Measurement of the SEISM - CERN · • The SEISM magnetic structure was built and set up on a test bench at LNCMI Grenoble • Continuous magnetic field has been produced on-site

Measurement of the SEISM (Sixty GHz ECR Ion Source using Megawatt Magnets)

magnetic field map

Mélanie MARIE-JEANNE

J. Jacob, T. Lamy, L. Latrasse from LPSC Grenoble

F. Debray, J. Matera, R. Pfister, C. Trophime from LNCMI Grenoble

August 28, 2010 Mélanie MARIE-JEANNE ECRIS’10 - Grenoble, France 2/18

Outline

• Why a 60GHz prototype ?

• Results of the design study

• Technical challenges of fabrication

• Magnetic field measurements

• What next ?

Why a 60GHz prototype ? (1)

For the beta-beam project

Multi-ionizing and bunching radioactive gases diffusing from the target

Neutrino beams

Ion production

6He, 18Ne storage and β- decay

RADIATION RESISTANTHIGH CURRENTS

Close to target 1013 at/s yields

EFFICIENT IONIZATION SHORT CONFINEMENT TIME

Short half-lives from 807ms (6He) to 1.67s (18Ne)

PULSED MODESHORT BUNCHES

10Hz to 25Hz duty cycle50-100μs pulses



Experimental Ar4+ preglow pulse from PHOENIX V2 at 28GHz

Izotov I.V., Lamy T., Latrasse L., Sidorov

A.V., Skalyga V.A. et al IEEE Transactions

on Plasma Science 36/4 (2008) 1494-1501

PREGLOW MODE

Polyhelix technology developed at LNCMI Grenoble

New ECR magnetic structures

using resistive polyhelix coils

60GHz high density plasma

RADIATION RESISTANT

Compact (combined to target)

Why a 60GHz prototype ? (2)

What kind of prototype

60GHz ECRIS

SHORT BUNCHES

HIGH CURRENTS

Preglow mode

PULSED MODE

SHORT CONFINEMENT TIME

Origins of 60GHz project: presentation by P. Sortais in Moriond - Les Arcs, March 17-22, 2003


Results of the design study (1)

L. Latrasse et al., SEISM: A 60 GHz cusp electron cyclotron resonance ion source, Rev. Sci. Instrum. 81, 02A324, 2010

Compact CUSP magnetic structure

4 T

3 T6 T

4 T

Injection Extraction

2.1 T

Field line

60

mm

Peak to peak ~100 mm

ECR zone

4 T

3 T6 T

4 T


2.1 T

Field line

60

mm


ECR zone

injection

extraction

Plasma chamber

Magnetic field above expectations

6.9 T

4.9 T

4.9 T

3.5 T

SEISM prototype

H1 aluminum prototype

0

50

100

150

200

250

300

350

400

-150 -100 -50 0 50 100 150 200

z (mm)

B s

ur

l'a

xe

(G

au

ss

) Expérience

Calcul

H4

H4

H2

H2

H1

H1

H3

H3

2 current leads per coil

bossage to avoid rotation



SEISM prototype

30kA current on each set of coils

Fake plasma chamber

Stainless steel rod



SEISM prototype

28l/s waterflow in each tank

Water inlet

Water outlets

Water inWater out


Valve

Flowmeter

Deionized

water inlet

Flexible pipes

Deionized

water outlet

Flexible power cables

Magnetic measurement

systems

Flexible power

cables

On-site connection to a running magnet for half-magnetic field test

Electric parameters

I = 15000 A

P = 0.75 MW (one coil set)

Cooling parameters

Q = 22 l/s

Pin = 20 bars

Max coil temperature

Tmean = 50 °C

TMax. loc. 70 °C.


Voltagemeasurement

Jack

Hall probe

Visit LNCMI on Thursday !

•Electrically in series

•Hydraulics in parallel

Test bench at LNCMI Grenoble

13T – 10MW magnet2x13000A applied


Technical challenges (1)Insulating between the helices windings

• Narrow insulators (2mm wide) to minimize local heating• Height calibration (0.32mm height)• 24 sectors on inner coil, 32 sectors on outer coil to avoid

contact between the windingsTemperature calculations

Winding displacement calculations

insulators Height calibrating pieces

Heating tests on prepreg

Damaged resin

« prepreg » for pre-impregnated composite fibres

G11 dry woven glass fabric impregnated with epoxy resinSpecified maximum continuous operating temperature: 165°C

Specified breakdown voltage condition: 35kV/mm

Out of stock ! Tests with frozen out-of-date prepreg20MPa at room temperature

Resin damaged for local temperatures around 300°C

0

5

10

15

20

0 2000 4000 6000 8000 10000 12000 14000

Wate

r fl

ow

(l/s)

Current (A)

Water flow vs applied current

Injection tank

Extraction tank

Injection tank with porous disks

Extraction tank with porous disks


Technical challenges (2)Hydraulic circulation from the inner to the outer coil

• Tests with fake aluminum helices

water circulation up to 18bars – 18l/s in each tank

cavitation noises, small damage marks on aluminum

• Tests with copper helices and measurements up to 7000A

10bar - 12l/s in each tank _ water speed up to 14m/s in the radial helices slit

cavitation noises

• Porous discs to slow down the flow in SEISM, but not in the LNCMI magnet running in parallel

no more cavitation, but a filter damaged after 24 hours of run28l/s waterflow in each

tank

∆P ~ 0.5 bar in the helices

19.5 bar on the porous disc


Voltage measurement to monitor the coils resistance / temperature:

Technical challenges (3)Temperature monitoring

insulators

Snapshot of the outer coil insulators after 40 hours of run up to 7000AInserting a camera to have a look:

++ insulators still aligned !

