1Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
PEP-II Interaction Region Upgrade
M. Sullivan
for the
Super-B Factory WorkshopHawaii
January 19-22, 2004
2Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Outline
•Present design
•Upgrade parameters
•Why upgrade the IR
•1st IR Upgrade Attempt
•Beta functions
•Crossing angle
•Present IR Upgrade Study
•SR fans and power
•SR backgrounds
•Summary
3Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Machine Parameters that are Important for the IR
PEP-II KEKBLER energy 3.1 3.5 GeVHER energy 9.0 8.0 GeVLER current 1.96 1.51 AHER current 1.32 1.13 A y
* 12.5 6.5 mm
x* 25 60 cm
X emittance 50 20 nm-radEstimated y
* 5 2.2 m
Bunch spacing 1.26 2.4 mNumber of bunches 1317 1284Collision anglehead-on 11 mradsBeam pipe radius 2.5 1.5 cm
Luminosity 7.21033 11.31033 cm sec
4Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
5Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
PEP-II Proposed Upgrade Plans
Now Projected UpgradeLER energy 3.1 3.1 3.1 GeVHER energy 9.0 9.0 9.0 GeVLER current 1.8 3.6 4.5 AHER current 1.0 1.8 2.0 A
y* 12.5 8.5 6 mm
x* 28 28 28 cm
X emittance 50 40 40 nm-radEstimated y
* 4.9 3.6 2.7 m
Bunch spacing 1.89 1.26 1.26 mNumber of bunches 1034 1500 1700Collision angle head-on head-on head-on mradsBeam pipe radius 2.5 2.5 2.5 cm
Luminosity 6.61033 1.81034 3.31034 cm sec
6Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Lower the vertical beta function to 6-7 mm
Keep maximum betas low
Lower the beam-beam effect from the parasitic crossings
Possibly get enough separation to allow filling every RF bucket
Why Upgrade (Initial motivations)
7Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Replace the last 20 cm of each B1 magnet with quadrupole field (50% stronger than QD1 field)
Introduce a crossing angle to recover beam trajectories so we don’t have to redesign or move any magnets. We need a certain amount of separation at QF2.
Relatively easy change to make
Moves the focusing closer to the IP
Initial upgrade attempt
8Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
0 2.5 5 7.5-2.5-5-7.5
0
10
20
30
-10
-20
-30
Meters
Ce
ntim
ete
rs2x1034 Interaction Region with a ±3.25 mrad Xangle
M. Sullivan Jun. 14, 2000
QD1
QD4QF5
QF2
QD4 QF5
QD1 QD1
QD1
QF2
Extra focusing
Extra focusing
9 GeV
9 GeV
9Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Beta functionsLow-energy beam
Present designbeta x* (cm) beta y*(mm) beta x max beta y max50 15 94 11235 12.5 135 13550 6 75 277
With Initial Attempt50 6 106 213
High-energy beamPresent design
50 15 520 45035 12.5 735 54050 6 528 1085
With Initial Attempt50 6 546 1057
10Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Crossing angle
Early last year, Ohmi-san from KEK announced that his beam-beam simulation code indicated a rapid luminosity degradation as a function of increasing crossing angle. Last summer, Yunhai Cai at SLAC confirmed Ohmi’s beam-beam result. The effect is most pronounced for very high tune shifts (~0.1).
Parasitic crossings
The introduction of a crossing angle increases the beam separation at the parasitic crossings which would lower the effect we presently see from parasitic crossings in by2 bunch patterns. Lowering y
* also increases parasitic crossing effects since the y at the PC is larger.
