PSI
contribution to the
ILC-EUROTeV
collaboration
Work
Package 6: Integrated
Luminosity Performance Studies (ILPS)
Task: Bunch Compression Design (BCDES)
1 Task quick overview
1.1
Magnetic chicane.
1.2
See as well: ILC chicane
design concepts
(http://www-project.slac.stanford.edu/ilc/acceldev/LET/BC/)
In a collider the bunch specification at
the interaction point, such as transverse and longitudinal dimensions, are
driven by the Luminosity requirement.
In order to focus the entire bunch at the
collision point for maximum luminosity the rms bunch length has to be ≤ than
the vertical beta function βy.
The luminosity needed for a 0.5-1 TeV
Center of Mass collider requires a very strong focusing which in the present
design studies corresponds to a βy of the order of only 400 μm. For this
reason in the baseline design concept one assume today an rms pulse length of
300 μm, with a possible extension down to 150 μm.
In the present ILC design concepts, the linear
collider injector complex (injector linac + damping ring) is generally
providing a low emittance beam with a longitudinal rms extension σz of the order of 6 mm. In order to achieve the specified
length the bunch must therefore be compressed by at least a factor of 20 before
being injected in the main linac.

Figure 1: schematic view of the accelerator
complex
As represented in figure 1, the needed
compression can be carried out using a magnetic compression chicane. Two or
more compression stages are more suitable for large compressor factors and
probably the only alternative for bunch length below 300 μm (30 μm
for CLIC).
Within the EUROTeV collaboration, the first
task of PSI will be:
A conceptual
design of a magnetic compression chicane suitable for multi-TeV colliders.
We will concentrate our efforts on the
final compression stage sitting in front of the main linac, which is the more
critical with respect to the Coherent Synchrotron Radiation (CSR) effects because
of the shorter bunches and the highest energy of the beam. As first approach we
will start from the existing ILC design concepts for a 500 GeV Center of Mass
collider (Tesla and US-cold) and CLIC, assuming minor changes in the injector
complex for a multi-TeV upgrade.
1.3
Path length tuning chicane
See as well: D.Schulte,
Proceedings of LINAC
2004, pp138, Lübeck, Germany (2004)
CLIC drive beam path length chicane layout
(http://clic-study.web.cern.ch/CLIC-Study/Report/34DriveFreq.html#342)
(from CLIC report: CERN 200-008)
CLIC Layouts
(http://clic-study.web.cern.ch/CLIC-Study/Layout/Overall.html)
In the two beams CLIC concept
a drive beam is accelerated up to 2 GeV and then decelerated to produce the RF
power needed to feed the main linac 30 GHZ accelerating structures. As shown on
figure 2, in order to obtain the required time structure and energy, the bunch
trains of the drive beam are first accelerated and then manipulated with a
delay loop and in two combiner rings.
The RF voltage and phase
jitter in the main linac accelerating structures depends directly on the drive
beam stability (current and phase). The main process which can lead to a
longitudinal jitter of the drive beam is connected to the small energy
deviations induced by RF phase or amplitude jitters that are turned in
longitudinal jitter by the bunch compressor. Temperature changes, ground motion
and timing errors can as well contribute to this phenomenon.
The relative energy error
after the main CLIC linac should not exceed 0.1, which correspond to a maximum
phase error of 0.25o (at 30 GHz). To accommodate this tight requirements
the jitter of the longitudinal drive beam must therefore be stabilized within 7
μm. For this reason a phase feedback system that should reduce all
incoming phase errors of the drive beam as to be
considered. The most efficient location for this system is immediately upstream
of the decelerator units in the so called “turn around loop” where the bunch
further compressed (figure 2).


Figure 2: Schematic view of the CLIC
complex
Figure 3 shows the principle of the drive
beam feedback system at the turn around loop. The correction range suited for
such a system is ~100 μm, the compression factor ~10.

Figure 3: Schematic view of the drive beam
phase feedback system.
Within the EUROTeV collaboration, the
second task of PSI will be:
A conceptual design of a path
length tuning chicane
The chicane will be used as well as bunch
compressor in order to match the decelerating structures. The total requested compression
corresponds to a reduction of the bunch length from 2mm down to 290-170 μm. In the early design
reported in the CLIC
study the correlated r.m.s.
momentum-spread needed was approximately 1.5-1.2%, which correspond to and R56
~0.15 m.
2
Beam parameter summary
2.1
Compression chicane
See as well: Proposed
ILC beam parameters (SLAC-ILC web server)
CLIC
parameter Table (CERN CLIC web server)
CLIC report CERN
2000-008, 28 July 2000
ILC,
Y. Kim proposal, 1st ILC workshop at KEK, 2004

Table 1: CLIC beam parameters (Multi-TeV
collider example)
Table
1 gives an overview of the ambitious beam parameter at the chicane and at the
interaction point for the CLIC project. Those values are continuously
re-discussed, and should be used as baseline for the ultimate chicane design.
The
chicane designs proposed for the ILC beam parameters (0.5 – 1 TeV Center of
Mass) summarized in table 2 could be considered as the starting point towards
the ultimate chicane design for a Multi-TeV collider. The optimum energies at
which the compression should take place and the compression factors should be
nevertheless adapted in order to minimize the strong collective effects due to
the shorter bunch and the highest peak current.

