Fluidised Powder Rig Development

Work by:Ottone Caretta, Peter Loveridge and Chris Denham (RAL)

Tom Davies (Exeter University)Richard Woods (Gericke Ltd.)

With special thanks to EPSRC Engineering Instrument Poolwww.eip.rl.ac.uk

Presented by Ottone CarettaO.Caretta@rl.ac.uk

UKNF Meeting, OxfordJune 2010

Is there a ‘missing link’ target technology?

Some potential advantages of a flowing powder:Resistant to pulsed beam induced shock waves

Favourable heat transfer

Quasi-liquid

Few moving parts

Mature technology

Areas of concern can be tested off-line

Open jets

SOLIDS LIQUIDS

Monolithic Flowing powder Contained liquidsSegmented

Schematic layouts of flowing powder targets for neutrino facilities

Superbeam target - contained within pipe

Neutrino factory target - open jet configuration used in test rig on day 1

(for MERIT comparison)

(1) pressurised powder hopper, (2) discharge nozzle, (3) recirculating helium to form coaxial flow around jet, (4) proton beam entry window, (5) open jet interaction region, (6) receiver, (7) pion capture solenoid, (8) beam exit window, (9) powder exit for recirculation, (10) return line for powder to hopper, (11) driver gas line

The rig

• Powder– Rig contains 150 kg Tungsten– Particle size < 250 microns

• Total ~8,000 kg powder conveyed– 90 ejection cycles– Equivalent to 15 mins continuous

operation

• Batch mode– Test out individual handling processes

before moving to a continuous flow loop

1

2

3

4

1. Suction / Lift2. Load Hopper3. Pressurise Hopper4. Powder Ejection and Observation

Ottone Caretta, Oxford, Nov 09

Experiments: the fun part!

Turbulent flow ~3bar

Dune flow ~1.5bar

Pulsing flow ~1.5bar Coherent jet ~2bar

Ottone Caretta, Oxford, Nov 09

Areas of work

1. Measure density and variations of

2. Obtain as high an coherent density as possible

3. Match the suction and ejection rates currently 1:10 ratio (necessary for continuous operation)

4. Minimise wear of the system and of the powder

Pulsing flow ~1.5bar Coherent jet ~2bar

1. PIV

2. Improve powder flow path

3. Optimise the suction cycle

4. Eliminate design flaws from the rig

Coherent jet characterisation

Coherent Jet workout

– Tungsten powder <250 um– 2.0 bar ejection hopper pressure– Jet “droops” by ~30 mm over a 300 mm length– Each particle takes ~0.1 sec to traverse

viewport– Coherent flow with separation between the 2

phases– Constant pressure in hopper throughout

ejection– Small velocity gradient from top to bottom– Velocity constant over time– Cross section of the jet remains constant

as the jet flows away from the nozzle– Geometry of the jet remains reasonably

constant with timeLow pressure ejection schematic

Vjet = 3.7 m/s

Vair ~30 m/s

Still from video clip(2 bar ejection hopper pressure)Ottone Caretta, Oxford, Nov 09

Jet Density Calculation

From hopper load-cell data log:63 kg in 8 sec = 7.875 kg/sec

h ID

Nozzle ID = 21.45 mmJet height = 14.6 mmJet Area = 262 mm2

• Recall: Solid Tungsten density = 19,300 kg/m3• Powder density “at rest” ~ 50% solid

Density Calculation for 2 bar ejection

Jet area, A= 262 mm2

(from nozzle dimensions and video still measurements)

Powder bulk velocity, V = 3.7 m/s

(from particle tracking)

Vol flowrate = A.V = 0.000968 m3/s

Mass flowrate = 7.875 kg/s

(from loadcell)

Jet Density = Mass flowrate / Vol flowrate = 8139 kg/m3

Jet Density = 42% Solid tungsten density

Uncertainty is of the order ± 5% density

Ottone Caretta, Oxford, Nov 09

Open Source - Particle Image Velocimetry (PIV)

– http://www.openpiv.net/– Allows calculation of relative velocities

between pairs of subsequent images.– Velocities can be scaled in space and time

Ottone Caretta, Oxford, Nov 09

PIV - Highlighting the odd grains

– Original images

– Average images

– Negative image (subtracted the average image) this highlights the odd grains

Ottone Caretta, Oxford, Nov 09

PIV – velocity range selector

– Allows taking out fringes and mistaken points from the average calculation

Ottone Caretta, Oxford, Nov 09

PIV - example

Ottone Caretta, Oxford, Nov 09

PIV – vertical velocity profile in the jet

Ottone Caretta, Oxford, Nov 09

Variations in the flow rate – typical 2bar ejection

How much material does the beam meets?

Density?

Is the amount of material in the nozzle (or jet) constant?

Ottone Caretta, Oxford, Nov 09

Ottone Caretta, Oxford, Nov 09

Future experiments – prevent phase separation

The commercial dense phase conveyer is

less than ideal!

Sharp bend

Horizontal reducer

Unused lower air supply

Vertical reducer (near continuous)

Long radius bend

Future experiments – artificial/regular slug formation

Ottone Caretta, Oxford, Nov 09

Pulsed air injection

Separate, bunch and accelerate slugs

Future experiments – continuous recirculation (contained target)

Ottone Caretta, Oxford, Nov 09

Analytical study on lifting power requirements

Ottone Caretta, Oxford, Nov 09

0 2 4 6 8 10 12 14 160

1000

2000

3000

4000

5000

6000

7000

8000

lifting work vs suction capacity

Powder Mass flow rate [kg/s]

Th

eore

tica

l li

ftin

g p

ow

er [

W]

Powder lifting flow rate depends on a few variables:

– Powder entrainment in the air stream– Powder size distribution– Sphericity of the grains– Diameter of the suction line– Air to powder ratio– Density of the powder– Density of the gas– Temperature of the gas– Etc.!

The blower in the rig (18kW) has so far been able to lift at 1kg/s vs an

ejection rate of ~10kg/s

But there is hope!

Different density requirements for Superbeam and NuFact

Ottone Caretta, Oxford, Nov 09

High density tungsten for NuFact

Low density alumina for Superbeam?

Thank you!

Q & A?

Ottone Caretta, Oxford, Nov 09