FCC-ee tuning WG meeting
R. Tomas reports that after the mid-term report a table of larger and more realistic misalignments will be published. It is expected that by 2025 BBA will be successfully performed and the optics corrected.
M. Hofer reports that he has been working on implementing T. Raubenheimer misalignments presented at the FCC-week in a MAD-X model. The implementation is done in a modular way. It is aimed to apply e.g. beam threading to these lattices. Following a question by F. Zimmermann, R. Tomas agrees to apply a 2.5sigma cut to these values and M. Hofer agrees to work on it. T. Raubenheimer suggests to check cross-correlation for misalignments over the circumference. T. Raubenheimer asks if the misalignments can be applied to the girder and not to the components. Furthermore, it should be clarified if 20 or 50 um pre-alignment of the girder should be assumed and F. Zimmermann adds that both options should be studied. Additionally, F. Zimmermann adds that these tolerances should be reduced to about 10um after BBA, preferably by trims coils.
F. Zimmermann asks if a different outcome would be expected if the dipole correctors are inside the quadrupole or next to it. R. Tomas reminds that it was reported by J. Bauche that horizontal dipolar corrections could be performed with the dipoles, since they have tapering coils. Furthermore, magnetic simulations have shown poorer magnet quality if dipole correctors are inside quadrupoles and sextupoles.
T. Charles presents tuning simulations with larger misalignments: girder = 200 um; arc sextupoles = 100 um; IR sextupole = 25 um. Although the majority of seeds has a vertical emittance below 1 pm, the beta-beating is not well corrected. With girder misalignments reduced to 150 um the beta-beating is only slightly better. T. Charles states that in these studies a trade-off between emittance and beta-beating must be made. F. Zimmermann asks if e.g. the algorithm yields 26% beta-beating, this could still be reduced with e.g. trim coils. T. Charles answers that no additional correction techniques have been considered so far and must be studied. T. Charles asks how exactly the misalignments should be applied. T. Raubenheimer answers that each end of the 6-7 m long girder should have a 150 um error, which effectively translates as a tilt. Furthermore, T. Charles asks if the proposed alignments include the variation of the magnetic center. T. Raubenheimer comments that about 1 um could be achieved with 10-20 % gradient variation.
B. Dalena reports on updates on orbit corrections for the FCC-ee booster, based on the segment-by-segments (SbS) technique. Errors are applied with and without girders. No tilts of girders are considered. The optics is not corrected. With girders the rms orbit is about 30% larger, and only 81 seeds converge (without 99 converge). F. Zimmermann suggests to also look at the emittance. K. Oide comments on the greater beta-beating fluctuations when no girders are assumed. B. Dalena comments that this could stem from seeds, which in the scenario with girders are not successful, and hence not shown.
Following a question by R. Tomas, T. Raubenheimer states that alignment, BBA, orbit and optics correction techniques and constraints should ideally be in a solid state by FCC-week 2024, followed by a follow-up discussion around September 2024. This allows to define mechanical and engineering by the end of 2025. K. Oide asks if a comparison of w and w/o girders are planned to be performed before deciding on the hardware requirements.
Y. Wu reports on orbit correction simulations for polarization studies. Simulations have to be done alternating corrections and adding components to the simulations (as RF and sextupole strength). With this approach polarization is good for seeds that succeed to converge. No IR errors could be considered so far as optics are not stable. Also comparing to assigning small errors without corrections gives similar performance when final rms orbit is similar. Some further realism is planed to be added in the simulations. Frank mentions that these 40um are in line with the targeted BBA performance. Rogelio added that this would need to be demonstrated with a more complete misalignment model (presented by Michael) as the current baseline has no movers and latest errors are 500um over 200m.
X. Huang pesents application of advanced BBA techniques to FCCee as follow-up from the results presented in the optics tuning workshop 2023 at CERN. The most important source of BBA performance error was identified to be the angle at the quadrupoles. Now an orbit bump with zero angle at the center of the quadrupole is used. 7 nearby orbit correctors are used per quad. The bump is steered by +-1mm. Initial misalignment error is 0.2mm. All quads have a BPM at the entrance. BPM error is set to 1 um. Systematic errors remain to 10-70 um. Reducing the number of quads probed simultaenously to 4 quads reduces systematic errors to max. of 20 um. Extrapolating to sextupoles requires minimizing rms orbit deformation which can be performed in few minutes.
Satya presents the development of IP tuning knobs for FCCee. Efficient dispersion and coupling IP knobs have been generated. Dy knob works up to Delta Dy=5mm at the IP. Skew quadrupoles are needed at the doublets in all IPs. Oide asks how to measure D_xy, which is indeed a challenge as discussed in one optics meeting (two methods proposed by Oide-san and Alain Blondel).
Peter presents studies of weak-strong beam-beam with non-linear lattice. A source of vertical emittance is needed to mimic imperfections. Wigglers are used for this. However significant discrepancies on the vertical emittance are seen between different codes. Issues identified so far: Emittance matching, bunch length also disagrees in simulation, Observation point at IP is not convenient for having a deformed phase-space. Frank asks if the strong beam is crabbed or not, which is not. Emittance is computed from rms values, not Gaussian fit (as Oide-san does): another source of discrepancy.