1. MEIC New Baseline: Part 10
No Bunched Beam Cooling during Collision
MEIC Accelerator R&D Meeting
January 4, 2016

2. Plan B if the Present Cooling Design Fails Completely
If the ERL circulator ring based electron cooler design fails completely
There will be no electron cooling in the collider ring (20 to 100 GeV)
It also means there will be no electron cooling during collision
Whatever the emittance after accumulation in the pre-booster (limited by space charge tune-shift) will be the final emittance for collision, assuming there is no further emittance degradation.
Without cooling during collision, the emittance will increase due to IBS. The IBS emittance growth time is also the luminosity life-time, and it should not be too short
Plan B
Add electron cooling in the pre-booster, after the energy is boosted to 3 GeV and before the beam is extracted for the large booster
Reduce the emittance to a lower level, but it should also not too small to cause too short IBS growth time in the collider ring
Electron cooling at 3 GeV is well below Fermilab cooler; and a new 2 MeV electron cooler COSY (3.8 GeV) is current under commissioning at Juelich, Germany
Response to Rolf Ent’s Request, 2/11/2013

3. MEIC Electron Cooling Rick Assessment & Plan B
Nominal
Total Failure
Plan B
Pre-booster
DC cooling for accumulation of positive ions at injection energy
Emitt. after accumulation
3.15 mm-mrad (vert. & horiz.) (limited by space charge)
Boosting energy to 3 GeV
DC cooling again
(1.5 MeV e energy)
Emittance at ejection
3.15 mm-mrad
1.2 mm-mrad
Ion collider ring
Multi-phase ERL-based electron cooling
At injection (20 GeV)
At top energy (up to 100 GeV)
During collision
Emittance at collision
0.35/0.07 mm-mrad
IBS growth time (x/y/s)
16/160/23 s
19/101/459 hours
2.6/150/38 hours
Proton bunch length (cm) 1 3
β* (x/y) (cm)
2/10
p (3/3), e (27.5/27.5)
p (3/3), e (10.4/10.4)
Beam-beam parameters
p (0.015), e (0.03)
p (0.001), e (0.0018)
p (0.0025), e (0.018)
Leslett tune-shift (proton)
0.06
0.0007
0.002
Luminosity (1032 /cm2/s)
5.6
0.41
1.1
Response to Rolf Ent’s Request, 2/11/2013

4. MEIC Phased Cooling Scheme
MEIC R&D Meeting, 3/12/2013
Loosely speaking, based on existing technologies
5.6
Add “weak” electron cooling &stochastic cooling (heavy ions) during collision
Full capacity electron cooling (ERL-circulator cooler)
Luminosity (1033 1/cm2/s)
3.3
Low energy DC cooling only at pre-booster injection
1.1
Add 3 GeV DC cooling at pre-booster
0.41

5. Why We Revisit This Issue?
Recent design development
Change of the baseline emittance design values (relaxed to 0.5 to 1 mm mrad from 0.35/0.07 mm mrad) makes the “weak cooling” scheme (i.e., no circulator ring) sufficient
 now the weak cooling is a part of the MEIC baseline
Large emittance leads to roughly a factor of 10 to 40 increase of IBS induced emittance growth time
 this opens a possibility of the concept of “no cooling” during collision (“Plan B” as discussed nearly two years ago)
New idea of a “stacking-cooling ring” (Peter McIntyre)
Peter recently proposed to add a cost-effective, fixed energy storage ring in the collider ring tunnel for stacking and cooling ion beams
Peter also suggested using the Fermilab DC cooler for this ring
These motivated us to revisit the “no-cooling” concept 5

