IHEP High Power Input Coupler Activity Report

First Mini-workshop on IHEP 1.3 GHz Superconducting RF project

June 10th, 2009, Beijing

IHEP High Power Input Coupler Group

Huang Tong-ming ( 黄彤明 )

Outline• The construction of the high power input coupler

for BEPCII 500MHz superconducting cavity;• The preliminary consideration about the input

coupler for IHEP 1.3GHz superconducting RF project;

• Some questions to be consulted.

Part I:

The Construction of the High Power Input Coupler for BEPCII 500MHz Superconducting Cavity;

Design

• Considering the short time available for the BEPCII construction, the high power input coupler is modified based on the KEKB 508MHz coupler.

• In the following, some simulation works will be listed.

个

1. doorknob cover;2. doorknob;3. DC-bias component;4. ceramic;5. outer conductor;6. inner conductor;7. water-cooling pipes;8, window;9. monitor ports for vacuum 、 Arc and electron ;10. BEPCII 500MHz superconducting cavity.

Window position

The ceramic window is usually located at the minimum E-filed of the standing wave mode.

Ceramic window

Tristan-type window optimization

-10 0 10 20 30 40 50 60 700

100000

200000

300000

400000

500000

Radial electric field of the ceramic window

Ele

ctri

c fie

ld (

V/m

)

Radial distance (mm)

dc=2 mm dc=4 mm dc= 6 mm

The area near Choke tip

Conclusion: • The E-field near the choke tip

increases obviously, so the distance from the choke tip to the ceramic surface should be carefully selected.

• The H-field gratitude along the ceramic radius is high, especially the H-field on the inner conductor is quite high, so the dimension of the ceramic window inner conductor is not allowed to be too small.

490 492 494 496 498 500 502 504 506 508 510

0.990

0.992

0.994

0.996

0.998

1.000

1.002

rw=171mm

Mag

(S21

)

Frequency (MHz)

rw=176mm

rw=166mm

rw f0,

f0=500MHz

(b)

488 490 492 494 496 498 500 502 504 506 508 510 5120.9840.9850.9860.9870.9880.9890.9900.9910.9920.9930.9940.9950.9960.9970.9980.9991.000

h=99mm

h=94.5mm

Mag

(S

21)

Frequency (MHz)

h ,f0

h=90mm

(c)

f0=500MHz

488 490 492 494 496 498 500 502 504 506 508 510 512

0.994

0.995

0.996

0.997

0.998

0.999

1.000

r_doorknob=15mm

r_doorknob=20mm

Mag

(S21

)

Frequency(MHz)

r_doorknob ,f0

r_doorknob=10mm

(d)f0=500MHz

Impedance matching optimization

Conclusion:

• ‘rw’, ‘h’, ‘r_doorkonb’ have a great impact to the central frequency;

• the parameter “h” is changed to 94.5mm to shift the central frequency to 500MHz.

1.000000E+04

1.000000E+05

1.000000E+06

1.000000E+07

-33 -28 -23 -18 -13 -7. 7 -2. 7 2. 3 7. 3 12.3 17.3 22.6 27.6 29.6

antenna penetrati on(mm)

10*l

ogQe

xt

measuredsi mul ati on

0. 000E+00

5. 000E+05

1. 000E+06

1. 500E+06

2. 000E+06

2. 500E+06

3. 000E+06

-32. 7 -27. 7 -22. 7 -17. 7 -12. 7 -7. 7 -2. 7 2. 3 7. 3 12. 3 17. 3 22. 6 27. 6 29. 6

antenna penetrati on(mm)

Qext

radi us_10radi us_15

0. 000E+00

5. 000E+05

1. 000E+06

1. 500E+06

2. 000E+06

2. 500E+06

3. 000E+06

3. 500E+06

4. 000E+06

4. 500E+06

antenna penetrati on(mm)

Qext Di a_SBP 220

Di a_SBP 200

0. 000E+00

1. 000E+06

2. 000E+06

3. 000E+06

4. 000E+06

5. 000E+06

6. 000E+06

7. 000E+06

-33 -28 -23 -18 -13 -7. 7 -2. 7 2. 3 7. 3 12. 3 17. 3 22. 6 27. 6 29. 6

antenna penetrati on(mm)

Qext D_215. 25

D_225. 25

Qe calculationConclusion:

The results provide theoretical instruction for coupler input port design and the positioning of the antenna penetration depth.

Multipacting simulation

Electron trajectory of the window :Tri-plots about the counter function, final impact energy of the window:

Conclusion:

By simulation, a one order MP appears on the vacuum side of the window at mixed wave mode.

