| Paper |
Title |
Page |
| TUPLT012 |
Adjusting the IP Beta-functions in RHIC.
|
1156 |
| |
- W. Wittmer, F. Zimmermann
CERN, Geneva
- F.C. Pilat, V. Ptitsyn, J. Van Zeijts
BNL, Upton, Long Island, New York
|
|
| |
The beta- functions at the IP can be adjusted without perturbation of other optics functions via several approaches. In this paper we describe a scheme based on a vector knob, which assigns fixed values to the different tuning quadrupoles and scales them by a common multiplier. The values for the knob vector were calculated for a lattice without any errors using MADX. Previous studies for the LHC have shown that this approach can meet the design goals. A specific feature of the RHIC lattice is the nested power supply system. To cope with the resulting problems a detailed response matrix analysis has been carried out and different sets of knobs were calculated and compared. The knobs are tested at RHIC during the 2004 run and preliminary results maybe discussed. Simultaneously a new approach to measure the beam sizes of both colliding beams at the IP, based on the tune ability provided by the knobs, was developed and tested.
|
|
| TUPLT013 |
Calculating LHC Tuning Knobs using Various Methods
|
1159 |
| |
- W. Wittmer, D. Schulte, F. Zimmermann
CERN, Geneva
|
|
| |
By measuring and adjusting the beta-functions at the IP the luminosity is being optimized. In LEP this was done with the two closest doublet magnets. This approach is not applicable for the LHC due to the asymmetric lattice and common beam pipe through the triplet magnets. To control and change the beta-functions quadrupole groups situated on both sides further away from the IP have to be used where the two beams are already separated. The quadrupoles are excited in specific linear combinations, forming the so-called tuning knobs for the IP beta-functions. We compare the performance of such knobs calculated by different methods: (1) matching in MAD, (2) inversion of the re-sponse matrix and singular value decomposition inversion and conditioning and (3) conditioning the response matrix by multidimensional minimization using Hessian method.
|
|