X-band Accelerator Structures

R&D at SLAC

Juwen Wang

SLAC/LLNL Discussion

March 5, 2011

1. Introduction• Brief history• Achievements

2. Basics of X-Band Accelerator Structures• Design Principle and Description• Accelerator Structures Performance

3. Structure Assembly Technology• Mechanical QC and Microwave QC• Chemical cleaning• Accelerator parts joining (diffusion bonding, brazing

and welding)• Microwave tuning and characterization• Vacuum baking• Alignment

4. Discussion on Fabrication Technology and Locations

Outline

Contributors

SLAC: C. Adolphsen, N. Baboi, K. Bane, G. Bowden, D.L.

Burke, J. Cornuelle, S. Doebert, V. Dolgashev, A. Hasse, H.

Hoag, E. Jongewarrd, K. Jobe, R.M. Jones, R. Kirby, k. Ko, Z.

Li, G.A. Loew, J. Lewandowski, R.J. Loewen, D. McCormick,

R.H. Miller, C. Nantista, C.K. Ng, E. Paterson, C. Pearson, N.

Phinny, T. Raubenheimer, M. Ross, R.D. Ruth, S. Tantawi, K.

Thompson, J. Van Pelt, F. Wang, J. W. Wang, P.B. Wilson.

KEK: Y. Funahashi, Y. Higashi, T. Higo, N. Hitomi, H. Kudo, T.

Kume, H. Matsumoto, Y. Morozumi, K. Takata, T. Takatomi, N.

Toge, K. Ueno, Y. Watanabe.

FNAL: T. Arkan, H. Carter, D. Finley, I. Gonin, T.

Khabiboulline, S. Mishra, G. Romanov, N. Solyak.

LLNL: J. Klingmann, K. Van Bibber.

1. Introduction• Brief history

• Achievements

Achievements of X-Band

Structures R&D at SLAC

• Motivation:Main Linac for the Future Linear Colliders (NLC, GLC and CLIC)

• Brief History:1. 1988 – 2004 X-Band accelerator structures R&D for the NLC/GLC

in collaboration with KEK, FNAL and LLNL.

Designed, fabricated and tested 50 X-Band accelerator structure sections.

(Among them, 8 were made with KEK collaboration and 12 were fabricated by

FNAL).

2. 2007 – Present X-Band accelerator structures R&D for the CLIC

main linac in collaboration with CERN and KEK. Participated the design, fabrication and testing of more than 12 X-Band

accelerator structure sections.

3. Ongoing Support Work for LLNL Project of the Compton

Scattering Light Source MEGa-Ray.

Evolution of Structures

New typesof couplers

Optimized cell shape

Contribution to the Accelerator Technology

through NLC/GLC X-Band Structures R&D

• Theoretical analysis for full understanding of HOM suppression in RF

accelerator structures.

• Damped and Detuned Structures can be applied to any low emittance,

high beam loading accelerators.

• Simulation methods for beam-structure interaction: structure wakefield,

emittance growth and analysis of structure alignment and dimension

tolerances.

• Optimization of accelerator parameters for highest RF efficiency and

dimension determination with sub-micron precision.

• Manifold damping gives structure position monitor with micron transverse

sensitivity and frequency multiplexed longitudinal resolution of the order

of several cells.

• Fabrication technologies for normal conducting accelerator structures

such as precision machining, diffusion bonding and long structure

alignment.

• Extensive studies for high gradient RF operation to meet the NLC

requirement rate at 65 MV/m: new types of couplers, Procedure for

structure treatments.

2. Basics of X-Band

Accelerator Structures

• Design Principle and Description

• Accelerator Structures Performance

Requirements of Accelerator

Structures for Linear Colliders

• High Accelerating Gradient to

Optimize Length and Cost.

• Control of Short and Long-

Range Wakefields to Ensure the

Preservation of Low Emittance for

Multi-Bunch Beams.

Transverse wakefields - II

Computed transverse δ-function wake potential per cell for S-Band SLAC structure.

Solid line: Total wakeDashed line: 495 modesDot-dashed line: lowest frequency dipole

mode (λ=7 cm)

Multi-bunch Beam Breakup due to the long range transverse wakefields.

Single Bunch Emittance Growth (Head-Tail Instability) due to the short range transverse wakefields

Early Studies on Two Types

of Heavily Damped Structures

Example of structure with radial slots in iris.Example of structure with circumferential-slot coupling: crossed-waveguidestructure, with two half-cells and one full cell.