Thermal simulation

Pin（kW）

Dynamic loss W）

50 100 150 200 250 300

Window 26 52 77 103 129 155

Inner conductor 26 51 77 102 128 153

Outer conductor 11 22 33 45 56 67

Temperature distribution

Thermal Stress distribution

Calculated Dynamic losses

50 100 150 200 250 3000

20

40

60

80

100

120

140

160

RF heat of the coupler for BEPCII SCC

Y =0.08+0.26526 X

Y =0.064+0.22209 X

Y =0.004+0.51043 X

RF

hea

t(W

)

Pin(kW)

inner conductor outer conductor ceramic

Fabrication

Clean

Ceramic window frame

machiningClean

Ceramic-copper brazing

Ceramic TiN coating

The whole window brazing

Stainless steel parts machining

CleanCopper or

nickel plating

OFHC window parts machining

Copper-copper brazing

Window subassemblies

leak check

Window subassemblies

leak check

Steel-copper brazing

Window subassemblies

leak check

Stainless steel water cooling pipes

machiningClean TIG

Subassemblies leak check

OFHC antenna parts machining

Clean

Antenna and window EBW

Final assembly leak checkand storage

The whole antenna EBW

The fabrication flows of the inner conductor and the window:

Copper plating and TiN coating

Copper plating A good sample (left) and a bad sample (right) of copper plating: bubbles on the bad sample due to too high current

TiN coating

A plate shape Ti target acted as anode and the outer conductor of the window as the cathode. The desired film thickness is 8 nm.

Copper- copper brazing

AuCu(65%) at 1020

The whole window brazingAuCu(20%) at 889 , AgCu(28%) at 779

Brazing and welding

99.5% Al2O3ceramic- copper brazing

Ag at 962

Steel- copper brazing

AuCu(65%) at 1020

TIG (Tungsten-arc inert-gas welding)

EBW (electron beam welding)

The ceramic shield before EBW (left) and after EBW (right)

Brazing and welding, cont.

Leakage checking and surface cleaning

Subassemblies leak check Final assembly leak check

Assembly is processed in a

class 100 clean room OFHC parts after cleaning

Test ResultsThe two key components (window and inner conductor) wer

e fabricated in China and then received high power test in KEK. Professor Takaaki Furuya has given us great help about the test. The maximum power reached is 270kW in CW.

Test result (Cont.) No electron and outgasing were found at CW of 20

0kW after conditioningElectron current state: Keeping power test at 100kW ,150kW and

200kW

Part II:

The preliminary consideration about the

input coupler for IHEP 1.3GHz

superconducting RF project

IHEP 1.3GHz SCC couplerTTF III @DESY ERL @Cornell

STF baseline 型 @KEK电容型 @KEK

Through carefully investigating and considering our fabrication experiences, we choose the STF baseline type coupler as IHEP 1.3GHz coupler design prototype in the first research phase.

超导腔和调谐器

高功率输入耦合器

低温恒温器

真空系统等

功率源系统

低电平控制系统

低温系统

Frequency 1.3 GHz

Max Power

（for ILC）

Test：1MW，Pulsed

Oper：300 kW，Pulsed

Pulse length & rep.rate

（for ILC）

Test：1.5msec，5Hz

Oper：1.5msec，5Hz

eQ for ILC 6102

What have done• Preliminary RF structures have been

calculated.

• Coupling factor calculation has been finished decide the coupler port position

Transition taper

Cold part Warm part doorknob

Doorknob RF Structures

85

784

39

R 14R 10

R 3

40

45.2R 1.6 R 2.0

82.5

5

110

104

7.0

4

4

R 2.0

23

3

R7.0

19

Two parameters belong to the doorknob( highlighted with yellow) have been modified to get the optimum power transmission performance.

104

11645.2

44 6.6

50

50

303

35.2

5010

2011

3Warm Part RF Structures

R 2.01.0

3.0

R 1.5

6.6

R 52 R 22.6

R 15R 58

25 28

R 22

10

R 1.5

R 3.01

30Inner conductorOuter conductor

R 1.25

6.0

2.5×

11

2.5

R 1.25

2.0×

9

R 1

R 1

3.0R 40.5

9.0

R 17.6

5.0

Inner conductorOuter conductor

Bellow

Warm window

Try to reduce the number of bellows from 4 to 2: only one bellow on each conductors?

Cold Part RF Structures

81

35.2

5035

5050

407

60

26

40

17.4

R 40.5 R 17.6

R 17.3

10

20

3

6.2

R 1.25

R 21

R 1.5

R 2.3 1

R 11R 46

25

30

Inner conductorOuter conductor

Cold window

In order to match with the 40mm coupler input port, a tapper was used to transit the coaxial diameter from 60mm to 40mm.

E-field and H-field distribution

1MW @TW

Try to reduce the E-field on ceramic surface, especially air side.

