Yoda

Experimental Results for Thin and Thick Liquid WallsM. Yoda and S. I. Abdel-Khalik: 

G. W. Woodruff School of Mechanical Engineering Atlanta, GA 30332-0405 Experimental Results for Thin and Thick Liquid Walls M. Yoda and S. I. Abdel-Khalik

Overview: 

Overview Thin liquid protection: Experimental study of high-speed thin liquid films on downward-facing surfaces around cylindrical dams [J. Anderson, D. Sadowski] Ports for beam entry and target injection Effect of surface wettability Thick liquid protection: Experimental studies of turbulent liquid sheets [S. Durbin, J. Reperant, D. Sadowski] Quantify surface smoothness of stationary liquid sheets using planar laser-induced fluorescence (PLIF) technique Impact of initial conditions: nozzle geometry, flow straightener design, flow straightener blockage

Thin Liquid Protection: 

IFE chamber (Prometheus) First Wall Injection Point Detachment Distance xd Liquid Film/Sheet X-rays and Ions Thin Liquid Protection

Objectives: 

2 mm nozzle 17 GPM 10.7 m/s 10o inclination Re = 20000 2 mm nozzle 17 GPM 10.7 m/s 10o inclination Re = 20000 Objectives Determine “design windows” for high-speed liquid films proposed for thin liquid protection of IFE reactor chamber first wall In the absence of film dryout, films most likely to detach on downward-facing surfaces on top endcap Chamber curvature probably negligible: chamber radius ~6.5 m, vs. film radius of curvature at detachment point O(1 cm) How does film flow around cylindrical obstructions, or dams (e.g. beam ports)?

Experimental Apparatus: 

A Glass plate (1.52  0.40 m) B Liquid film C Flow straightener D Film nozzle A B C D Adjustable angle  x z gcos  g Experimental Apparatus

Cylindrical Obstructions: 

Cylindrical Obstructions Cylindrical dam/obstruction at x = 7.6–9 cm from nozzle exit Held in place by permanent rare-earth magnet above glass plate Vary cylindrical dam height H and diameter D Height (axial dimension) H ~  How does high-speed film flow around obstructions (e.g. beam ports)?

Experimental Parameters: 

Experimental Parameters Nozzle exit thickness (z-dimension)  = 0.1, 0.15, 0.2 cm Average speed at nozzle exit U0 = 1.9–5.1 m/s Jet injection angle  = 0°, 30° Cylindrical dam outer diameter D = 1.58, 2.54 cm Cylindrical dam height H = 0.051, 0.12, 0.24 cm Reynolds number Re = U0 / = 3800–9800 Froude number Fr = U0 / [(g cos ) ]½ = 15–55 Weber number We = U02 / = 100–700 Cylindrical dam aspect ratio AR = H/D = 0.02–0.093 Film nozzle aspect ratio ARf = (5 cm)/ = 25–50

Results: 

Results Detachment Type I  = 0.15 cm Re = 7600 H = 0.24 cm D = 2.54 cm AR = 0.093 Dam Detachment Type II  = 0.15 cm Re = 3800 H = 0.24 cm D = 2.54 cm AR = 0.093 Dam

Detachment Type I: H > : 

 = 0.1 cm Re = 3800 H = 0.12 cm D = 2.54 cm Detachment Type I: H >  AR = 0.047  = 0  = 30

Detachment Type II: H > : 

 = 0.2 cm Re = 3800 H = 0.24 cm D = 1.59 cm AR = 0.15 Detachment Type II: H >   = 0  = 30

Flow Over Dam: H < : 

 = 0.1 cm Re = 3800 H = 0.051 cm <  D = 2.54 cm AR = 0.02 Flow Over Dam: H <   = 0  = 30 Flow Over Dam

Summary: Obstructions: 

H > : Detaches around cylindrical obstructions at outer leading edge (Type I) or inner trailing edge (Type II) H < : Flows over obstruction, blocking hole Occurs at lowest speeds (and Re) Cylindrical beam ports incompatible with wet wall concept “Streamlined” fairings? No beam ports on upper endcap  fewer beams? Summary: Obstructions For all cases studied, film flow on downward-facing surfaces either:

Surface Wettability: 

Compare film average detachment length xd, lateral spread W for water on two surfaces with very different contact angles/wettability Water on glass: contact angle  30° Water on glass coated with Rain-X®: contact angle  85° Surface Wettability Water on glass w/Rain-X (drop dia. ~4 mm; vol. 0.4 mL) Water on glass (drop diameter ~5 mm; volume 0.4 mL)

Wettability Effects: 

Wettability Effects x /  W /Wo At Re = 3800, xd /   100 for Rain-X surface; 180 for glass At Re = 14700, xd /   550 for Rain-X surface; 700 for glass  Glass, Re = 3800  Rain-X, Re = 3800  Glass, Re = 14700  Rain-X, Re = 14700  = 0.2 cm  = 0 Preliminary data Non-wetting surface  Earlier detachment

Future Work: 

Streamlined beam ports Examine effect of surface wettability/contact angle Non-wetting surfaces (a la Prometheus) worse: earlier detachment, smaller lateral spread Measure film thickness with ultrasonic probes Measure lateral (y) velocity profile across film using laser-Doppler velocimetry (LDV) Future Work Sketch courtesy L. Waganer

Thick Liquid Protection: 

Protect IFE reactor chamber first walls by using molten salt or liquid metal “curtain” to absorb neutrons, X-rays, ions and target debris from fusion events HYLIFE-II conceptual design based on turbulent liquid sheets as “building block” Oscillating slab jets, or liquid sheets, create protective pocket to shield chamber sides Lattice of stationary liquid sheets shield front and back of chamber while allowing beam propagation, target injection Sketches courtesy P.F. Peterson Thick Liquid Protection

