ILC Damping Rings: Configuration Status and R&D Plans

Andy Wolski

Lawrence Berkeley National Laboratory

January 19, 2006

2/25

Baseline Configuration

Item Baseline Alternatives

Circumference (e+) 26 km(e-) 6 km

1. (e+) 6 km2. (e+) 17 km

Beam energy 5 GeV

Injected emittance & energy spread 0.09 m-rad & 1% FW 0.045 m-rad & 2% FW

Train length (bunch charge) 2700 (2×1010) - 4050 (1.3×1010)

Extracted bunch length 6 mm - 9 mm

Injection/extraction kickertechnology

Fast pulser/stripline kicker 1. RF separators2. Fourier pulse compressor

Wiggler technology Superconducting 1. Normal-conducting2. Hybrid

Main magnets Electromagnetic Permanent magnet

RF technology Superconducting Normal conducting

RF frequency 500 MHz (650 MHz)

Vacuum chamber diameter,arcs/wiggler/straights

50 mm/46 mm/100 mm

3/25

Top Priority: Baseline lattice design by end of March 2006

Circumference 6476.7163 m

Energy 5 GeV

RF frequency 500 MHz

Harmonic number 10802

Transverse damping time e+ (e-) <25 ms (<50 ms)

Normalized natural emittance 5 µm

Equilibrium bunch length 6 mm

Equilibrium energy spread <0.13%

Momentum compaction ~ 4×10-4

Damping wiggler peak field 1.67 T

Damping wiggler period 0.4 m

Energy acceptance ||<0.5%

Dynamic aperture Ax+Ay<0.09 m-rad (up to ||<0.5%)

There are additional specifications on tunes and optics…

4/25

Design studies of dogbone alternative will continue

Circumference 17227.9195 m

Energy 5 GeV

RF frequency 650 MHz

Harmonic number 37353

Transverse damping time e+ (e-) <25 ms (<50 ms)

Normalized natural emittance 5 µm

Equilibrium bunch length 6 mm

Equilibrium energy spread <0.13%

Momentum compaction ~ 1.5×10-4

Damping wiggler peak field 1.67 T

Damping wiggler period 0.4 m

Energy acceptance ||<0.5%

Dynamic aperture Ax+Ay<0.09 m-rad (up to ||<0.5%)

5/25

Baseline lattice specification allows flexibility in fill patterns

Ring circumference [m] 6476.7163

Harmonic number 10802

Ring RF frequency [MHz] 500

Linac RF frequency [GHz] 1.3

Linac pulse length [ms] 0.97

Linac bunch spacing [linac RF wavelengths] 468 390 351 312 234

Linac bunch spacing [ring RF wavelengths] 180 150 135 120 90

Linac bunch spacing [ns] 360.00 300.00 270.00 240.00 180.00

Ring bunch spacing [linac RF wavelengths] 5.2

Ring bunch spacing [ring RF wavelengths] 2

Ring bunch spacing [ns] 4.00

Bunches per train 45

Number of bunch trains 60 72 80 90 120

Gaps per train 45 30 22.5 15 0

Gap length [ns] 184.00 124.00 94.00 64.00 4.00

Total number of bunches 2700 3240 3600 4050 5400

Particles per bunch [×1010] 2.07 1.73 1.56 1.38 1.04

6/25

Alternative lattice specification also allows flexibility in fill patterns

Ring circumference [m] 17227.9195

Harmonic number 37353

Ring RF frequency [MHz] 650

Linac RF frequency [GHz] 1.3

Linac pulse length [ms] 1.03

Linac bunch spacing [linac RF wavelengths] 540 360 180

Linac bunch spacing [ring RF wavelengths] 270 180 90

Linac bunch spacing [ns] 415.38 276.92 138.46

Ring bunch spacing [linac RF wavelengths] 18 12 6

Ring bunch spacing [ring RF wavelengths] 9 6 3

Ring bunch spacing [ns] 13.85 9.23 4.62

Bunches per train 6 9 18

Number of bunch trains 415

Gaps per train 12

Gap length [ns] 60.00

Total number of bunches 2490 3735 7470

Particles per bunch [×1010] 2.25 1.50 0.75

7/25

ILC Damping Rings R&D Tasks List is in development

1. Parameter specifications and system interfaces1.1 Injected beams

1.2 Extracted beams

1.3 Fill patterns and timing issues

2. Beam dynamics2.1 Single-particle dynamics

2.2 Multi-particle dynamics

3. Technical subsystems3.1 Injection/extraction kickers

3.2 Damping wiggler

3.3 Main magnets

3.4 Orbit and coupling correction

3.5 RF system

3.6 Vacuum system

3.7 Fast (bunch-by-bunch) feedback system

3.8 Instrumentation and diagnostics

8/25

ILC Damping Rings R&D Tasks List: Excerpt

2. Beam dynamics2.1 Single-particle dynamics

2.1.1 Lattice design2.1.1.1 Lattice design for 6 km baseline positron damping rings

Produce a lattice design for the 6 km baseline positron damping rings. The lattice should meet the specifications for damping time, equilibrium emittance, acceptance etc. and include all major subsystems, including injection/extraction sections, orbit and coupling correction systems, RF cavities etc. The circumference should be around 6 km, and should allow for a variety of different fill patterns (different numbers of bunches) without changes in circumference or RF frequency.

