1. Injection design of CEPC
Cui Xiaohao
And CEPC Group
2018/11/12

2. 1. Introduction: A Linac accelerate electrons and positrons to 10 GeV.
A booster with the same circumference with the collider.
For high luminosity, top-up injection is needed.
For the simplicity and robustness of the injection system, a conventional off-axis injection is chosen as our baseline design.
For Higgs mode, an swap-out on axis injection is chosen as another choice, in case of small DA in the collider.

3. Some key parameters of the collider ring
Higgs W Z
Energy (GeV)
120 80
45.5
Bunch number
242
1524
12000
Bunch Charge (nC) 24
19.2
12.8
Bunch Current (mA)
17.4
87.9
461
Revolution Period (ms)
0.3336
Emittance x/y (nm)
1.21/0.0031
0.54/0.0016
0.17/0.004
Life time (Hour)
0.67
1.4 4

4. The geometry of the injection system

5. 2. Off-Axis Injection: Injection from Linac to booster
Booster ramping to Extraction Energy
Inject from booster to collider
Booster ramping to 10 GeV
repeat

6. Injection time structure:
Different injection time
Structure for different mode

7. This injection system fulfills the 3% top-up injection and full injection.
Mode
Higgs W Z
Injection Mode
Top-up
Full
Bunch number
242
1220
6000
Bunch Charge (nC)
0.72 1
0.384
0.55
Beam Current (mA)
0.5227
0.726
2.63
3.67
6.91 10
Current threshold
1 mA
4 mA
10 mA
Number of Cycles 2
Current decay 3%
Ramping Cycle (sec)
(Up + Down）
6.6
3.8
Filling time (sec)
（e+，e-)
25.84
39.6
275.2
Injection period (sec) 47
131
438
Full Injection time
10 min
15 min
2.2 Hour (collide from 230mA

8. Acceptance >4sxc + 6sxi + S
For Higgs:
e_x = 1.21 nm.rad; e_inject = 3.1 nm.rad
sx = 0.47 mm, s_inject = 0.45 mm
Septum = 2 mm
Acceptance > 13 s X’ X
Bumped
Stored beam
Septum
2mm
Injected beam
For W:
e_x = 1.21 nm.rad; e_inject = 1.59 nm.rad
sx = 0.47 mm, s_inject = 0.45 mm
Septum = 2 mm
Acceptance > 14 s
For Z:
e_x = 0.17 nm.rad; e_inject = 0.51 nm.rad
sx = 0.18 mm, s_inject = 0.16 mm
Septum = 2 mm
Acceptance > 16 s
Bump Height
Acceptance >4sxc + 6sxi + S

9. Current evolution in collider-Higgs
Top up  one injection (26s/35s) every 47s
Full injection from empty  10 min
Collider lifetime (non-collision) 924 hours: dominated Toucheck lifetime
Top up
Full injection

10. Current evolution in collider-W
Top up  one injection (46s) every 153s
Full injection from empty  15 min
Collider lifetime (non-collision) 105 hours: dominated Toucheck lifetime
Top up
Full injection

11. Current evolution in collider-Z
Top up  one injection (4.6min) every 7.3min
Full injection from empty  2.2 hours (bootstrapping start from half of the design current)
Collider lifetime (non-collision) 53 hours: dominated Toucheck lifetime
Top up
Full injection

12. Kickers and Septums for booster injection and extraction:
Component
Length (m)
Waveform
Deflection angle (mrad)
Field (T)
Beam-Stay-clear
H(mm)
V(mm)
Septum 2 DC
9.1
0.152 63
Kicker
0.5
Half_sin
0.034
Booster Extraction
Component
Length (m)
Waveform
Deflection angle (mrad)
Field (T)
Beam-Stay-clear
H(mm)
V(mm)
Septum 10 DC
10.4
0.41 20
Kicker 2
Half_sin
0.2
0.04