Test Cells for Damped Structures

in 1980s at SLAC

CERN CLIC

Waveguide Damped (WDS) Structure

• Minimize

E-field

• Minimize

H-field

• Provide

good HOM

damping

• Provide

good

vacuum

pumping

Long Range Dipole Mode Suppression

- Idea of Detuning of Dipole Modes

1

dnk

df

Cells for a Detuned Structure have profiles with Gaussian dimensional distribution.

Dipole mode distributionfor Detuned Structure

In the time domain, the excited wakefield by the cells with Gaussian distribution dipole frequencies has Gaussian amplitude profile.

•Treat each cell as periodic.

•Calculate several sample

cells to obtain dispersion

curves for studying

synchronous kick factor

and avoided crossing

(coupling).

•Fit dispersion curves

of sample cells to

obtain cell parameters

for equivalent circuits.

• Interpolate to obtain

parameters of all cells

• Solve coupled circuit

system.

• Integrate spectrum for

wake in order to provide all

important design

information to optimize

cell-manifold parameters

12

14

16

18

20

22

0 30 60 90 120 150 180

Phase (deg)

F1 (

GH

z)

avoided

crossing

Long-Range Wakefield Calculation

Precision Fabrication

for Accelerator Discs

Profile tolerance 1 μm and Surface finishing better than 50 nm

Microwave QC for Single Disk

Stack Microwave QC Setup

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

0 50 100 150 200

Single-disk RF-QCdel_sf00del_sf0pidel_sf1pidel_sf20

Fre

qu

en

cy D

evia

tio

n [

MH

z]

Disk number

Single-Crystal Diamond Turning Polycrystalline Diamond Turning

Super Precision Machining with Single

Diamond Cutter – Tuning Not Needed

-3

-2

-1

0

1

2

3

0 50 100 150 200

Accelerating mode frequency

2b

off

set

[mic

ron

]

Fre

qu

en

cy [

MH

z]

Inte

grate

d P

hase

Sli

p [

deg

ree]

Disk number

2b_offset

Integrated Phase SlipMeasured Frequency

Regular Precision Machining with

Polycrystalline Diamond Cutter – Tuning Needed

Microwave QC of Fundamental Modes for H60VG4SL17A/B Regular CupsTemperature and humidity corrected

Microwave QC of Dipole Modes for H60VG4SL17A/B Regular CupsTemperature and humidity corrected

Regular Precision Machining with Polycrystalline

Diamond Cutter – Tuning Needed (Continued)

Port for Terminatingand Extracting Dipole Mode Power

High PowerRF Coupler

Prototype Accelerator Structure

for the NLC/GLC Main Linac

Cutoff view of a structure end

A 60 cm structure with most of final design features

Damped Detuned Structures for

the NLC/GLC

DDS1 (Round Damped Detuned) 2π/3 Mode TW StructureSingle diamond turning discs without tuning; Micron level cell-to-cell alignment.

High Gradient Test Structures

One of four T-type Structures --

T53VG3, 60-Cell 2π/3 Mode TWSW20PIL

15-Cell π Mode SW

One of more than

10 High Phase

Advance 5π/6

Mode TW

Structures,

H60VG3S18 with

HOM Slots and

Manifolds.For the LLNL Campton

Scattering Light Source.

Theoretical and Experimental Proof of

Transverse Wakefield Suppression

Comparison of the measurement for a pair of dipole Interleaved

60 cm Damped Detuned X-Band Structures with error bars (red)

and calculated wakefield (black) – Data from early 2005.

RFRF

Beam’s eye view of input coupler.

SEM picture of input matching iris.

Pulse heating was in excess of 100°C.

Performances of some structures

were found to be limited by pulse

heating of coupler matching irises.

Distribution of Breakdowns

(70 MV/m, 400 ns, 10 hr run)

T53VG3

Input coupler Output coupler58 Cells

Rate in cells

.1/hr

Edge Damage on Cavity Side of

Coupling Iris due to RF Pulse Heating

RF Pulse Heating causes:

• Surface Roughening and Cracks

• Local Surface Melting

Field Distribution in Coupler

Region and RF Pulse Heating

Surface temperature

distribution in the region of

coupler iris for 400 ns pulses,

48 MW.

The maximum temperature

increase was 127º C.