6. IBS Induced Emittance Growth w/o Cooling (100GeV)
Bunch length cm 2
He Zhang, BetaCool simulation, 1/12/2015 6

7. IBS Induced Emittance Growth w/o Cooling (30GeV)
0.5
Bunch length cm 2
He Zhang, BetaCool simulation, 1/12/2015 7

8. Implication of Proton Emittance Degradation
Increase of proton emittance leads to large beam sizes
If there is no change of optics (particularly in IR)
Beam may not be able to pass through the final focusing quads (aperture issue)
Beam sizes may not match at IP (beam-beam issue)
Mitigations
Static beta-star
Pick a β* that satisfies both aperture and beam-beam requirements at the maximum emittance (at termination of collisions for a refill of the ring)
Then at all other emittance, the conditions should also be satisfied
Simple but not optimized in luminosity
Dynamic beta-star (revise of beta-squeezing)
Synchronizing β* (both electron & proton) along with proton emittance evolution such that the beam can always pass through the aperture and also reach beam spot matching
Luminosity is optimized 8

9. 100x5 GeV2 Luminosity under Dynamic β*
Beam-stay-clear (IR): +/-4 cm
Maximum betatron: m 9

10. 100x10 GeV2 Luminosity under Dynamic β*
Beam-stay-clear (IR): +/-4 cm
Maximum betatron: m 10

11. 30x3 GeV2 Luminosity under Dynamic β*
Beam-stay-clear (IR): +/-4 cm
Maximum betatron: m 11

12. Luminosity Performance
Beam energy
ERL Cooling
Matching Beam
Operation cycle
Luminosity
(1033 1/cm2/s)
GeV2
min
Starting
Terminal
Average
100 x 5
Yes
6x60
4.6 No 60
3.3
1.2
2.0
120
1.0
1.5
2.1
1.7
100 x 10
1.04
0.77
0.61
0.68
0.55
0.63
30 x 4
1.85
0.21
0.67
0.15
0.42 12

13. Luminosity Performance13

14. Cycle in the booster ring
Beam Formation Time
Beam formation cycle
Eject the expanded beam from the collider ring, cycle the magnet
Injection from the ion linac to the booster
Ramp to 2 GeV (booster DC cooling energy)
DC electron cooling
Ramp to 7.9 GeV (booster ejection energy)
Inject the beam into the collider ring for stacking
The booster magnets cycle back for the next injection
Repeat step 1 to 6 for 8 to 24 times for stacking/filling the whole collider ring (number of injections depends on the linac energy)
Cooling during stacking in the collider ring
Ramp to the collision energy (20 to 100 GeV)
Coasting/rebunching (or bunch splitting) to the designed bunch repetition rate
Nominal formation time: min
Cycle in the booster ring

15. Luminosity Performance
Average also includes beam formation time
Beam energy
ERL Cooling
Beam Spot matching
Operation cycle
Duty factor
Luminosity
(1033 1/cm2/s)
Integrated Luminosity
GeV
min
Starting
Terminal
Aver
fb-1
100x5
Yes
6x60
0.92
4.6
4.2
46.5 No 60
0.67
3.3
1.2
2.0
1.34
14.7
120
0.8
1.0
1.5
1.20
13.2
2.1
1.41
15.5
1.7
1.36
15.0
100x10
1.04
0.96
10.5
0.77
0.61
0.68
0.46
5.0
0.55
0.63
0.50
5.5
30x4
1.85
1.70
18.7
0.21
0.45
4.9
0.15
0.43
0.34
3.8
Assuming the beam formation time is 30 min
The integrated luminosity for 26 wk 70% duty factor is 11 fb-1 per 1033 cm-2s-1 15

16. Proposal of Combined-Function Dipoles for a Fixed Energy e-Cooler
ICR dipole
50 K vapor
4.5 K LHe
Peter McIntyre ~10/2015
Fermilab’s Recycler was designed as a fixed-energy 8 GeV storage ring, using combined-function permanent magnets.
The choice of permanent magnet made sense because it was installed in the Main Injector tunnel with no cryogenics.
For MEIC, we have designed a 1 T combined-function super-ferric dipole. Two CF dipoles are configured piggy- back on opposite ends of the two ICR dipoles, so that each ICR half-cell becomes a Cooler full-cell.
In the Recycler there was a troublesome field coupling between the Main Injector (cycling GeV) and the Recycler 8 GeV).
We modeled the same question for the mounting of the Recycler dipole piggy-back on the ICR dipole.
Result: no coupling at level of 0.5 unit.
Figures and texts from Peter McIntyre’s talk in MEIC Internal Cost Review, 12/192015 16