Crossing angle and parasitic crossings
11Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Plot of luminosity degradation as a function of increasing crossing angle (courtesy of Yunhai Cai)
12Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
0
0.1
0.2
0.3
0.4
0.5
0.6
0 2 4 6 8 10 12
LER First PC y,
l = 9 mm
by*=0.9 cmby*=0.8 cmby*=0.7 cmby*=0.6 cmby*=0.5 cm
2PC/I P
/2 (mrad)
0
0.1
0.2
0.3
0.4
0.5
0.6
0 2 4 6 8 10 12
LER First PC y,
l = 7 mm
by*=0.9 cmby*=0.8 cmby*=0.7 cmby*=0.6 cmby*=0.5 cm
2PC/I P
/2 (mrad)
0
0.1
0.2
0.3
0.4
0.5
0.6
0 2 4 6 8 10 12
HER First PC y,
l = 9 mm
by*=0.9 cmby*=0.8 cmby*=0.7 cmby*=0.6 cmby*=0.5 cm
2PC/I P
/2 (mrad)
0
0.1
0.2
0.3
0.4
0.5
0.6
0 2 4 6 8 10 12
HER First PC y,
l = 7 mm
by*=0.9 cmby*=0.8 cmby*=0.7 cmby*=0.6 cmby*=0.5 cm
2PC/I P
/2 (mrad)
LER PC tune shifts vs /2 for different y* normalized to
the IP tune shift for l (bunch length) = 9 and 7 mm
HER PC tune shifts vs /2 for different y* normalized to
the IP tune shift for l = 9 and 7 mm
The tune shift from the first parasitic crossing normalized to the main collision tune shift as a function of crossing angle and plotted for various y
* values for PEP-II (courtesy of Marica Biagini)
13Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Keep head-on collisions
Increase the beam separation at the 1st parasitic collision as much as possible
Allow for 6-7 mm y*
Do not change QF2 septum magnet
Present Working Design Constraints
14Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Stronger B1 magnet to increase separation at 1st parasitic crossing
20% stronger first 5 slices (first 12.5 cm with the weakest field and the largest lever arm)
Slightly increase the beam energy asymmetry 9.1 x 3.08 GeV
Stronger, closer QD1 magnets30% stronger slices for 1st 5 slices
Move radial ion pump behind B1 to behind QD1
Put higher strength focusing in present pump place
Minimal hardware change
Higher strength material has higher temperature coefficient
Present Working Design
15Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Modified Head-on design
16Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Orbit comparison with upgrade and present design
Work in progress
1st horizontal corrector
1st vertical corrector
+z side
LER beam
17Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
+z side
HER beam
1st horizontal corrector
1st vertical corrector
Work in progress
18Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Need to correct the position and angle in the x and y orbit on the +Z and –Z sides for each beam (8 orbits in total – two constraints for each orbit)
Too many variables for 8 total correctors
Allow adjusting the vertical collision point
Allow adjusting the horizontal collision angle
Allow the new sections of QD1 to have a variable offset (variable bending – but it affects both beams)
Allow (small) beam energy adjustments as long as we are still on the 4S resonance
Steps toward a new design
19Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
Beta functions (revisited)Low-energy beam
Present designx* (cm) y* (mm) x max (m) y max (m)50 15 94 11235 12.5 135 13550 6 75 277
With upgrade #150 6 106 213
With upgrade #250 6 99 21335 6 134 213
High-energy beamPresent design
50 15 520 45035 12.5 735 54050 6 528 1085
With upgrade #150 6 546 1057
With upgrade #250 5 570 97035 5 795 970
20Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
The proposed high beam currents (4.5A LER and 2A HER) will generate a large amount of SR in the IR
The HER vacuum elements were designed for 2A so the HER parts should be OK. There is some question about the High-Power Downstream Dump that absorbs the HER B1 SR power
There are 2 vacuum chambers that see the LER SR power that need to be looked at more closely
The LER downstream crotch chamber that sees B1 radiation
The upstream LER SR mask for the Be beam pipe. It sees upstream QD1 radiation
SR fans and power
21Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
22Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
23Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
The present power levels on the multi-tipped LER SR mask are about 30 W/mm at 2.1A beam current. This goes to 50 W/mm at 3.6A and 65 W/mm for 4.5 A of LER beam. This mask is under study. Presently it looks like this chamber may just work at 3.6 A but that it will need to be rebuilt for a 4.5 A beam
The crotch chamber design allowed for overlapping B1 radiation fans. This can only happen when the detector solenoid is off. We never intend to run high current beams with the solenoid off so we have some margin here.
SR fans and power (cont.)
24Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
The present LER SR mask (multi-tipped mask) does? (not completely checked yet) an adequate job of shielding the detector if we upgrade the B1 and QD1 magnets. We are trying to maintain the beam orbits such that the present design will work with minor modifications.
In any case, the extra power from the higher current LER beam means that SR backgrounds have to be studied in more detail
The very high beam current of the LER means that we need to check to make sure that back-scattered photons from the downstream crotch chamber do not strike the detector beam pipe
SR backgrounds
25Super-B Factory WorkshopJanuary 19-22, 2004
IR UpgradeM. Sullivan
The initial upgrade proposal replaced the last 4 slices of the B1 magnets with quadrupole field. This allows for lower beta y* values with a smaller increase in the maximum beta y.
The replacement of the B1 slices with quad field introduces a ± 3.3 mrad crossing angle at the IP which reduces the beam-beam effect at the 1st parasitic crossing. However, recent beam-beam simulations indicate a luminosity reduction for beams with a crossing angle.
An alternative proposal currently under study is to strengthen the IP end of QD1 effectively moving the center of the magnet closer to the IP. At the same time, increase the beam separation at the 1st parasitic crossing by increasing the strength of the initial B1 slices. This maintains the PEP-II head-on collision.
The high beam currents of the upgrade plans generate significant SR power in the IR that must be handled
SR backgrounds look like they can be controlled but have not yet been thoroughly studied
Summary