Surface resistivity

Thermal conductivity

Specific heatMagnetic field Pulse width

Temperature increase

due to RF pulse heatingc

RTHT s

pt

2

Improved Coupler Design

By proper choice of

matching cell b dimension,

matching cell length can be

made equal to standard cell

length.

|Es|max= ~34 MV/m @ 48 MW

|Hs|max= ~98.4 kA/m @ 48 MW Pulse Heating ~ 3° C

NLC Prototype Structures Can Stably Operate at 65

MV/m to Meet the Required RF Breakdown Rate

Average breakdown rates for a series of NLC test structures as a function of accelerating

gradient after 500 hours (upper line) and 1500 hours (lower line) of RF processing.

Breakdown Rate Dependence on Pulse

Length for Various NLC Structures.

Some of KEK/SLAC Made Accelerator Structures

for Testing CLIC Main Linac Design

T18_VG2.4_DSC with SLAC Flanges TD18_VG2.4_DISC with SLAC Flanges

TD18_VG2.4_DISC with KEK Flanges

T28_VG2.9 (T26) with SLAC Flanges

T18_VG2.4_DISC Structure Test

Cumulated

Phase Change120°

Field

Amplitude

5.1~__ inaccoutacc EE

Microwave Tuning and test High power test set-up

RF BKD Rate Gradient Dependence for 230ns Pulse at Different

Conditioning Time

After 250hrs RF

Condition

After 500hrs RF

Condition

After 900hrs RF

Condition

RF BKD Rate Pulse Width Dependence at Different

Conditioning Time

G=108MV/m

G=108MV/m

G=110MV/mAfter 1200hrs RF

Condition

This performance maybe good enough for 100MV/m structure for a warm collider, however, it does

not yet contain all necessary features such as wake field damping. Future traveling wave structure

designs will also have better efficiencies

CLIC Prototype Structures Can Stably Operate at 100

MV/m to Meet the Required RF Breakdown Rate

3. Structure Fabrication

Technology

• Mechanical QC and Microwave QC

• Chemical cleaning

• Accelerator parts joining (diffusion bonding,

brazing and welding)

• Microwave tuning and characterization

• Vacuum baking

• Alignment

Lathe with Twin Spindles and Twin Turrets

Profile tolerance 5 μm and Surface finishing 300 - 400 nm

ZYGO Surface Flatness Measurement

for Typical Cups of T18_VG2.4_DISC Structures

Both sides show less than 1 micron concaved

17D-A 17D-C

14D-C16D-A

Stacking for Body Diffusion

Bonding of a CLIC Structure

Diffusion Bonding of T18_vg2.4_DISC

Pressure: 40 PSI

Holding for 1 hour at 1020º C

Brazing of QUAD with Water Flange

Au/Cu Alloy: 25/75

Brazing temperature: 1041-1045º C

First Assembly Brazing of T18_vg2.4_DISC

Body / Two Coupler Assemblies / Cooling/One Beam Pipe / Tuning Studs

Au/Cu Alloy: 35/65Brazing temperature: 1021-1025º C

Final Brazing of T18_vg2.4_DISC

Au/Cu Alloy: 50/50Brazing temperature: 979-983º C Adding One Beam Pipe

Flange Welding for a Accelerator Structure

Microwave Tuning and Characterization

Tuning and

Structure Characterization

Example of Phases and Amplitudes

along the Axis of a 77-Cell TW Accelerator

Phases and amplitudes

plotted in a complex

plane (5π/6 mode

structure, 2x150º=300º

per cell for reflection)

Electrical field amplitudes along the structure.

There are small amplitude and phase

modulation due to slightly imperfections of the

couplers, which almost no impact to the power

efficiency and beam acceleration.

Typical modulation due to uncompleted tuning of output coupler

Reflection and Transmission

Reflected S11 from input coupler as a function of frequencies.

Transmission S12 from input port to output port as a function of frequencies.

Wiggles due to the beating of reflections from slightly mismatched input/output couplers.

Transmitted power (left to the load) =

fTf

2

1

2

12S

2

21Sf

Vacuum Baking of Two Structures

650° C

10 days

Alignment Measurement

Using CMM Machine

4. Fabrication Technology and

Locations

Fabrication Locations

1. National Laboratories for the X-Band Structures:• SLAC

• KEK

• LLNL

2. Private Vendors for the X-Band Structures:• US

Robertson Precision, Inc. (California)

LeVezzi Precision, Inc. (Illinois)

• Japan

IHI

Morikawa co.

• Europe

VDL Enabling Technologies Group (Netherland)

Manufacturability

• Case of Small Amount Production (less

than few hundreds)

• Case of Mass Production (10k for future

X-Band compact FEL or even 1.8 millions

precisely machined parts for Linear Collider,

which was studied extensively in late 1990s)

Design of Manufacturability (DFM) Studies

with Huge Cost Reduction