13. Kickers and Septums for collider injection:
Component
Length (m)
Waveform
Deflection angle (mrad)
Field (T)
Beam-Stay-clear
H(mm)
V(mm)
Septum
8.75 DC 14
0.64 20 7
0.32
3.5
0.16
1.75
0.08
Kicker 2
Half_sin
0.1
0.02

14. 3. On-Axis Injection: 1. Inject 242 small bunches into the booster.
Linac
booster
collider
2. Ramp the booster to high energy and inject several (7 for example) bunches from the collider into the booster.
Ramp up
booster
collider
3. Large bunches stay in the booster for 4 damping time(200ms), so that the injected bunches merge with booster bunches.
booster
collider
Repeat
4. Inject the merged large bunches back to the same empty buckets left by the last step.
5. Repeat the last steps.
booster
collider

15. Merge bunches in the booster
1. The booster is used as an accumulator ring.
2. Inject the large bunch in the collider into the booster, not the other way around.
3. off-axis injection and bunch mergence are performed in the booster, whose dynamic aperture is large enough.
4. With this on-axis injection, the required dynamic aperture in the collider is only 8 sx.
5. For W, Z mode, the dynamic aperture in the collider is large enough, and we don’t need this on-axis injection.

16. DA requirement: Off-axis 13sx X 15 sy On-axis 8sx X 15 sy
The required horizontal DA in the booster reduce from 13sx to 8sx.
The required vertical DA also decrease a little. X Y
On-axis
8sx X 15 sy X

17. Time structure of the booster:
Exchange bunches between the booster and collider 26 times.
2.4 s
5.2 s
5 s
Linac to booster
Ramp up
Ramp down
Injection cycles
Positron injection
Electron injection
3% current decay in the collider
The small bunch charge in booster is 3% of the charge of the colliding bunch in the collider.
With a 3% current decay, the injection period of CEPC should be less than 47s.
Current limit in the booster
The bunch charge in the collider is 24nC, which is less than the booster's single bunch charge limit of 100 nC.
At the beginning, only 7 bunches are injected into the booster, so that the current in the booster does not exceed the 1mA threshold limited by the RF power.
35.2 s

18. Injection from the collider to booster
Bunch after injection
Bunch before injection
Septum
A vertical off-axis injection
Transfer beam from the collider to booster vertically.
4-kicker bump in the booster.
Septum
kicker
Phase space of the first 100 turns

19. Injection from booster to the collider
Bunch after injection
Bunch before injection
A vertical on-axis injection
Septum
Transfer beam back to the collider vertically.
Septum
quad
kicker
Phase space of the first 100 turns

20. Longitudinal phase shift
We need to inject the bunch back into the same bucket after damping in the booster.
The booster and collider has the same circumference, which makes this easier.
Due to the different paths traveled by the injection beam and the circulating beam, a longitudinal deviation is introduced.
Booster
261.2 m
2.4 m
collider m
Longitudinal phase shift
The length of the transport line is meters, and the path difference between the injected beam and the circulating beam is only meters, or ns in time.
This is equivalent to a 17 degree phase shift in the booster, and is in the stable region. This phase shift can be damped by synchrotron radiation.
Or move the RF phase of the booster
Did experiment at BEPC II, we found that 200 ms is enough to move a 17 degree.

21. Beam loading effect is weak
Transient beam loading in the booster by the injection of large bunch and larger total beam current
With 7 large bunches (0.07mA) evenly distributed among the other small bunches. Max cavity voltage drop is 0.48 %. Max phase shift is 0.63 deg.
For 13 large bunches. Assume they are in a very short bunch train. Max cavity voltage drop is 5.8% . The maximum phase shift is 7.7 deg.

22. flip-flop instability
We considered bunch instability due to the absence of several bunches in the collider.
A beam-beam simulation shows that there is no flip-flop instability.
Simulation result of Zhang Yuan

23. 4. Summary
An baseline injection design from Linac to the collider is discussed, which can work for Higgs, W, Z mode.
An on-axis injection is designed to reduce the required dynamic aperture for Higgs energy injection.
Future work will be more focused on the optimization of the injection design, and considering the hardware.

24. Thank you For your attention!