1. ILC Damping Rings: Configuration Status and R&D Plans Andy Wolski Lawrence Berkeley National Laboratory January 19, 2006

2. 2/25 Baseline Configuration ItemBaselineAlternatives Circumference (e + ) 2  6 km (e - ) 6 km 1. (e + ) 6 km 2. (e + ) 17 km Beam energy5 GeV Injected emittance & energy spread0.09 m-rad & 1% FW0.045 m-rad & 2% FW Train length (bunch charge)2700 (2×10 10 ) - 4050 (1.3×10 10 ) Extracted bunch length6 mm - 9 mm Injection/extraction kicker technology Fast pulser/stripline kicker1. RF separators 2. Fourier pulse compressor Wiggler technologySuperconducting1. Normal-conducting 2. Hybrid Main magnetsElectromagneticPermanent magnet RF technologySuperconductingNormal conducting RF frequency500 MHz(650 MHz) Vacuum chamber diameter, arcs/wiggler/straights 50 mm/46 mm/100 mm

3. 3/25 Top Priority: Baseline lattice design by end of March 2006 Circumference6476.7163 m Energy5 GeV RF frequency500 MHz Harmonic number10802 Transverse damping time e + (e - )<25 ms (<50 ms) Normalized natural emittance5 µm Equilibrium bunch length6 mm Equilibrium energy spread<0.13% Momentum compaction~ 4×10 -4 Damping wiggler peak field1.67 T Damping wiggler period0.4 m Energy acceptance |  |<0.5% Dynamic aperture A x +A y <0.09 m-rad (up to |  |<0.5%) There are additional specifications on tunes and optics…

4. 4/25 Design studies of dogbone alternative will continue Circumference17227.9195 m Energy5 GeV RF frequency650 MHz Harmonic number37353 Transverse damping time e + (e - )<25 ms (<50 ms) Normalized natural emittance5 µm Equilibrium bunch length6 mm Equilibrium energy spread<0.13% Momentum compaction~ 1.5×10 -4 Damping wiggler peak field1.67 T Damping wiggler period0.4 m Energy acceptance |  |<0.5% Dynamic aperture A x +A y <0.09 m-rad (up to |  |<0.5%)

5. 5/25 Baseline lattice specification allows flexibility in fill patterns Ring circumference [m]6476.7163 Harmonic number10802 Ring RF frequency [MHz]500 Linac RF frequency [GHz]1.3 Linac pulse length [ms]0.97 Linac bunch spacing [linac RF wavelengths]468390351312234 Linac bunch spacing [ring RF wavelengths]18015013512090 Linac bunch spacing [ns]360.00300.00270.00240.00180.00 Ring bunch spacing [linac RF wavelengths]5.2 Ring bunch spacing [ring RF wavelengths]2 Ring bunch spacing [ns]4.00 Bunches per train45 Number of bunch trains60728090120 Gaps per train453022.5150 Gap length [ns]184.00124.0094.0064.004.00 Total number of bunches27003240360040505400 Particles per bunch [×10 10 ]2.071.731.561.381.04

6. 6/25 Alternative lattice specification also allows flexibility in fill patterns Ring circumference [m]17227.9195 Harmonic number37353 Ring RF frequency [MHz]650 Linac RF frequency [GHz]1.3 Linac pulse length [ms]1.03 Linac bunch spacing [linac RF wavelengths]540360180 Linac bunch spacing [ring RF wavelengths]27018090 Linac bunch spacing [ns]415.38276.92138.46 Ring bunch spacing [linac RF wavelengths]18126 Ring bunch spacing [ring RF wavelengths]963 Ring bunch spacing [ns]13.859.234.62 Bunches per train6918 Number of bunch trains415 Gaps per train12 Gap length [ns]60.00 Total number of bunches249037357470 Particles per bunch [×10 10 ]2.251.500.75

7. 7/25 ILC Damping Rings R&D Tasks List is in development 1.Parameter specifications and system interfaces 1.1Injected beams 1.2Extracted beams 1.3Fill patterns and timing issues 2.Beam dynamics 2.1Single-particle dynamics 2.2Multi-particle dynamics 3.Technical subsystems 3.1Injection/extraction kickers 3.2Damping wiggler 3.3Main magnets 3.4Orbit and coupling correction 3.5RF system 3.6Vacuum system 3.7Fast (bunch-by-bunch) feedback system 3.8Instrumentation and diagnostics

