Transcript RFQ development

Imperial College London, FETS
Work towards high performance
accelerator structures
1. Introduction
2. Cavity & electrode design, particle dynamics
3. Critical fields, sparking, multipacting
4. Surface treatment and conditioning
5. Design of RFQ and test models
6. Conclusions
1
Imperial College London, FETS
2
Work to be performed
Year 1
Year 2
Year 3
Year 4
Year 5
1) Decision of frequency  (RFQ = 324 MHz , MC = 201 MHz)
2) Particle dynamic design (field level, Kilpatrick) => Length of RFQ
3) Electro dynamic design of resonator and Endplates => Choice of RFQ type
4) Electro dynamic design of tuners, couplers
5) Design of models, production and tests of models
6) Mechanical design of RFQ, Endplates, Positioners
7) Mechanical design of tuner, couplers, etc
8) Production of RFQ
9) Production of tuner, couplers, and support
10) Assembly of RFQ, test and commissioning
11) Test of production methods and surface conditioning
Year 6
Imperial College London, FETS
3
Particle dynamics in the RFQ
A RFQ electrode design has to be
found by numerical simulations
which fulfils :
-
High beam transmission
- low emittance growth
and defines
-
Improved design
Classic design
- electrode voltage
electric surface field strength
HERA
Imperial College London, FETS
4
Electrodynamic design of the RF resonator
U2
R 
 30(40)kWm
P
Qexp
  Rth 
Rexp
(4Vane)
QSF
N
 330kW / m (U  100kV )
L
N  1.5  2MW
Ploss  50kW / m
 DF  15%
R’ [kWm]
Challenges are : High shunt impedance, low resistive losses, concentration of
fields onto axis
4-Vane
4-Rod
Split coaxial
D-H-resonator
P
~30 kWm
For the FETS (DF~7.5 %) the average
resistive power loss is app. 25 kW/m.
Therefore 4 vane and 4 rod structures
are possible.
f [MHz]
Imperial College London, FETS
5
Field emission of electrons from surfaces
Fowler Nordheim :
1.5

7 0 
  6.810

E 

2
J  B E e
J: emission current density (A/cm2); B: field-independent constant [A/V2]; E: applied field
(V/cm) F0: work function (eV)
Imperial College London, FETS
6
The maximum electric field and the Kilpatrick limit
Due to sparks and dark discharge, the maximum
potential on the Electrodes is limited. This limit
is not only a function of the aperture of the
RFQ, but also of the RF frequency and the
quality of the electrode surfaces.
f ( MHz)  1.64 10 E  e
4
2
 0.085 


 E 
(E in MV/m)
U
(a  average aperture)
a
but also thefollowinglinear approximation is used :
E  1.36 
U K (kV )  10(1  g (m m))  (1  1.5 103  f ( MHz))
g  smallest distanceof electrodes
U K (kV )  10(1  4)  (1  1.5 103  324)  74 (~ 1.4)
The Kilpatrick factor (usually between 1.5 and 2 for
RFQ’s) is the factor between the applied
potential on the electrodes and the spark limit
given by the semi empirical theory of Kilpatrick.
Electrode potential as a function of gap
distance and RF frequency
Imperial College London, FETS
7
Multipacting 1
E( n )
4m
f 2l


e (2n  1)
Multipacting in a acceleration cavity (left - field plot), occurs at low and medium
power levels, when primary electrons are accelerated by the electric field and
generate (in resonance with the electric field) an avalanche of secondary electrons.
Imperial College London, FETS
8
Multipacting 2
In opposite to rectangular cavities, in elliptical
cavities the path of the multipacting electrons is
directed towards the axis of the cavity. Therefore
the avalanche is interrupted when the electrons
reach the cavity iris. The magnetic self fields in
the cavity do not prevent the propagation of the
electrons towards the axis !
Imperial College London, FETS
9
Multipacting 3
Range of electric field for
multipacting measured
for different materials.
Due to the impact of the
electron current on the
surfaces a conditioning
of the surface might
occure and the
multipacting levels off.
*
*
Imperial College London, FETS
Influence of external
magnetic field
10
Imperial College London, FETS
Sparking and surface treatment
Los Alamos experiments using a 420 MHz 4 Vane RFQ in 1981
11
Imperial College London, FETS
Surface damage due to sparks
Pitting of surface
12
Imperial College London, FETS
Conditioning by RF
13
Imperial College London, FETS
Gas discharge and low energy ion impact on surfaces for surface
conditioning by sputtering of field emitters
14
Imperial College London, FETS
15
Mechanical design & construction of the RFQ-Testmodel
(example TRASCO RFQ Legnaro)
20 cm technology test
model
Results of
measurements at test
model
RFQ Parameters
Imperial College London, FETS
16
Mechanical design & construction of the RFQ
(example TRASCO RFQ Legnaro)
While the first section of the RFQ
fulfilled the requirements after
production and brazing, the second
one showed a deviation of 0.1 mm
due to bending of the vanes and
there from a deviation of the
resonance frequency which could not
be adjusted by the tuners.
Drawing of the first section and final first
section
Imperial College London, FETS
17
Conclusions
Work to be performed :
1. Numerical Design of RFQ resonator and RFQ electrodes for
- High beam transmission
- Low emittance growth
- moderate electric field
- Low power losses
2. Numerical determination of critical fields in RFQ and MICE cavity
design – investigation on influence of magnetic fields on
HV break down effects
3. Definition on required surface and production quality
Validation of production techniques and (cold / hot) test of models
4. Experimental determination of required surface treatment and surface
conditioning procedures and sparking test on models
and verification of resonator design.