HOM Third harmonic cavities, coupled
modes and FEL optimisation
Manchester University Christmas talk 21/12/2010
Dr Ian Shinton
The University of Manchester; Cockcroft Institute, Daresbury, UK
FLASH
 FLASH is a user free electron laser facility at DESY, Hamburg Germany
 The Linac of “currently” 8 TESLA modules
 FLASH produces wavelength in the range of 80nm to 120nm and is a pilot facility for XFEL and
the ILC
 The FLASH produces pulse trains with up to 800s length at a repetition rate of up to 10Hz.
The pulse spacing is usually 1s (1MHz)
ACC39
XFEL
 XFEL is the European X-Ray Free Electron Laser facility currently
under construction at DESY
XFEL is a linear FEL machine that is 3.4km in length utilising
superconducting RF technology
XFEL accelerates electrons up to 20GeV, built on much the same
design as FLASH.
2
A 3.9GHz bunch shaping cavity ACC39
A third harmonic “bunch shaping cavity” counteracts the energy
spread a bunch sees as a result of cosine fields used to accelerate
the bunch. ACC39 has been built by FNAL and installed in FLASH.
FLASH uses one ACC39 module, XFEL will use three ACC39.
Need to understand the higher order modes (HOM) – can lead to
beam break-up, wakefields, unwanted multicavity modes and
prevent the accelerator from working….. We have made an
electromagnetic dictionary up to 10GHz mapping out the modes.
Also can use the HOM’s (which are a by-product) to produce an
ultra fast beam position monitoring system – this is currently being
designed.
Part of collaborative effort with DESY, FNAL, University of
Rostock and Royal Holloway - The University of London
ACC39 GSM
0
0
-20
-40
GSM simulation
Full HFSS simulation
Cutoff frequency
-30
S21 DB
S21 DB
Experimental measurement at FLASH
GSM simulation
-20
-10
-40
-60
-80
-50
-100
-60
-120
4
4.1
4.2
4.3
4.4
4.5
/2
4.6
4.7
4.8
4.9
5
-140
4
4.5
5
GSM is a semi analytical technique used to simulate large
RF structures that cannot be simulated by numerical
techniques due to time and memory constraints
GSM successfully applied to complicated structures –
coaxial ports.
Full ACC39 modelled using GSM.
Good qualitative agreement with experimental data above
cut-off. Results very similar to those calculated by the CSC
technique of Rostock.
GSM calculations confirm strong presence of HOM
multicavity coupling – problem for present diagnostic design
5.5
6
6.5
7
7.5
8
 /2 : GHz
0
-50
C1 - Single uncoupled cavity
C1+C2 - Two coupled cavities
Idealised eigen modes (sans coupler)
-100
S21 DB
-70
-150
-200
-250
-300
-350
4
4.1
4.2
4.3
4.4
4.5
/2
4.6
4.7
4.8
4.9
5
A Crab cavity for CLIC
The present design requires the beams to collide at a small crossing angle of 10mrad per line.
Transverse deflecting cavities, referred to as "crab cavities", are installed in the beam delivery
system (BDS) of linear collider designs in order to make the final luminosity at the interaction
point (IP) comparable to that for a design for a head on collision.
Courtesy of J.Jones, DL
Manchester’s Crab cavity for CLIC obligations
1) Basic tolerance calculations of the CRAB cavity using the
special edition of the PLACET beam tracking code and Guinea-pig
2) Long range wakefield study of the CLIC crab cavity using the
special edition of the PLACET beam tracking code
 3) Tolerance study using beam off-sets for the CLIC CRAB cavity
using the special edition of the PLACET beam tracking code and
Guinea-pig.
4) Short range wakefield study of the CLIC crab cavity using the
special edition of the PLACET beam tracking code
CLIC CRAB Beam dynamics
34
2.54
CLIC CRAB cavity voltage scan
x 10
Plot of first bunch electron lines with and without Crab at IP yz after coordinate transform
- Common: Phase=0, Roll=0
0.02
e line: V=2.587
e line: V=0
0.015
first bunch
Poly fit
2.53
2.52
0.01
2.5
0.005
y: um
Luminousity cm -2s-1
2.51
2.49
0
2.48
2.47
-0.005
2.46
-0.01
2.45
2.44
2.5
2.55
2.6
2.65
2.7
2.75
-0.015
-150
-100
-50
0
z: um
V: MV
100
150
Plot of first bunch electron lines with and without Crab at IP xx' after coordinate transform
- Common: Phase=0, Roll=0
20
e line: V=2.587
15
e line: V=0
10
5
x': um
Used PLACET and GUINEA-PIG to complete
suicidal beam dynamic study on CLIC BDS and
determined working gradient and tolerances to
achieve 98% working luminosity
Discovered an anomaly in the luminosity i.e. a
loss of 10% as a result of adding the crab cavity.
An investigation proved that the entire CLIC
BDS must be redesigned as a result of the
multipole magnet geometric effects as a result of
the installation of the CRAB cavity – Deepa Angal
Kalinin has assigned James Jones to work on this
aspect in conjunction with me.
50
0
-5
-10
-15
-20
-25
-14
-12
-10
-8
-6
x: um
-4
-2
0
2
OTHER THINGS
Did many, many other side projects…….
Several conference papers….
Various local committees
/workshops
Various bits of teaching
THINGS TO DO
Publish everything I’ve done over the years – in proper journals!
Complete shift work at DESY on the 3.9GHz and also the 1.3GHz cavities at FLASH this
January/March
Finish all side projects…. Get journal papers…..
Have two NIM papers being reviewed this January
EGRESS