Priority/Need: High priority. Required for Reference Design Report, and to allow dynamics studies, engineering designs and costing.

Deadline: March 31, 2006

Experimental facilities: None

Investigators: Louis Emery (ANL), Aimin Xiao (ANL), Yi Peng Sun (IHEP)

9/25

Comments on the R&D Tasks List

The intention is to coordinate activities through a working document that lists R&D objectives, and that can be periodically revised and updated.

Short-term and long-term (ongoing) goals are included.

Objectives are ideally stated in terms of deliverables with deadlines.

Objectives should be developed in consultation between the investigators, the DR Area System Leaders. We want to avoid micromanaging the R&D process.

Resources are widely distributed between different laboratories. This approach provides a coherent framework for collaboration.

We are still in the very early stages. We hope that this approach provides sufficient flexibility to respond to changing project needs.

10/25

R&D Tasks List Summary Spreadsheet (Excerpt)

11/25

Links

Final version of Damping Rings Configuration Recommendation Summary Report:http://www.desy.de/~awolski/ILCDR/DRConfigurationStudy.htm

Final draft of Damping Rings Configuration Studies Report (300 pages):http://www.desy.de/~awolski/ILCDR/DRConfigurationStudy.htm

Present version of Damping Rings R&D Tasks List:http://www.desy.de/~awolski/ILCDR/

Present version of Damping Rings Lattice Specifications:http://www.desy.de/~awolski/ILCDR/

ILC Damping Rings: Fill Patterns and Timing Issues

Andy Wolski

Lawrence Berkeley National Laboratory

January 19, 2006

13/25

General comments

We assume that there will be a benefit in being able to vary the bunch charge and fill pattern in the damping rings.

Lower charge benefits single-bunch instabilities (e.g. microwave).Fewer bunches can allow longer gaps in some schemes, with potential benefits for electron cloud and ion effects: the benefits need to be better understood.Effects at the IP drive for lower charge (down to 1×1010 particles per bunch).Optimization during commissioning and operation will probably be of value.

Designing for flexibility in the number of bunches places strong constraints on the damping rings’ circumference and the lengths of other systems in ILC.

There are many solutions: here, we consider just two possible schemes.

We assume that gaps in the bunch train in the linac are to be avoided. If gaps are acceptable, this opens up further possibilities.

14/25

Scheme A: “Fixed bunch spacing” (increase no. of bunches by reducing the gaps)

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

1 2 3 4 5

1 2 3 4 5

1 2 3 4 51 2 3 4 51 2 3 4 51 2 3 4 5

1 2 3 4 51 2 3 4 51 2 3 4 51 2 3 4 51 2 3 4 51 2 3 4 51 2 3 4 5

bunch separation in linac = Tlinac

bunch separation in linac = Tlinac

bunch separation in linac = Tlinac

Extraction kicker fires regularly at intervals of Tlinac

Bunches numbered “1” are extracted on first turn;

bunches numbered “2” are extracted on second turn, etc.

We always extract over a fixed number of turns, so linac RF pulse length does not change.

RF buckets corresponding to extracted bunches are filled immediately by bunches arriving at regular intervals of Tlinac

15/25

Example A1: A 6476 m damping ring with 500 MHz RF frequency

Numbers in bold face must be integers in a valid solution.

Input values are in red; values in black or blue are calculated from these.

Grey cells indicate an invalid solution.

16/25

Example A2: A 6643 m damping ring with 650 MHz RF frequency

Numbers in bold face must be integers in a valid solution.

Input values are in red; values in black or blue are calculated from these.

17/25

1 3 5 2 4 6 1 3 5

bunch separation in linac = Tlinac = 24 ring RF buckets

Extraction kicker fires regularly at intervals of Tlinac

Bunches numbered “1” are extracted on first turn;

bunches numbered “2” are extracted on second turn, etc.

We always extract over a fixed number of turns, so linac RF pulse length does not change.

RF buckets corresponding to extracted bunches are filled immediately by bunches arriving at regular intervals of Tlinac

bunch separation in linac = Tlinac = 12 ring RF buckets

1 3 52 4 61 3 52 4 61 3 52 4 6

Scheme B: “Fixed gaps” (increase no. of bunches by reducing the bunch spacing)

18/25

Example B: A 16.2 km damping ring with 500 MHz RF frequency

Numbers in bold face must be integers in a valid solution.

Input values are in red; values in black or blue are calculated from these.

19/25

Pros and cons…

Scheme A: Fixed bunch spacingProvides greater flexibility than fixed gaps: more possibilities for numbers of bunches (e.g. 2700, 3240, 3600, 4050 or 5400 in example A1).