1MW @TW

RF Performance

1.1 1.2 1.3 1.4 1.50.0

0.2

0.4

0.6

0.8

1.0

Ma

g(S

21

)

Frequency (GHz)

Before optimize After optimizeMesh check

Power transmission performance

0 10 20 30 40 50

200000

400000

600000

800000

1000000

1200000

1400000 Radial E-field distritubion on ceramic air side

Ab

s(M

ag

_E

)

Radial distance (mm)

Mag(S21)=0.9979@1.3GHz

Maximum Mag_E on ceramic air side is 1.28e6 V/m, @ 1MW, TW, 115deg

Qe calculation

• This work discusses the positioning of the high power input coupler for IHEP 1.3GHz 9-cell SCC. It examines the method used to reduce the model to enable faster solutions whilst still maintaining accurate results.

In order to faster solving, the model was reduced to 4.5-cell since Qe scales with the number of cells. The field flatness in every cell should be assured before the coupling analysis. Two parameters related with Qe were studied: 1) the coupler input port position; 2) the antenna penetration depth

Model of 4.5-cell and simplified coupler

field magnitude along the on axis curve

E-field distribution on Z-Plane

D

HParameters related with Qe

Qe calculation Plotted results

• To obtain the optimum 2×106 ， we can:– Chose the distance from

end cell to coupler port center D=40mm, adjust the antenna penetration near 4mm;

– Chose the distance from end cell to coupler port center D=45mm, adjust the antenna penetration near 7mm;

– Chose the distance from end cell to coupler port center D=50mm, adjust the antenna penetration near 10mm;

0. 0000E+00

2. 0000E+06

4. 0000E+06

6. 0000E+06

8. 0000E+06

1. 0000E+07

1. 2000E+07

1. 4000E+07

0 2 4 6 8 10

1 3 5 7 9

beam tube = 80mm input port = 40 mm Distance from end-cell = ‘D’The radius of antenna = 3mm Coaxial line = 50

Part III:

Some questions to be consulted

one MP resonance conditioned satisfied at operating power on flat wall

If the crest is wide enough (~2-3 mm), it may be able to support a MP buildup, or MP may jump across

Based on scaling, for a given bellows undulation or step, more MP resonances may be supported simultaneously, potentially increasing the conditioning area and gas load

Question 1: Have you done MP simulation? 2D or 3D simulation? Tools?Did you find MP occurs inside the coupler, especially on the bellow area during high power test? How to avoid MP?

“Coupler component Test stand Activities at SLAC/ LLNL”, TTC meeting @ KEK, WG1-input coupler, 2006

Bellows or Steps May Lead to More Pervasive Multipacting

Question2:

What’s the fabrication difficulties?

• Copper Plating on Bellows?

• Vacuum sealing of bellows?

• Other?

Question3:About ceramic:

• KEK coupler choose 95% Al2O3, why don’t you choose high purity ceramic?

• Is the thermal shock tests of the ceramic necessary? Put into liquid nitrogen repeatedly: repeated times?

Thermal Shock Test

Question4: Does the dimension shrinking from room temperature to cryo-temperature impact the power transmission performance?

Why is the central frequency a little below 1.3GHz? Does it have relationship with dimension shrinking?

My guess: Right?

The RF structures simulated

are at room temperature and after cooling down, the central frequency will shift to 1.3GHz.

5K cooling80K cooling

Beam pipe

Warm window

Doorknob conversion

Cold window

Vacuum port

Independent Vacuum? Why?

Question5:

Why does KEK STF coupler make a independent Vacuum in the inner conductor, since it results in DC bias impossible?

Questions 6 and 7:What kind of sealing gasket of flanges used between warm and cold parts, doorknob and warm parts connecting?What kind of method used in bellow connecting? Brazing or welding?

What’s the sealing gasket of the flange?

What’s the method used in bellow connecting?

Question8: Except copper plating, does the coaxial lines received other plating, e.g. ion plating? Plating method? Thickness?

One of trial to reduce electron multipactoring amplitude, Ti ion plate with ~2-µm applied (gold color) to the inner and outer conductors for the coaxial transmission line.

TTC at KEK

Question 9: When does TiN been coated on ceramic vacuum side? Before whole window assembled or after? What’s method used in TiN coating?

A TiN coating

B TiN coating

A or B?

Question10:During brazing and welding, is there any metal vapor coming from? If there is, how to protect the ceramic?

e-beam

ceramic with copper collars

shielding against metal vapor

for support and removal

protection mask for ceramic and RF surface, TTF-III coupler

ERL mini-workshop at Beijing University, 7th – 8th November 2005

Question 11: Is it necessary to use four bellows? Why not just use two bellows (it’s similar with TTF-III coupler)?

2 bellows4 bellows

Question 12:

• what’s the stainless steel type? 316L or 316LN?