8. 8/25 ILC Damping Rings R&D Tasks List: Excerpt 2. Beam dynamics 2.1 Single-particle dynamics 2.1.1 Lattice design 2.1.1.1 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. 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. 10/25 R&D Tasks List Summary Spreadsheet (Excerpt)

11. 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/

12. ILC Damping Rings: Fill Patterns and Timing Issues Andy Wolski Lawrence Berkeley National Laboratory January 19, 2006

13. 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 1×10 10 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. 14/25 Scheme A: “Fixed bunch spacing” (increase no. of bunches by reducing the gaps) 12345 12345 12345 12345 12345 12345 12345123451234512345 12345 12345 12345 12345 12345 12345 12345 bunch separation in linac = T linac Extraction kicker fires regularly at intervals of T linac 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 T linac

15. 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. 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. 17/25 135 246 135 bunch separation in linac = T linac = 24 ring RF buckets Extraction kicker fires regularly at intervals of T linac 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 T linac bunch separation in linac = T linac = 12 ring RF buckets 135246135246135246 Scheme B: “Fixed gaps” (increase no. of bunches by reducing the bunch spacing)

18. 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. 19/25 Pros and cons… Scheme A: Fixed bunch spacing Provides 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 gaps Limited 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. 20/25 Lengths of different sections in ILC cannot be chosen arbitrarily If L 1, L 2, L 3 and L 4 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 (L LCM ) of the various linac bunch separations, L linac. L 1, L 2, L 3 and L 4 should then all be integer multiples of L LCM. e - source e - damping ringe + damping rings e - linac e + linac e + source IP L1L1 L2L2 L3L3 L4L4

21. 21/25 Lengths of sections are determined by linac bunch separation e - source e - damping ringe + damping rings e - linac e + linac e + source IP If L 1, L 2, L 3 and L 4 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 (L LCM ) of the various linac bunch separations, L linac. L 1, L 2, L 3 and L 4 should then all be integer multiples of L LCM. L1L1 L2L2 L3L3 L4L4 snapshot of bunch positions

22. 22/25 We can retain flexibility by choosing lengths carefully In example 1, the bunch separations in the linac are T linac = (360, 300, 270, 240, 180) ns. LCM(360, 300, 270, 240, 180) = 10800, or L LCM = 3237.8 m. This is inconveniently large. LCM(360, 300, 270, 240, 180) = 2160, or L LCM = 647.55 m. This is better. LCM(360, 300, 270, 240, 180) = 720, or L LCM = 215.85 m. This could be appropriate for the 2 nd IP. e - source e - damping ringe + damping rings e - linac e + linac e + source IP L1L1 L2L2 L3L3 L4L4

23. 23/25 Example, using 6476 m damping ring with 500 MHz RF frequency L1L1 L2L2 L3L3 L4L4 L1L1 L2L2 L3L3 L4L4 6×647.55 = 3885.3 m10×647.55 = 6475.5 m16×647.55 = 10360.8 m33×647.55 = 21369.2 m T linac L 1 /(c×T linac )L 2 /(c×T linac )L 3 /(c×T linac )L 4 /(c×T linac ) 360 ns366096198 300 ns43.272115.2237.6 270 ns4880128264 240 ns5490144297 180 ns72120192396 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. 24/25 Example, with 2 nd IP L1L1 L´2L´2 L´3L´3 L4L4 L1L1 L´2L´2 L´3L´3 L4L4 6×647.55 = 3885.3 m10×647.55 – 215.85 = 6259.65 m 16×647.55 + 215.85 = 10576.65 m 33×647.55 = 21369.2 m T linac L 1 /(c×T linac )L´ 2 /(c×T linac )L´ 3 /(c×T linac )L 4 /(c×T linac ) 360 ns365898198 300 ns43.269.6117.6237.6 270 ns4877.33130.66264 240 ns5487147297 180 ns72116196396 IP IP´

25. 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.