Design Issues: 

Effective protection  minimize clearance between liquid sheet free surface and driver beams, or minimize surface ripple Irradiation of final focus magnets Interferes with target injection, beam propagation How do various jet (nozzle, flow straightener) designs impact the free-surface geometry and its fluctuations? Robust protection  thick liquid protection system must withstand occasional disturbances How does partial blockage of the flow straightener (due, for example, to debris) affect the free-surface geometry and hence surface ripple? Design Issues

Objectives: 

Objectives Liquid probability distribution (LPD): probability of finding liquid at any given spatial location Mean surface ripple z: Average standard deviation of the z-position of the free surface Study turbulent vertical sheets of water issuing downwards into atmospheric pressure air at Reynolds numbers Re = Uo/ = 53,000–130,000 (prototypical Re = 200,000) Uo average speed at nozzle exit;  nozzle thickness (short dimension);  fluid kinematic viscosity Quantify impact of nozzle designs and blockage on surface ripple in liquid sheets typical of HYLIFE-II

Experimental Apparatus: 

A Pump B Bypass line C Flow meter D Pitot tube E Flow straightener F Nozzle G Oscillator H Sheet I 400 gal tank J Butterfly valve K 350 gal tank Pump-driven recirculating flow loop Test section height ~1 m Overall height ~5.5 m Experimental Apparatus

Nozzle Geometries: 

A B C x y z Fabricated with stereolithography rapid prototyping Nozzle exit dimensions 1 cm ()  10 cm 2D contractions: nozzle z-dimension contracts from 3 cm to 1 cm at exit Three different nozzles A Matched circular-arc contraction B 5th order polynomial contraction C B with rounded corners Nozzle Geometries

PLIF Technique: 

Water dyed with fluorescein Jet illuminated by Ar+ laser light sheet at 514 nm Free surface imaged obliquely from below by CCD camera 100 (1008  1008 pixel) consecutive images acquired at 30 Hz over 3.3 s for x  25 cm Image exposure 5 = 4.3–11.2 ms, where  = /Uo Visualize free surface as interface between fluorescing (white) water and (black) air x y 10 cm z PLIF Technique

Liquid Prob. Distribution: 

Threshold individual images Grayscale > threshold  liquid; < threshold  air Average over 100 images  LPD; assemble composite LPD over half the flow by overlapping side, edge sections Probability of finding liquid inside 50% contour  50% Distance between contours measure of surface ripple LPD (part of side view) LPD (edge view) Liquid Prob. Distribution

Nozzle Geometry Effects: 

5 25 50 75 95% LPD Contours A B LPDs for nozzles A, B, C: Re = 130,000; x = 25 cm C Surface ripple similar for all 3 nozzles Surface ripple greatest at edge C has largest surface ripple B “best” nozzle 1 cm Nozzle Geometry Effects

Reynolds Number Effects: 

Re = 22,000 Side and edge fluctuations increase with Re LPDs for Nozzle B: x = 25 cm 5 25 50 75 95% LPD Contours Re = 53,000 Re = 97,000 Re = 130,000 1 cm Reynolds Number Effects

Mean Standard Deviation: 

A D Edge, side fluctuations increase with x z = 0.10 mm at x = 10 cm from nozzle exit z = 0.19 mm at x = 25 cm [scaled HYLIFE-II pocket: x  30 cm]  Max. surface ripple for HYLIFE-II 1.4 mm LPDs for Nozzle B: Re = 130,000 Mean Standard Deviation 5 25 50 75 95% LPD Contours

Flow Straightener: 

Perforated Plate (PP) Open area ratio 50% with staggered 4.8 mm dia. holes Honeycomb (HC) 3.2 mm dia. hexagonal cells Fine Screen (FS) Open area ratio 37.1% 0.33 mm dia. wires woven with open cell width of 0.51 mm 195 mm from FS to nozzle exit All elements stainless steel x y z PP HC FS Flow Straightener

Flow Straightener Blockage: 

Flow straightener blocked just upstream of fine mesh screen (element most likely to trap debris) “Blockage” = 1.5 cm  0.5 cm rectangle (blockage area = 2.5% of total screen area) Studied blockage at two different locations Centered along y, z [“center blockage”] On right edge centered along z [“edge blockage”] Flow Straightener Blockage

Blockage Effects: 

E G 2.5% area blockage 19.5 cm upstream of nozzle y z No Blockage Edge Blockage No blockage Center blockage 5 25 50 75 95% LPDs for Nozzle B: Re = 97,000; x = 25 cm Blockage Effects

Summary: Thick Liq. Prot.: 

Results at 2/3 the prototypical Re imply: Standard deviation of side free-surface geometry ~1.4 mm at bottom of lattice typical of HYLIFE-II At higher Re and x, free surface ripple on sheet side AND edge can “clip” driver beams Surface ripple appears relatively insensitive to small changes in nozzle geometry (circular-arc, vs. 5th order polynomial, contraction) Rounding nozzle corners ( elliptical nozzle) does not reduce surface ripple Blockage of flow straightener (due to debris, for example) will drastically increase surface ripple  filtration required Summary: Thick Liq. Prot.

What remains to be done?: 

All concepts: Chamber clearing Droplet formation/ejection High-speed film/Wet wall: Beam port designs compatible with film flow [GT] Surface wettability [GT] Surface curvature [GT] Porous wall/Wetted wall: How does heat transfer affect film stability? Thick liquid protection: Vacuum effects (We) on sheet breakup Oscillating sheets at high Re What remains to be done?