CEPC pretzel scheme study H. Geng, Y. Zhang, J. Gao, Y. Wang, D. Wang, G. Xu, Q. Qin, W. Chou, N. Wang, Y. Peng, X. Cui, Z. Duan, S. Bai, F. Su Beihang University, Beijing, China CEPC-SPPC Study Group Meeting September 2-3, 2016

Outline Introduction on pretzel scheme DA optimization with downhill simplex method DA optimization with MODE Combination with FFS Summary

Intro. on pretzel scheme In single ring collider, pretzel orbit is used to avoid the beam collision at positions except for the IPs Beam with off-center pretzel orbit will see extra fields in magnets, which will break the periodicity of lattice, and will significantly reduce the DA In CEPC, the pretzel orbit has been designed for 50 bunches Every 12-FODO cells were regrouped as a super-period in the arc, to correct the betatron aberration

Layout of the ring FODO cell length: 47.2 m (60/60degrees) ARC length:5852.8m Short straight: 18 FODO cells, 849.6m Long straight: 34 FODO cells, 1604.8m Circumference: 56.640km For 50 bunches: Every nearest parasitic collision points will have a phase advance of 4Pi.

DA results before adding pretzel Pure standard FODO cells, plus matching section adjusting working points DA: (30sx,450sy) for ±2% momentum spread

DA with pretzel (no FFS) DA tracking is done with SAD 240 turns (or 3 times transverse damping time ) is tracked The DA is (5sx*50sy) at ± 2.0% momentum spread DA: (5sx,50sy) for ±2% momentum spread

DA optimization with Downhill Simplex method In Oide’s code Sextupoles are adjusted DA is the only optimization objective Target modifications Quadrupoles are adjusted (to be done) DA, closed orbit, dispersion function, phase advance are the optimization objectives

Modifications Weight Objective function Dispersion Phase advance Closed orbit DA

Results No significant improvement found ! DA: (5sx,50sy) for ±2% momentum spread DA: (5sx,100sy) for ±2% momentum spread

A Multi-Objective coDE (MODE) DA result directly from SAD matching (use 0.3% coupling) Horizontal: ~5sx @0.0% dp/p ~4sx @2.0% dp/p

Results DA result after optimization (about 3-4days), use 13 families of sextupoles Horizontal: ~20sx @0.0% dp/p ~15sx @2.0% dp/p

Constraints have been set for orbit, dispersion etc. Beta, dispersion function and orbit of the optimized lattice

Check DA at different direction Not all directions are equally improved Other directions can be further optimized…

Combination with FFS One version of FFS (which has been optimized for the ring without pretzel orbit) is inserted to the lattice with pretzel The betatron and dispersion functions of the FFS are shown in the left plot The whole lattice of the ring is shown in the right plot Courtesy of Yiwei Wang, by*=3mm

Emittance change No pretzel orbit/FFS ex=6.8 nm , sp/p=0.13% With pretzel orbit ex=7.9 nm , sp/p=0.13% Increased by 16% With pretzel orbit, and FFS ex=21.1 nm , sp/p=0.17% Increased by 167% The contribution of FFS is too big, but should be managable by tuning the bending magnet strength in FFS.

DA with pretzel and FFS No DA after combining with pretzel and FFS Further optimization needed

Summary Downhill simplex method and a multi-objective code MODE has been used to optimize DA The DA for lattice with pretzel reached ~20sx @0.0% dp/p and ~15sx @2.0% dp/p The FFS has be inserted to the lattice with pretzel orbit Further optimization will be carried out……

Thank you !