Table 2: ILC proposed beam parameters
2.2 Tuning chicane
See as well: CLIC
study (chapter 3)

Table 3: CLIC drive beam parameters
3 Chicane principle and difficulties – qualitative
short overview.
See as well:
Andy
Wolsky, Introduction to Bunch Compression
The longitudinal size of a relativistic
beam is not changed by the acceleration along the linac, and the collective effects
with short bunches in the damping rings prevent us to reach the specified
length already at the linac injection. A strong compression can nevertheless be
achieved by a rotation of the longitudinal phase space in a magnetic chicane,
which reduces the bunch length to the expense of the energy spread. The
principle of the compression chicane is shown at figure 4. The incoming beam
travels of crest trough an accelerating structure, which introduces a
correlated energy spread along the bunch (less energy at the head, more energy
at the tail). The beam is then bended using few dipole magnets and since the
bending angle depends on the energy the head and the tail of the bunch
experience a different path along the chicane.

Figure 4: Compression process in a magnetic
chicane. (courtesy of Paul Emma).
The compression chicane principle is
relatively simple, but it is necessary to find a compromise between the
following aspects:
1) The final energy dispersion ΔE/E introduced by the energy chirp
needed for the compression process, can’t be too large in order to avoid a
possible reduction in the luminosity (usually ΔE/E ≤1%).
2) The bending angle at each dipole and the
optimum beam energy must to be selected in order to prevent a too large
emittance dilution due to Coherent Synchrotron Radiation (CSR) and Incoherent Synchrotron Radiation (ISR). Since the last
compression stage take place at high energy (5-9 GeV), both effects have to be
taken into account.
3) A compromise between costs/size and
performances has to be found.
4) The chicane should accommodate some small
fluctuations of the incoming beam (energy, alignment).
CSR and ISR effects are the main unwanted
effects driving the chicane design. Figure 5 describe schematically the
process. While bending the beam, each bunch radiate coherently at wavelength
longer than the bunch length and incoherently at longer wave length. Since the
bunch get compress the amount of radiation a short wavelengths increase along
the chicane and overtake together with the ISR the bunch itself increasing the incoherent
energy dispersion. Because of the different bending angle at the chicane dipole
magnets, the additional incoherent energy spread causes a transversal size growth
(emittance) of the beam on the bend plane. Moreover recent numerical
observations show that small current modulations along the bunch can be
strongly amplified by CSR. This phenomenon can be reduced by long bends, and
introducing some additional un-coherent energy spread in front of the chicane.

Figure 5: Synchrotron radiation .in a
bending magnet interacts with the bunch generating an uncorrelated energy
spread which break the achromatic system (courtesy of Paul Emma).
4
Work plan - strategy
4.1
First order design
For the first design step we assume a two
stage compression scheme and we concentrate on the second stage. Longitudinally
we will use a perfect Gaussian distribution and a bi-Gaussian distribution for
the transverse plane, the energy chirp will be linear. The CLIC beam parameters
described here above, which are the most difficult to achieve, represent our
goal for a multi TeV collider. While the ISR effects can be estimated
analytically, the CSR contribution to the emittance dilution will be here
evaluated using 1-D CSR numerical models.
In a similar way the path length chicane will
be designed assuming the beam parameters described above, and optimized with
respect to the CSR effects.
Few tracking codes include a CSR module
capable to models the interaction between the radiation and the electron bunch.
The simplest and faster models are one-dimensional and can be used as first
approximation of the CSR effects before to go for more complicated and cpu consuming approaches. The
program ELEGANT
by M. Borland is a general multi purpose tracking code widely used for storage
rings or linac optic designs, it include a one-dimensional model for CSR.
One-dimensional computations can be made as well using the 1D solver option of
the CSR-track
code developed at DESY. As benchmark the simulation results made by those two
codes will be compared.
4.2
Second order design
The chicane optimizations made here above
will be investigated and re-optimized using a two dimensional and
three-dimensional CSR models. Comparisons between CSR-Track and Trafic4
results are suitable. Sensitivity to micro-bunching instabilities should be
investigated as well.
4.3
Third order design
A more realistic phase space for the
incoming beam has to be used in order to evaluate the impact of non-linearity
of the longitudinal phase space on the compression process. For this purpose we
expect an input from the start-to-end simulations carried out within the
EURO-TeV program.
4.4
Fourth order design
The chicane performances, the matching
optic and the turn around lattice need to be characterized and optimized with
respect to the possible dynamic and static imperfections (energy jitter,
alignment errors).