M. Yoda, S. I. AbdelKhalik,

1
Update on Thermal Performance of the Gas-Cooled
Plate-Type Divertor

• M. Yoda, S. I. Abdel-Khalik,
• D. L. Sadowski and M. D. Hageman
• Woodruff School of Mechanical Engineering

2
Objective / Motivation

• Objective
• Experimentally evaluate and validate thermal
  performance of gas-cooled divertor designs in
  support of the ARIES team
• Motivation
• Leading divertor designs rely on jet impingement
  cooling to achieve desired performance
• Accommodate heat fluxes up to 10 MW/m2
• Performance is robust with respect to
  manufacturing tolerances and variations in flow
  distribution
• Extremely high heat transfer coefficients (50
  kW/(m2?K)) predicted by commercial CFD codes used
  for the design
• Experimentally validate such numerical
  predictions

3
Approach

• Design and instrument test modules that closely
  match divertor geometries
• Conduct experiments at conditions matching and
  spanning expected non-dimensional parameters for
  prototypical operating conditions
• Reynolds number Re
• Use air instead of He
• Measure temperature distributions and pressure
  drop
• Compare experimental data with predictions from
  CFD software for test geometry and conditions
• Nu(Re), ?P(Re)

4
Some History

• Investigated leading gas-cooled divertor designs
• FZK Helium-Cooled Multi-Jet (HEMJ) Finger
• Norajitra et al. 2005
• Ihli et al. 2007 Crosatti et al. 2009
• ARIES-CS T-Tube Ihli et al. 2007 Raffray et
  al. 2008
• Crosatti et al. 2007 Abdel-Khalik et al. 2008
  Crosatti et al. 2009
• ARIES-Pathways Plate-Type Design
• Malang Wang et al. 2009
• Variant with metal open-cell foam Gayton et al.
  2009 Sharafat et al. 2007
• Variant with pin-fin array In progress

5
Outcomes

• Enhanced confidence in predicted performance by
  commercial CFD codes at prototypical and
  off-normal operating conditions
• FLUENT
• Use validated CFD codes to optimize/modify
  divertor designs
• Predict sensitivity to changes in geometry and
  operating conditions to define and establish
  manufacturing tolerances

6
Plate-Type Divertor Design

• Covers large area (2000 cm2 0.2 m2) divertor
  area O(100 m2)

• HEMJ, T-tube
• cool 2.5, 13 cm2
• Accommodates
• up to 10 MW/m2
• without exceeding
• Tmax ? 1300 C, ?max ? 400 MPa

20
100 cm
Castellated W armor 0.5 cm thick

• 9 individual manifold units with 3 mm thick
  W-alloy side walls brazed together

7
GT Test Module
Heated brass shell

• Jet issues from 0.5 mm slot, then impinges on and
  cools underside of W-alloy pressure boundary
• Coolant flows along 2 mm gap, exits via outlet
  manifold
• Original design Malang 2007
• Use air as coolant
• Reynolds number Re 1.1?104 6.8?104 (vs.
  3.3?104 at nominal operating conditions)
• Nominal heat flux q?nom 0.2 0.75 MW/m2

Al inner cartridge
8
Modified Manifold Design

• To reduce thermal stress, current modified design
  adds Wang et al. 2009
• 2 mm stagnant He region outside outlet manifold
• 8 mm thick W-alloy base

9
Al Inner Cartridge

• Inlet, outlet manifolds embedded inside Al
  cartridge
• Manifolds 19 mm ? 15 mm ? 76.2 mm
• 2 mm ? 76.2 mm slot
• Coolant enters outlet
• manifold via holes
• Side wall bolted on

In
Out
10
Brass Outer Shell

• Models pressure boundary
• 5 TC in shell to measure cooled surface
  temperature distribution 2 in center (1,5) and
  (3,4) at same depth 0.5 mm from surface
• Brass shell heated
• by heater block
• k for brass
• similar to
• that of W-alloy

11
Pin-Fin Array

• Can thermal performance of leading divertor
  designs be further improved?
• Mo open-cell foam in 2 mm gap increased HTC by
  4050, but also increased ?P by similar
  fraction Gayton et al. 2009
• In HEMP, a variant of HEMJ, coolant impinges on
  pin-fin array Diegele et al. 2003
• Combine plate with pin-fin array
• 808 ?1.0 mm ? 2.0 mm pin fins (nearly) contacting
  Al cartridge on 1.2 mm pitch
• 2 mm clear area for impinging jet
• Pin fins EDMed into inside of brass shell

12
Heated Test Section
Copper heater block
Graphite shim
Brass outer shell
Gasket
Aluminum cartridge
13
GT Air Flow Loop
Outlet P, T measurement
Inlet P, T measurement

• Cu heater block
• 3 cartridge heaters
• 6 TC in neck measure q?
• 2 TC at top monitor max. Cu temperature

14
Effect of Pin Fins

• Nu from TC data
• Nearly uniform T along slot
• Nu based on gap width, k at 300K and effective
  HTC (for pin fins)
• q?nom 0.20.75 MW/m2

Pin fins Bare surface
Nu
Nominal operating condition

• ? TC 1 ? TC 4
• ? TC 2 ? TC 5
• ? TC 3

Re (/104)
15
Comparison Pins vs. Bare
Mass flow rate g/s

• Ratio of Nu and ?P for cooled surface with,
  without pin-fins
• Pin-fins with 260 more surface area improve
  cooling performance by 150200 while
  increasing pressure drop by 4070

?Pp / ?P
Nup / Nu
Nominal operating condition
Re (/104)
16
Summary

• Designed and studied experimental test modules
  modeling leading He-cooled divertor designs
• T-tube, HEMJ finger, plate
• Conducted dynamically similar thermal-hydraulics
  experiments matching and spanning expected
  prototypical operating conditions
• Used commercial CFD software to predict
  performance of experimental test modules
• Good agreement between experimental data and
  model predictions (including those from other
  groups)
• Use validated codes to predict performance of
  gas-cooled components with complex geometries

17
Conclusions

• Plate-type divertor pin-fin array promising
  design
• Smaller number of divertor modules required ?
  reduced cost, complexity
• Two- to three-fold enhancement by pin fins ? can
  accommodate heat fluxes much higher than 10 MW/m2
• Initial results for un-optimized configuration
  use CFD to suggest improvements to current
  experimental design
• Effect of pin pitch, diameter
• Effect of slot width

18
Next Steps

• To complete ARIES-Pathways study
• Validate CFD codes (e.g. FLUENT) and plate models
  with experimental data
• Model pin-fin array
• Use validated CFD codes to optimize/modify
  pin-fin layout
• Predict maximum heat flux that can be
  accommodated by optimized pin-fin/plate-type
  divertor
• Predict pressure drop across optimized pin-fin
  array