Used 6 different laser fluence
  energies to determine the critical energy.

Laser Fluence 1065 mJ
Interfacialstress history
Laser Fluence (mJ) 613 1065 1329 1577 1708 1737
Failure No Failure No Failure Some Cracks generates at the interface Severe interfacial damage Severe interfacial Damage Severe interfacial Damage
Results Severe interfacial damage occurred above
1577 mJ. Analysis results in a bond strength
between 300 and 450 MPa (depending on Youngs
Modulus of W 390 and 195 GPa, respectively)
5
Ion Implantation(1) Helium(2) Carbon
6
Carbon Implantation Profile (SRIM)

• Threat Spectra (405 MJ)
• C 1.07x1019 /shot
• Peak 0.4 um
• Avg. C/ W Ratio 2.5x10-7 (apa)

• Pellet Geometry
• CH layer 3 mm
• r 2.264 mm
• C/ W4.1x10-7 (apa)

(Perkins, HAPL Oct04)
7
C-Buildup No Carbon Diffusion (SRIM 2003)
HAPL June 05

• 154 MJ Target

8
Carbon Diffusion Model

• Multi-physics problem
• Consider Carbon Diffusion with R 0 No WC or
  W2C
• Future work will consider carbide formation (R ?
  0)

Carbon Diffusion in Tungsten ?
http//FusionNET.seas.ucla.edu
(Eckstein, 1999)
(Eckstein, 1999)
9
Multi-physics Diffusion Solution

• Used ANSYS to solve the diffusion equation
• it worked
• Needed to couple temperature with diffusion in
  ANSYS
• did not work
• Briefly considered
• Neutron Transport

Use COMSOL (PDE Solver) to solve the coupled
thermal diffusion model
10
Carbon Concentration Profiles
IC Profile C(to)
14 mm
1 mm
11
Carbon Concentration Profile
12
Carbon Concentration Profile
13
Quasi Steady State Diffusion Approximation

• Quasi Steady State Parameters
• Carbons per shot 1.07 ? 1019 ions
• Avg. Carbon Flux (r10.1 m, 10 Hz) 8.35 ?
  1016 m-2 s-1
• Avg. W-Temperature (50 um) 500 oC
• Diffusion Pre-exponential (Do) 3.15 ?107 m2 s-1
  
• Activation Energy (Q) 1.78 eV

14
Carbon Diffusion Observations

• The high W-surface layer temperatures facilitate
  C diffusion in W.
• Carbon does not preferentially diffuse out from
  surface but also spreads inwards (towards F82H).
• C-to-W ratio of 1 can be reached in lt10 days (2
  mm at 10 Hz r 10.1 m, 405 MJ).
• Quasi Steady State Analysis shows that F82H steel
  wall is protected from C pickup (C reaches 20
  um lt 1 year).
• Formation of W2C and WC was not considered
  (would slow C-diffusion until C/W ratio gt 1).
• Evaporation of C from WC surface was not
  considered It would increase C loss from
  surface WC?W2C above 2000 oC(1).

1Yamada, JNM 2000
15
WC-W2C Thermo-Mechanical Impact

• Concerns
• Lower KIC , s affects mechanical response (crack
  nucl. growth).
• Lower k, Tm may impacts thermal performance of
  FW.
• High C-content might impact Tritium release rates
  
• W2C forms at 800 oC and WC forms(1) at 1000 oC
• Helium release from WC above 1200 oC is similar
  to W(2).

2 SiC-B4C data Hino-JNM-1999
1Hatano 2005 Roth 2001