Damping Rings
James Jones
ASTeC, Daresbury Laboratory
Personnel
ASTeC
Oleg Malyshev
–
Vacuum Studies
James Jones
–
Low emittance tuning studies
+ Engineering Support
Liverpool University
Andy Wolski
–
Low emittance tuning, Vacuum Design
1 RA currently employed looking partially at impedance effects
RA 7.1.4
–
Vacuum Design (LC

ABD 2)
RA 7.2.1
–
Vacuum Design (LC

ABD 2)
Larisa Malysheva
–
Polarisation issues (As part of a wider
collaboration)
Damping Rings must provide very high quality, very stable beams
ILC DR
PEP II LER
HERA e
SPring

8
Energy
5 GeV
3.5 GeV
27.5 GeV
8 GeV
Circumference
6694 m
2199 m
6360 m
1435 m
Average current
400 mA
2450 mA
58 mA
100 mA
Number of bunches
2767
–
5782
1588
156
2016
Particles per bunch
2
×
10
10
–
1
×
10
10
7
×
10
10
3.7
×
10
10
0.15
×
10
10
Extracted bunch length (rms)
6 mm
12 mm
9 mm
6 mm
Horizontal normalised emittance
8
μ
m
200
μ
m
1000
μ
m
90
μ
m
Vertical normalised emittance
0.02
μ
m
10
μ
m
180
μ
m
0.09
–
0.27
μ
m
DR
DR
Damping ring parameters are very demanding in terms of beam stability:
No operating machine meets all the parameters simultaneously.
Previous Work
ASTeC was, and is, committed to damping ring work as part
of the EUROTeV framework:
Both O. Malyshev and J. Jones have small but significant
work packages within this framework
EUROTeV WP 3 Task 2 deals with electron cloud issues
within the positron damping ring, and mitigation schemes.
O. Malyshev was a major contributor to vacuum simulations
for this task.
EUROTeV WP 3 Task 3 deals with low emittance tuning
simulations for the damping rings. The work is coordinated
by J. Jones.
Vacuum Simulations
Complete simulations of damping ring vacuum systems, including both the dipole induced
photo desorption, along with analysis of pump location and speeds
Conventional vacuum technology does not allow to reach required vacuum after 100 Ah
conditioning
NEG coated chamber provides cheapest and simplest vacuum solution for dumping ring:
less number of pumps
less pumping speed required
lower bakeout/activation temperature (180C in stead of 250

300C)
low SEY to suppress e

cloud effect
Low Emittance Tuning
Have a large scale emittance tuning simulation environment
for the RDR damping rings.
Includes full orbit, dispersion and coupling correction.
Models the effects of ATL

like ground motion on the time
evolution of the output emittances.
None
CO
Full
Initial Correction
Only
Full Correction every 6 Days
Major issues for beam stability
Electron cloud effects in the positron damping rings
One of the top priorities for damping rings R&D: already receiving major
attention from groups around the world.
Ion effects in the electron damping rings
Still some uncertainty in likely impact on damping rings performance. Can
probably be mitigated with feedback systems and a well

designed vacuum
system.
Impedance

driven beam instabilities
Wide experience from operating facilities; we expect the damping rings to
operate in a challenging regime.
Long

range wake fields can drive multibunch instabilities, and couple jitter from
freshly

injected bunches to damped bunches awaiting extraction.
Short

range wake fields can drive single

bunch instabilities, which can appear
as emittance increase, or a “bursting” type of instability.
These effects require careful study, with beam dynamics models closely
connected to the technical design of the vacuum system.
Using a time

domain simulation code, we studied the coupling of injection jitter to
damped bunches in the NLC damping rings.
Task 2.1 Goal 1
Evaluate the effects of beam loading, injection/extraction transients and long

range wake fields in the damping rings under a range of operational conditions.
t = 0
ms
t = 10
ms
Similar (or stronger) effects are expected in ILC.
Studies must include a detailed impedance model (resistive wall and HOMs),
lattice model, radiation damping and feedback system. Our present code does
include these effects: alternatives may be available (e.g. MULTI

TRISIM).
Task 2.1 Goal 2
Single

bunch instabilities are diverse and complicated. There is a lot of operational
experience of these effects, but a good understanding for any given machine
generally requires a lot of hard,
detailed
work.
Single

bunch instabilities were a major problem for the SLC damping rings:
eventually, the vacuum chamber had to be rebuilt.
Evaluate impedance

driven instability thresholds and growth rates.
Single

bunch instability in
the SLC damping rings.
Left: Experimental
observation
(B. Podobedov, BNL).
Right: Simulation
(K. Oide, KEK).
We shall collaborate with LBNL and SLAC in the construction of impedance
models (using technical designs of the vacuum chamber, to be performed in Task
2.2) and the evaluation of the resulting instabilities.
Task 2.1 Goal 3
Lowest achieved vertical emittance (after significant effort) is 4.5 pm in KEK

ATF. The ILC
specification is for 2 pm.
Develop techniques for low

emittance tuning.
Several techniques (orbit/dispersion/coupling correction; orbit response matrix
analysis…) work well in simulation, but practical implementation with the necessary
accuracy and precision is still extremely challenging.
We need to demonstrate a technique that can be routinely applied to a (6 km) ring to
achieve vertical emittance of 2 pm on a regular basis.
Emittance

tuning using ORM analysis in the KEK

ATF.
Task 2.1 Goal 3
The main facilities used so far for experimental studies of low

emittance tuning have been the
KEK

ATF and the LBNL

ALS.
The ATF will continue to be available.
The main limitation so far has been the availability of personnel.
Producing a high

quality beam from the storage ring will be essential for ATF2.
Beam time at the ALS is generally available at monthly intervals.
The main limitation tends to be the availability of staff to run the experimental studies.
There are presently two serious proposals for future damping rings test facilities:
CESR

tf could start operations for damping rings studies as early as June 2008.
HERA

DR could start operations in late 2009.
In both proposals, low

emittance tuning would be an important part of the programme.
Further opportunities are provided by other machines.
Light sources, e.g. DIAMOND.
KEK

