DEVELOPING AN UNDERSTANDING OF

BREAKDOWN IN NF CU CAVITIES

Arash ZarrebiniUKNF Meeting– 8th January 2010

U.K Cavity Development Consortium

OLD BUT ATTRACTIVEThe most common problem encountered in both

Normal and Superconducting accelerating structures is:

RF breakdown – W. D. Kilpatrick (1953)

A large number of mechanisms can initiate breakdown. However, this occurs Randomly and Rapidly

It is believed surface impurities and defects are dominant cause of breakdown (must be verified)

No matter what mechanisms are involved, the end results are similar:

Fracture/Field evaporation High local Ohmic heating Hence, the loss of operational efficiency

RF BREAKDOWN

J. Norem, 2003, 2006 , Jens Knobloch, 1997

Breakdown is initiated locally while its effects are global

MuCool Button Test

Much of the effort has gone towards evaluating various material and coatings

MTA Testing Area

805 MHz Cavity

A. Hassanein , 2006

Button Test Results: 2007 – 2008

– LBNL TiN_Cu2LBNL TiN_Cu2

D. Huang – MUTAC 08

No Button 40 MV/m no

field 16 MV/m @ 2.8

T

Performance is considerably improved by usingstronger material and better coatings

A number of questions exist: o Reliability of Existing Results o Reproducibility

Experiment

To examine the effects of manufacturing on surface quality, hence the performance of the RF structure

Simulation

Investigate the relations between Surface defects and RF breakdown in RF accelerating Structures

Examine the effects of Surface features on field profile

Track free electrons in RF cavities

Investigate various phenomena such as secondary electron emission, Heat and stress deposition on RF surface due to particle impact

Proposed Research Program

EXPERIMENT (Button Test) MuCool

Single part

New Design

• 2 Individual Parts

Cap Holder

Surface is characterised by: Interferometer (Physical) XPS (Chemical)

Experimental Procedure

Cap Forming Surface Characterisation Holder Forming

Cap Material Selection Surface Characterisation

Final Cap Surface Characterisation High Power Testing

Cap Surface Treatment Surface Characterisation

A Typical Surface After Mechanical Polishing of OFHC Copper

Up to 1500 Angsrom Evidence of re-crystallisation due

to plastic strain and /or local temperature increases

Lower Slab shaped cells with sharp

boundaries

Deeper still More defuse boundaries

Virgin CopperMatthew Stable - 2008

INTERFEROMETER RESULTS

Matthew Stable - 2008

Mechanical polish and chemical etch remove deep scratches while EP reduces the average roughness

EXPERIMENTAL SETUP AND EP RESULTS

Characteristic Plot for Electro Polishing

XPS RESULTS

Matthew Stable - 2008

BUTTON POLISHING

Current Technique

U shaped cathode inserted within electrolyte

Rotated on Axis until bright finish observed

Changing electric field, Constant I/V

Proposed Technique

Dual `bobbin` 180˚ out of phase

Twin formed cathodes

Changing electric field, constant I/V

PARALLEL RESEARCH In collaboration with BNL (Diktys Stratakis , Harold Kirk, Juan Gallardo,

Robert Palmer)

0.07 cm

0.06

cm

CAVEL

201.23 MHz

Diktys Stratakis, 2008

Model Setup

On-Axis Defect Off-Axis DefectModel 1

805 MHz cavity with no defect (top view)Models 2 & 3

805 MHz cavity with a single defect (bottom view)

700 μm

600 μm

ELECTRIC FIELD PROFILE (MODEL 1 )

The colour bar is a good representation of the field. However, it needs to be scaled in order to represent the

actual field values

803.45 MHzMaximum E Field at the Centre of

Cavity

ELECTRIC FIELD PROFILE (MODEL 2 – OFF AXIS )

803.46 MHz Maximum E Field at the Tip of the Asperity

The overall Field profile is similar to model1, as the Asperity enhances the field locally. This is due to the small defect size compared to the

actual RF cavity

COMSOL IN BUILT TRACKER

Model 2 – Particles emitted from a distance of 0.00071m away from the RF surface (tip of the Asperity)

Model 1

Particle Tracking Procedure

Obtain Cavity’s Field Profile in Comsol

Contact with wall ?

Extract E & B Field Parameters at particle’s position (primary & new)

Obtain new particle position using 4th & 5th order

Runge Kutta Integration

Does Particle go through the Surface ?

Measure the amount of energydeposited onto the Impact surface

Dead Particle

Yes

No

Yes

Measure the number of SEs and their Orientation

Stage 1

Define a new set ofcoordinates for each particles

Investigate surface deformation and heating

No

Stage 2

Stage 3

SO WHERE WE ARE? New Batch manufactured EP and Scanning underway New Acid Mixture is being studied High power RF test (date depending on MTA

refurbishing and above work)

Validating stage 1 results (code almost finished)

Identifying the requirements for stage 2 and 3

Both studies will be ready to be presented at IPAC10