Can be applied in both 6 km and 16 km damping rings…

…but gaps vanish for largest number of bunches in 6 km rings.

“Local current” increases as number of bunches decreased (bunch charge increases) – may adversely affect ions or electron cloud effects.

Scheme B: Fixed gapsLimited flexibility: probably only two options for number of bunches(e.g. 3010 or 6020 bunches in example B).

Realistically requires a 16 km ring.

Fixed gaps means that ion clearing should be as effective at either number of bunches.

Local current remains constant as number of bunches is changed.

+

+

+

+

--

--

20/25

Lengths of different sections in ILC cannot be chosen arbitrarily

If L1, L2, L3 and L4 are all integer multiples of the bunch separation in the linacs, then by “time invariance” we see that bunches are always at the right place at the right time.

To retain flexibility in the fill patterns, we need to look for the least common multiple (LLCM) of the various linac bunch separations, Llinac.

L1, L2, L3 and L4 should then all be integer multiples of LLCM.

e- source

e- damping ring e+ damping rings

e- linac e- linac e+ linace+ source

IP

L1 L2 L3

L4

21/25

Lengths of sections are determined by linac bunch separation

e- source

e- damping ring e+ damping rings

e- linac e- linac e+ linace+ source

IP

If L1, L2, L3 and L4 are all integer multiples of the bunch separation in the linacs, then by “time invariance” we see that bunches are always at the right place at the right time.

To retain flexibility in the fill patterns, we need to look for the least common multiple (LLCM) of the various linac bunch separations, Llinac.

L1, L2, L3 and L4 should then all be integer multiples of LLCM.

L1 L2 L3

L4snapshot of bunch positions

22/25

We can retain flexibility by choosing lengths carefully

In example 1, the bunch separations in the linac are Tlinac = (360, 300, 270, 240, 180) ns.

LCM(360, 300, 270, 240, 180) = 10800, or LLCM = 3237.8 m. This is inconveniently large.

LCM(360, 300, 270, 240, 180) = 2160, or LLCM = 647.55 m. This is better.

LCM(360, 300, 270, 240, 180) = 720, or LLCM = 215.85 m. This could be appropriate for the 2nd IP.

e- source

e- damping ring e+ damping rings

e- linac e- linac e+ linace+ source

IP

L1 L2 L3

L4

23/25

Example, using 6476 m damping ring with 500 MHz RF frequency

L1 L2 L3

L4

L1 L2 L3 L4

6×647.55 = 3885.3 m 10×647.55 = 6475.5 m 16×647.55 = 10360.8 m 33×647.55 = 21369.2 m

Tlinac L1/(c×Tlinac) L2/(c×Tlinac) L3/(c×Tlinac) L4/(c×Tlinac)

360 ns 36 60 96 198

300 ns 43.2 72 115.2 237.6

270 ns 48 80 128 264

240 ns 54 90 144 297

180 ns 72 120 192 396

Note: If we do not start e+ DR extraction before there are new e+ bunches arriving at the injection point, then a number of e- bunches at the head of the train have nothing to collide with. We would lose about 10% of the luminosity this way, compared to the case where all bunches collide.

24/25

Example, with 2nd IP

L1 L´2 L´3

L4

L1 L´2 L´3 L4

6×647.55 = 3885.3 m 10×647.55 – 215.85= 6259.65 m

16×647.55 + 215.85= 10576.65 m

33×647.55 = 21369.2 m

Tlinac L1/(c×Tlinac) L´2/(c×Tlinac) L´3/(c×Tlinac) L4/(c×Tlinac)

360 ns 36 58 98 198

300 ns 43.2 69.6 117.6 237.6

270 ns 48 77.33 130.66 264

240 ns 54 87 147 297

180 ns 72 116 196 396

IP

IP´

25/25

Final Remarks

If the damping ring circumference is chosen carefully, there is significant operational flexibility (up to a factor of two) in the number of bunches in a full ILC bunch train.

There is little benefit in a 650 MHz RF system in the damping ring, compared to a 500 MHz RF system, in terms of the flexibility in fill patterns.

In a ~ 6 km damping ring, operating with ~ 5400 bunches means eliminating any gaps. This could cause problems with ions or electron cloud. The rings can still operate with ~ 4000 bunches, with gaps of 64 ns.

A damping ring circumference of ~ 17 km would allow retention of the gaps with a large number of bunches.

Before a change to the baseline DR configuration is proposed (e.g. from 6 km to 17 km rings)

- the impact on the damping rings (ion effects, acceptance etc.) needs to be quantified;

- the benefits of lower bunch charge at the IP need to be quantified.

Lengths of other sections in the ILC (linacs, e+ transport lines, distance between IPs) must be chosen carefully if operational flexibility in the numbers of bunches is desired.

There are many solutions. Some example have been shown; there may be better solutions. It is not clear how to approach optimization of the parameters.