B (proposed damping rings study programme starting in 2009, to include low

emittance tuning).
Develop techniques for low

emittance tuning.
Task 2.2 Goal 1
Calculate the average pressure and pressure profile in the damping
rings and, in the context of the results of these calculations,
evaluate the technology options for the damping rings.
4
6
8
10
12
14
16
18
20
1
10
11
1
10
10
1
10
9
H2
CH4
CO
CO2
Thermal desorption
Required CO pressure
Stainless steel tube, S=100 l/s
L (m)
P (Torr)
10
100
1
10
3
1
10
12
1
10
11
1
10
10
1
10
9
H2
CH4
CO
CO2
Thermal desorption
Required CO pressure
Distance between pumps L=10 m
S (l/s)
P (Torr)
5
10
15
20
25
30
1
10
13
1
10
12
1
10
11
1
10
10
1
10
9
H2
CH4
CO
CO2
Required CO pressure
NEG coated tube, S=20 l/s
L (m)
P (Torr)
10
100
1
10
3
1
10
13
1
10
12
1
10
11
1
10
10
1
10
9
H2
CH4
CO
CO2
Required CO pressure
Distance between pumps L=10 m
S (l/s)
P (Torr)
Initial evaluations have been performed, as part of the EUROTeV programme, and have indicated
the benefits of NEG

coated vacuum chamber.
Detailed studies are now needed to evaluate the benefits of NEG

coating, and to produce
technical specification for the vacuum system (apertures; antechambers; material and coating;
pumping locations; pumping speeds etc.)
Vacuum studies must be well

integrated into studies of electron cloud and ion effects.
Calculation of the pressure in a
section of the ILC damping rings
in two different scenarios for the
vacuum system, as a function of
the spacing between the pumps.
Left: Stainless steel tube.
Right: NEG

coated tube.
(O. Malyshev, ASTeC)
Task 2.2 Goal 2
Determine conditioning rates for NEG coatings under various conditions.
We know that:
the initial pressure in a NEG coated vacuum chamber activated at 180
C is better by two orders of
magnitude than that in a stainless steel vacuum chamber baked in

situ to 300
C;
NEG outgassing rates reduce with accumulated photon dose. Data from the ESRF show that this
reduction could be up to 2 orders of magnitude;
NEG coating simplifies and reduces the cost of the pumping system, and works to mitigate
multipacting.
In other words, we know that NEG coating is worth using.
For the design of the vacuum system, we need to know the photon and electron

stimulated
desorption yields:
as functions of photon or electron dose, up to very large doses;
as functions of photon or electron energy;
as functions of NEG activation temperature (from room temperature up to 250
C
);
after air vent to different pressures (from 10

6
mbar to atmosphere), to determine whether
recovery after an accident requires reactivation.
As a result of the experimental studies:
we will be able to produce (for the TDR) a vacuum system design optimised for performance and
cost, including spacing and required pumping speeds of the lumped pumps;
we will gain invaluable experience in the use of NEG coatings under a wide range of conditions.
Task 2.2 Goal 3
Produce technical designs for components in the vacuum chamber in the arcs
and straights, and use these designs for developing an impedance model.
Technical designs of components in the vacuum chamber are essential for constructing an
impedance model.
Need to include bellows, flanges, tapers, pumping ports, BPMs, antechambers, kickers
and septa…
Close collaboration with other technical groups (e.g. instrumentation) is essential.
Producing a complete, detailed model is a significant amount of work, but is essential for a
reliable evaluation of the impact of collective effects.
Calculation of trapped modes in
PEP II bellows. Higher

order
mode heating is a significant
problem for PEP II, and a potential
problem for the ILC damping rings.
(Cho Ng, SLAC).
We will collaborate with LBNL on the technical design, and with SLAC on the
impedance modelling.
Task 2.2 Goal 3
The goal of producing a detailed impedance model for the TDR, based on technical
designs of the important components, is ambitious.
The Damping Rings Workshop at Cornell, 26

28 September, outlined a staged
plan, with specified milestones towards the goal of a complete evaluation of the
impedance

driven collective effects.
Begin with constructing an impedance model based on scaling components
from existing facilities, in parallel with the technical design of the damping rings
vacuum.
Proceed iteratively to improve the model, using the results of the scaled
impedance model to guide the design work, so as to achieve a specified
impedance budget.
Our proposed work on the vacuum system fits extremely well with the timescales
and methodologies.
If the hoped

for contributions from other labs (LBNL and SLAC) are not provided,
we still make an essential contribution towards a reliable impedance model.
Produce technical designs for components in the vacuum chamber in the arcs
and straights, and use these designs for developing an impedance model.
Final Words
The damping ring work proposed addresses two critical and related issues for the
ILC damping rings:
Dynamical effects that potentially limit beam quality and stability.
Vacuum system specification and design.
We will make a leading contribution to the ILC in these areas.
The work we are proposing will produce results needed for the TDR on an
appropriate timescale.
We will collaborate with identified international partners to maximise the benefit of
the resources that are available.
Vacuum studies have the potential for industrial involvement, and a major
contribution (> 13 km of vacuum system) during construction.
The tasks are closely connected to other work packages within Cockcroft, for
example:
Effects of linac wakefields depend on beam stability from damping rings
Instrumentation and feedback essential for maintaining stability
Comments 0
Log in to post a comment