-- color indicates local temperature is higher than expected

Voltagemeasurement

Mean temperature evolution

0

5

10

15

20

25

30

35

40

2000 4000 6000 8000 10000 12000 14000 16000

Current (A)

Delt

a T

injection

extraction

injection with poral discs

extraction with poral discs


Jack

Hall probe

Controllers

• 300mm-course jacks with a step-by-step motorequipped with a probe holder

Magnetic field measurements (1)Measurement setup

• Measurement on 3 horizontal axes along z and on one radial axis

4 T

3 T6 T

4 T


2.1 T

Field line

60

mm


ECR zone

4 T

3 T6 T

4 T


2.1 T

Field line

60

mm


ECR zone

30mm15mm0mm

Axial field in LNCMI M5 magnet

0

1

2

3

4

5

6

7

0 2000 4000 6000 8000

Current intensity (A)

Axia

l fie

ld B

z (

T)

Bell615 Hall probe

M5 magnet calibration

curve

Axial field in LNCMI M5 magnet

0

0,5

1

1,5

2

2,5

3

3,5

4

0 1000 2000 3000 4000

Current intensity (A)

Axia

l fie

ld B

z (

T)

Bell5080 Hall probe

M5 magnet calibration

curve

• Two gaussmeters equipped with single-axis axial and radial Hall probes

• LabView interface to move jacks and acquire data


As expected:Radial field measurement on axis 15mm at 3500A

-0,2

-0,1

0

0,1

0,2

0,3

-250 -200 -150 -100 -50 0 50 100 150

z (mm)

Br x

-com

pon

en

t (T)

probe HORIZONTAL at angle 90deg

probe VERTICAL at angle 0deg

Magnetic field measurements (2)Preliminary results vs simulations

Axisymmetric field

Increase with distance to central axis

Scaling with increase of the intensity

Unexpected: Shift in maxima positions

Lower amplitude on extraction side

-1,5

-1

-0,5

0

0,5

1

1,5

2

2,5

3

-200 -100 0 100 200

Bz (

T)

z (mm)

Axial magnetic field on central axis

1500A

3500A

5250A

7000A

8750A

10500A

-3

-2

-1

0

1

2

3

4

-200 -100 0 100 200

Bz (

T)

z (mm)

Axial magnetic field on central axis

1500A

3500A

5250A

7000A

8750A

10500A

14000A (extrapolated)

14000A (calculations)


Magnetic field measurements (2)Preliminary results vs simulations

Axial field on central axis at 7000A

-1,5

-1

-0,5

0

0,5

1

1,5

2

2,5

-200 -150 -100 -50 0 50 100 150 200

z (mm)

Bz (

T)

simulation-injection

simulation-extraction

simulated sum


-1,5

-1

-0,5

0

0,5

1

1,5

2

2,5

-200 -150 -100 -50 0 50 100 150 200

z (mm)

Bz (

T)



measurement-injection

measurement-extraction


-1,5

-1

-0,5

0

0,5

1

1,5

2

2,5

-200 -150 -100 -50 0 50 100 150 200

z (mm)

Bz (

T)



measurement-injection

measurement-extraction

simulated sum

sum of measurements


• Mechanical error (tank dimensions, helix positioning)Injection and extraction coil sets are too closeSum of the amplitudes is modified Not likely to cause 10mm difference, tanks dimensions were checked and within tolerances

Magnetic field measurements (3)Possible explanations

• Misplaced magnetic center because helix shape is wrongElectric discharge machining with a 0.25mm wire

For example real split is 0.37mm instead of expected 0.32mm

Magnetic centers can be checked individually for each helix after dismounting

• Calculation errorConsidered heat transfer is wrong

Copper resistance is higher, current density is lower, and resulting magnetic field is lower

Comparative simulations should be performed


Conclusion

• The SEISM magnetic structure was built and set up on a test bench at LNCMI Grenoble

• Continuous magnetic field has been produced on-site for 70h up to now

• Results show an axi-symmetric field map with a lower amplitude and closer maxima than expected from the simulations

• Possible explanations involve mechanical errors in the fabrication of the polyhelix coils or the water tanks, and are still under investigation


What next ?

• Right now: Damaged poral disc to be replaced Gaussmeter with triple-axis Hall probe to be bought Magnetic field measurements up to 14000A (2 weeks run to be scheduled in automn 2010)

• Next year: Plasma chamber design permanent room at LNCMI is under funding request for first tests at 28 GHz

• In a near future: Preparation to raise the current to full intensity (30kA)

Adding direct voltage reading on each individual helix Adding temperature reading on local non-cooled parts (insulators) Insulators replacement ? What kind ?

60 GHz gyrotron is currently under construction at IAP-NN (ISTC contract)


Acknowledgements

To numerous people from the SERM-LPSC workshop

To the LNCMI Magnet Group

Documents

Measurement of the SEISM - CERN · • The SEISM magnetic structure was built and set up on a test bench at LNCMI Grenoble • Continuous magnetic field has been produced on-site