17. Advantages of A Stacking-Cooling Ring
While one proton/ion beam is in collision in the collider ring, another beam is formed in the stacking-cooling ring
Eliminating the waiting time for beam formation
Bring the duty factor to 100% (theoretically)
Pushing to extreme, we can replace the proton/ion beam in the collider ring as soon as the new beam is formed
The beam formation time is largely determined by multiple cycles in the booster ring.
Right now the nominal formation time is 30 min 17

18. Luminosity Performance with a Stacking-Colling Ring
Beam energy
ERL Cooling
Beam Spot matching
Operation cycle
Duty factor
Luminosity
(1033 1/cm2/s)
Integrated Luminosity
GeV
min
Aver
fb-1
100x5
Yes
6x60
0.92
4.20
46.2 No 30 1
2.57
28.3 60
2.14
23.5
120
1.73
19.0
100x10
0.96
10.5
0.72
7.9
0.68
7.5
0.63
6.9
30x4
1.90
20.9
1.08
11.9
0.67
7.4
0.43
4.7
The integrated luminosity for 26 wk 70% duty factor is 11 fb-1 per 1033 cm-2s-1 18

19. Integrated Luminosity w/ a Stacking-Cooling Ring

20. Optimization: Short Beam Cycle Time
If there is no cooling with a bunched beam in the collider ring, a short beam cycle provides an optimization of luminosity in both cases either with or without a stacking-cooling ring
In the present MEIC baseline, the beam cycle time is largely determined by
Number of cycles in the booster ring
the DC cooling time.
beam cycle ~ number of booster ring cycle x DC cooling time at 2 GeV
+ ramp time in the collider ring
(assuming the booster magnets can be ramped very fast, ~ a few seconds)
Presently, we consider 8x3 injections from the booster to the collider ring (we like to push down the ion linac energy, thus the space charge during accumulation in the booster is the main limit)
Cooling is much efficient at lower energy

21. He Zhang, BetaCool simulation, 2/3/2015
Where We Do DC Cooling?
Proton number 2.8x1012
Cooler length 10 m
Electron beam current 2A
e-beam radius 0.8 cm
Magnetic field 1T
@booster
~1.4x9+1~15 min
~(1.4/3)x27+1~15 min
~(1.4/6)x63+1~15 min
@collider ring
~35+1~36 min
He Zhang, BetaCool simulation, 2/3/2015

22. Cooling at A Lower Energy
In booster ring
@0.8 GeV
If the DC cooling is performed at a lower energy, the cooling time is reduced
In the pre-booster, at 0.8 GeV, the cooling time is ~15 s, leading to an injection cycle 16x63~16.8 min
In the collider ring, at 5 GeV, the cooling time is about 12 min
It seems there is no advantage to cool in the booster ring
Beam cycle time is ~ 15 to 20 min
In collider ring
@5 GeV
He Zhang, BetaCool simulation, 2/5/2015

23. What is the Beam Cycle Time?
LHC
As long as 5 hours  2 to 3 hours per proton ring
LHC ring is about 12 times longer than the MEIC collider ring
The equivalent filling time of the MEIC collider ring is about 15 min (no cooling)
RHIC
About 30 min for filling two proton rings after a normal dump of used beam  each ring is about 15 min (also no electron cooling in AGS and RHIC)
RHIC ring circumference is about 1.7 times longer than the MEIC ring
The equivalent filling time of the MEIC collider ring is about 9 min
MEIC
With cooling, a 15 min beam cycle is possible, but needs more study
Implication of a short beam cycle
Could be used for more frequent refill of the collider ring
Improvement of integrated luminosity by a stacking-cooling ring is also diminished

24. Conclusions of This Preliminary Study
Effective and integrated luminosities are still OK even without bunched beam cooling, thus providing a safe full-back positions
A stacking-cooling ring will eliminate the beam cycling time, potentially improves the integrated luminosity by 20% to 33%.
It is not clear whether there is other factor which will prevent frequent replacement of the hadron beams in a storage ring.
A short beam cycle time (15 to 20 minutes) decrease the advantage of a stacking-cooling ring.